摘要:

ANNUAL REPORTSON THEPROGRESS OF CHEMISTRYANNUAL REPORTSH. E. ARMSTRONQ, LL.D., F.R.S.G. M. BENNETT, M.A., Sc.D.H. M. DAWSON, D.Sc., Ph.D., F.R.S.H. W. DUDLEY, O.B.E., Ph.D., F.R.S.F. P. DUNN, B.Sc., F.I.C.A. C. G. EQERTON, M.A., F.R.S.H. J. T. ELLINOHAM, B.Sc., Ph.D.G. I. FINCH, M.B.E.J. J. Pox, O.B.E., D.Sc.W. E. GARNER, D.Sc., A.I.C.J. M. GULLAND, M.A., D.Sc.C. R. HARINGTON, M.A., F.R.S.W. N. HAWORTH, D.Sc., P.R.S.I. M. HEILBRON, D.S.O., D.Sc., P.R.S.J. T. HEWITT, D.Sc., F.R.S.C. N. HIFSHELWOOD, M.A., F.R.S.ON THEJ. KENDALL, M.A., D.Sc., F.R.S.J. KENNEH, D.Sc., F.R.S.R. P. LINSTEAD, DSc., Ph.D.T. M. LOWRY, C.B.E., D.Sc., F.R.S.W. H. MILLS, Sc.D., F.R.S.EMILE S. MOND.T. S. MOORE, M.A., B.Sc.G. T. MORGAN, O.B.E., D.Sc., F.R.S.J.R. PARTINGTON, M.B.E., D.Sc.F. M. ROWE, D.Sc., F.I.C.S. SMILES, O.B.E., D.Sc., F.R.S.S. SUGDEN, D.Sc.T. W. J. TAYLOR, M.A.J. F. THORPE, C.B.E., D.Sc., F.R.S.W. WARDLAW, D.Sc.PROGRESS OF CHEMISTRYH. R. AMBLER, B.Sc.R. P. BELL, B.A., B.Sc.J. D. BERNAL, M.A.E. J. BOWEN, M.A.B. A. ELLIS, M.A.E. H. FARMER, D.Sc.J. J. Fox, O.B.E., D.Sc.C. N. HINSHELWOOD, M.A., F.R.S.H. KING, D.Sc., P.R.S.G. A. R. KOK, M.A., D.Sc.A. G. POLLARD, B.Sc., A.I.C.(MISS) D. M. CROWROOT, B.A.S. GLASSTONE, D.Sc., Yh.D.F O X 1933.J. PRYDE, IM.Sc.B. W. ROBINSOX* PbD.A. RUSSELL, D.Se.A. S. RUSSELL, M.C., M.A., D.Sc.H. W. THOUPSON, M.A., P1i.D.W. WARDLAW, D.Sc.J. H. WOLFENDEN, MA.W. A. WOOSTER, Ph.D.I N. v. sIDGWICK, o.B.E., M.A., s~.D.,P.R.S.R.WHYTLAW-GRAY, O.B.E., Ph.D.,F.R.S.ISSUED BY THE CHEMICAL SOCIETYVOl. xxx.LONDON:T H E C H E M I C A L S O C I E T Y1934PRINTED IN GREAT BRITAIN BTRIUHABD W Y & SONS, UhlIl'ED.BUNGAY, 8uBBY)I.XCONTENTS.PAGEGENERAL AND PHYSICAL CHEMISTRY. By R. P. BELL, B.A., B.Sc.,E. J. BOWEN, M.A., C. N. HINSHELWOOD, M.A., F.R.S., H. W.THOMPSON, M.A., Ph.D., and J. H. WOLFENDEN, M.A.. . . 13INORGANIC CHEMISTRY. By N. V. SIDGWICK, O.B.E., M.A., Sc.D.,P.R.S., W. WARDLAW, D.Sc., and R. WHYTLAW-GRAY, O.B.E.,Ph.D., F.R.S. . . . . . . . . . . 82ORGANIC CHEMISTRY :-Part ~.-ALIPHATIC DIVISION. By E. H. FARMER, D.Sc. . . . 133Part II.-HOMOCYCLIC DIVISION.Part III.-HETEROCYCLIC DIVISION. By H. KING, D.Sc., F.R.S. . 226By G.A. R. KON, M.A., D.Sc., andA. RUSSELL, D.Sc. . . . . . . . . . 176gNALYTICAL CHEMISTRY. By H. R. AMBLER, B.Sc., B. A. ELLIS,M.A., J. J. Fox, O.B.E., D.Sc., and S. GLASSTONE, D.Sc., Ph.D. 268BIOCHEMISTRY. By A. G. POLLARD, B.Sc., A.I.C., and J. PRYDE, M.Sc. 305RADIOACTIVITY AND SUB-ATOMIC PHENOMENA. By A. S.CRYSTALLOGRAPHY (1932-33). By J. D. BERNAL, M.A., (MISS) I). M.CROWFOOT, B.A., B. W. ROBINSON, Ph.D., and W. A. WOOSTER,Ph.D. . . . . . . . . . . . 360RUSSELL, M.C., M.A., D.Sc. . . . . . . * 34TABLE OF ABBREVIATIONS EMPLOYED IN THEREFERENCES.Abbreuiutd Title.A . . . . . . .Acta Comm. Univ. Il’artu. .ActaPhytochim. . . .ActaBci.Pennicae . .Amer. Inst. blin. illet. Eizg.A m r . J . Sci. . . .Amer. 2llin. . . . .Anal.Pis. Quim. . . ,Analyst . . . . .Angew.Chem. . . .Annulen . . . .Ann.agron. . . . .Ann. Bot. . . . .Ann.Chim.. . . .Ann. Chim. awlyt. . .Ann. Fakif.. . . .Ann. Imt. Pmteur . .Ann. Imt. Platine . .Ann. Inst. Polytech. Lenin-Ann. Physik . . .Ann. Physique . . .Ann. Reports . . .Arch. Easenhiittenu-. . .Arch. Eleldrotechnik . .Arch. Pharm. . . .Arch. Ph9.s. biol. Chit)&.-Arch. Sci. phys. nat.Arch. Suikerind. Ned.-indidgradPhYS.Arhiv Hemiju . . .Arkanam Agric. Expt. Sta.Arkh. Biol. Nauk. . . Bull.Arkiv Kerni, Xin. Geol. .Arkiv Hat., Astron. FysdX: .Astrophys. J . . . .Atti B. Bccad. L i ? L C € i . .€3. . . . . . .Ber. . . . . .Biochem. J . . . . .Biochem. 2. . . . .Bot. Gaz. . . . .Bul. Chim.Soc. KomBne .B d l . A d . PolonnMe . .FULL TITLE.Abstracts in Journal of the Chemical Society (until1926) or in British Chemical Abstracts,* SectionA.Acta et Commentationes Universitatis Tartnensis(Dorpatensis).Acta Phytochimicrt.Acta Societatis Scientiarum Fennicae.American Institute of Mining and MetallurgicalAmerican Journal of Science.American Mineralogist.h a l e s de la Sociedad Espanijla Fisica y Quimica.The Analyst.Angewandte Chemie (formerly Zeitschrift fur ange-Justue Liebig’s Annden der Chemie.Annales agronomiques.Annals of Botany.Annales de Chimie.Annales de Chimie analytique et de Chimie appliquhe.Annales des Falsifications e t des Fraudes.h a l e s de d’Institut Pasteur.h a l e s de 1’Institut du Platine et des autres M6tauxAnnales de 1’Institut Polytechnique de Leningrad.Annalen der Physik.Annales de Physique.Annual Reports of the Chemical Society.Archiv fur das Eisenhuttenwesen.Archiv fiir Elektrotechnik.Archiv der Pharmazie.Archives de Physique biologique e t de Chimie-Archives des Sciences physiques et natnrelles.Archief voor de Suikerindustrie in Nederlandach-Arhiv za Hemiju i Farmaciju.Arkansas Agricultural Experiment Station Bulletins.Arkhiv Biologicheskikh Nauk (Archives des Sciencesbiologiques, U.S.S.R.).Arkiv for Kemi, Mineralogi och Geologi.Arkiv for Matematik, Astronomi och Fysik.The Astrophysical Journal.Atti (Rendiconti) della Reale Accademia Nazionaledei Lincei, classe di scienze fidche, matematichee naturali, Roma.British Chemical Abstracts,* Section B.Berichte der deutschen chemischen Gesellschaft.The Biochemical Journal.Biochemische Zeitschrift.The Botanical Gazette.Buletinul de Chimie p u i si applicatii a1 SocietgtiiBulletin Internationale de I ’AcadCmie Polonaise desEngineers, Technical Publications.wandte Chemie).Wcieux.physique des Corps organist%.Indie.Romhe de Chemie.Sciences e t des Lettres.xr is nut inserted in references to 1033.Thc yViii TABLE OF ABBREVIATIONS EMPLOYED IN TRE REFERENCES.Abbreviated Title.Bull. Acad. roy. Belg. . .Bull. Acad. Sci. U.R.S.S. .Bull. Assoc. Chim. Sucr. .Bull. Chem. SOC. Japan .Bull. Inst. Phys. Chem. Res.Bull.Soc. chim. . . .Bull. Sm. chirn. Belg. . .Bull. BOG. Chim.biol. . .Bull. SOC. franc. Min. . .Bull. Wagner Free Inst. .Bur. Stand. J . Res. . .Canadian J . Res. . .Chem. Erde . . . .Chem.and Ind. . . .Chem. Fabr. . . . .Chern.Listy . . . .TohyoChem.News . . . .Chem. Reviews . . .Chern. Weekblad . . .Chem. Zentr. . . .Chem.-Ztg. . , . .Chirn.et Id. . . .Coll. Czech. Chem. Cornin. .Coll. Symp. Monographs .Compl. Abs. Jap. Chem. Lit.Compt.rend. . . .Compt. rend. SOC. Biol. .Compt. rend. Trav. Lab.Danak Tidsskr. Farm. .Deut. Arch. Elin. Med. .Emiihr. Pflanze . . .Portschr. Min., lX?Gf. Pelr.Gazzettu. . . . .Geol. For. Fork . . .Georgia Agric. Zxpt. Sta.Giorn. C‘him. ind. appl. .Helv. Chim. Acta. . .Irrd.Chem. . . . .Ind.chim. . . . .Id. Eng. Chem. .I&. ~ n g .cbm. mi.) 1Indian J . Phy8ics . ,Int. Sugar J . . . .J . . . . . . .J . Agric. Ckm. SOC. JapanJ . Agric. Res. . . .J . Agric.Sci. . . .CarlsbergBull.FULL TITLE.AcadBmie royale de Be1gique.-Bulletin de la Classcdes Sciences.Bulletin de 1’AcadBmie des Sciences de 1’Union desRBpubliques SoviBtiques Socialistes.Bulletin de YAssociation des Chimistes de Sucrerieet de Distillerie.Bulletin of the Chemical Society of Japan.Bulletin of the Institute of Physical and ChemicalBulletin de la SociBtB chimique de France.Bulletin de la SociBtB chimique de Belgique.Bulletin de la SociBtB de Chimie biologique.Bulletin de la SociBtB franqaise de MinBralogie.Bulletin of the Wagner Free Institute of Science ofPhiladelphia.Bureau of Standards Journal of Research.Canadian Journal of Research.Chemie der Erde.Chemistry and Industry.Die Chemische Fabrik.ChemickB Listy pro V8du a PrSmysl.Organ de la“ CeskA ,yhernick& SpoleEnost pro V6du aPriimysl.Chemical News.Chemical Reviews.Chemisch Weekblad.Chemisches Zentralblatt.Chemiker-Zeitung.Chimie et Industrie.Collection of Czechoslovak Chemical Communications.Colloid Sym osium Monographs.Abstracts o f Japanese Chemical Literature, Com-Comptes rendus hebdomadaires des SBances deComptes rendus hebdomadaires de SBances de IsComptes rendus des Travaux du LaboratoireDansk Tidsskrift for Farmaci.Deutsches Archiv fiir klinische Medizin.Die Erniihrung der Mame.Bortschritte der Mineralogie, Kristallographie undGazzetta chimica italiana.Geologiska Forheningens i Stockholm Fordandlingar.Georgia Agricultural Experiment Station Bulletins.Giornale di Chimica industriale ed applicata.Helvetica Chimica Acta.The Industrial Chemist and Chemical Manufacturer.L’Industria chimica, mineraria e metallurgica.Industrial and Engineering Chemistry.Industrial and Engineering Chemistry : AnalyticalIndian Journal of Physics.The International Sugar Journal.Journal of the Chemical Society.Journal of the Agricultural Chemical Society ofJournal of Agrioultural Research.The Journal of Agricultural Science.Research, Tokyo.plete.l’Acad6mie des Sciences.SociM de Biology.Carlsberg.Petrographie.Edition.JapanTABLE OF ABBREVIATIONS EMPLOYED IN THE REFERENCES.ixAbbreviated 2’itle.J . Arner. Ceramic Soc. .J . Amer. Chem. SOC. . .J . Amer. Leather Chem.J . Amer. Pharm. Assoc. .J . Appl. Chem. Russia .J . Bact.. . . , .J . Bid. Chem. . . .J . Cell. Comp. Physiol. .J . Chem. Physics. . .J . Chim.physique . .J . Franklin Inst.. . .J . Gen. Chem. (U.S.S.K.) .Assoc.J . Ind. Eng. Chem. . .J . Indian Chem. 9oc. . .J . Infect. Dis. . . .J . Inst. Metals . .J . Lab. Clin. &fed: . .J.Landw. . . . .J . Opt. SOC. Amer. . .J . Pharm. Chim. . . .J . Pharm. SOC. Japan .J . Phys. Radium. . .J . Physical Chem. . .J . pr. Chem. . . . .J . Buss. Phys. Chent. Soc. .J . Sci. In&. . .J . Soc. Chem. Ind: . .J . SOC. Chem. Ind. Japan .J . SOC. Leather Trades’J . Text, Inst.. . .Jernlc. Ann. . . . .K. Norske Vidensk. Selsk.K. Svemka VetenskapsKlin. Woch. . . . .Kolloid - Z. . . .Landw. Jahri. . . .L’Ind. sacc. Ital. . . .Nasloboino Zhir. Dels .Medd. Centralanst. Fdrs6sksv.Mem. Coll. Agric. Kyoto .Mem. Coll. Eng. Kyushu .Mem. Coll. Sci. Kyoto .Mem. R. Accad. d’ltalia .Metallwirt. . . . .Mikrochem. . . . .Min. Mag. . . . .Chem.Forhundl.A M . Handl.Jwdbrub.FULL TITLE.Journal of the American Ceramic Society.Journal of the American Chemical Society.The Journal of the American Leather Chemists’Journal of the American Pharmaceutical Association.Zhurnal prikladnoi Chimii.Journal of Bacteriology.Journal of Biological Chemistry.Journal of Cellular and Comparative Physiology.The Journal of Chemical Physics.Journal de Chimie physique.Journal of the Franklin Institute.Journal of General Chemistry (U.S.S.R.) (formerlychemical part of the Journal of the Physical andChemical Society of Russia).Journal of Industrial and Engineering Chemistry(now Industrial and Engineering Chemistry).Quarterly Journal of the Indian Chemical Society.The Journal of Infectious Diseases.Journal of the Institute of Metals.The Journal of Laboratory and Clinical Medicine.Journal fur Landwirtschaft..Journal of the Optical Society of America.Journal de Pharmacie et de Chimie.Journal of the Pharmaceutical Society of JapanJournal de Physique et le Radium.The Journal of Physical Chemistry.Journal fur praktische Chemie.Journal of the Physical and Chemical Society ofRussia (now Journal of General Chemistry,U.S.S.R.).Journal of Scientific Instruments.Journal of the Society of Chemical Industry.Journal of the Society of Chemical Industry, Japan(K6gy6 Kwagaku Zasshi) .Journal of the International Society of LeatherTrades’ Chemists.The Journal of the Textile Institute.Jernkontorets h a l e r .Kongelige Norske Videnskabers Selskabs, Forhand-Kongliga Svenska Vetenskaps Akademiens Hand-Klinische Wochenschrif t.Kolloid-Zeitschrift.Landwirtschaftliche Jahrbucher.L’Industria saccarifera italiana.Masloboino Zhirovoe Dels (Oil and Fat Industry).Meddelande f r h Centralanstalten far Forsosks-Memoirs of the College of Agriculture, Kyoto Im-Memoirs of the College of Engineering, KyushuMemoirs of the College of Science, Kyoto ImperialMemorie della reale Accademia d’Italia.Metallwirtschaft, Metallwissenschaft, Metalltechnik.Mikrochemie.3Iineralogical Magazine and Journal of the Minera-Association.( Y akugakuzasshi) .linger.lingar.viisendet pB JordbruksomrAdet.perial University.Imperial University.University.logical Society.A H TABLE OF AkBREmATIONS EMPLOYICD IN T€IB &EFIZRENCJk?LWiek in-Wien .Abbreviated Title.Monatsh. .. . .Nach. Ges. Wiss. Gottingen .Naturwiss. . . . .Natuurwetensch. Tijds. .New Humps. Agric. Expt.New Jersey Agric. Expt.Norsk Geol. Tidsskr. . .Nuoz~o cim. . . * .P. . . . . . .Papier-Fabr. . . .Perf. & Essent. Oil Rec. .PJEanzenbau . . . .Sta.Bull.Sta. Bull.Pharm. Acta Helv. . .Pharm. Weekblad . .Pharm. Zentr. . . .Pharm. Ztg. . . . .Phil. Mag. . . . .Phil. Trans. . . . .Physical Rev. . . .Physikal. Z. . . . .Phytopath. . . . .Planta . . . . .Plant Physiol. . . .Proc. Camb. Phil. SOC. .PTOC. Imp. Acad. Tokyo .Proc. Indiana Acad. Sci. .Proc. K. Akad. Wetensch.Proc. Nut. Acad. Sci. . .Proc. Physical SOC. . .Proc. Roy. SOC. . . .Proc. Roy. Soc. Edin. . .Proc. SOC. Exp. Biol. Hed.AmsterdamProc. Tech. Sect. PaperProc. Univ. Durham Phil.Makers’ Assoc.SOC.PrzemysE Chern. . .Rec. trav. chim. . . .Rep. Brit. Assoc. . . .Rev. Mod, Physics . .Rev. Sci. Imtr. . . .Rocz. Chem. . . . .Sci. Papers Inst. Phys.Chena. Res. Tokyo . .Sci. Proc. Roy.Dublin SOC..Sci. Rep. Tdhoku Imp. Univ.Sitzunqsber. Akd. Wiss.FULL TITLE.Monatshefte fur Chemie und verwandte TheileNachrichten von der Gesellschaft der Wissenschaftenanderer Wissens chaf ten.zu Gottingen.Die Natunvissenschaften.Natuurwetensc happelij k Ti j dschrift .New Hampshire Agricultural Experiment StationBulletins.New Jersey Agricultural Experiment StationBulletins.Norsk Geologisk Tidsskrift.I1 Nuovo Cimento.Proceedings of the Chemical Society.Der Papier-Fabrikant.Perfumery and Essential Oil Record.Wissenschaftliches Archiv fur Landwirtschaft. Ab-Pharmaceutica Acta Helvetiae.Pharmaceutisch Weekblad.Pharmazeutische Zentralhalle.Pharmazeutische Zeitunn.teilung A. Pflanzenbau.Philosophical Magazine (The London, Edinburgh andPhilosophical Transactions of the Royal Society ofDublin).London.Physical Review.Physikalische Zeitschrift.Phytopathology .Zeitschrift fur wissenschaftliche Biologie. AbteilungE.Planta. Archiv fur wissenschaftlicheBotanik.Plant Physiology.Proceedings of the Cambridge Philosophical Society.Proceedings of the Imperial Academy of Japan.Proceedings of the Indiana Academy of Science.Koninklijke Akademie van Wetenschappen te Am-Proceedings of the National Academy of Sciences.Proceedings of the Physical Society of London.Proceedings of the Royal Society.Proceedings of the Royal Society of Edinburgh.Proceedings of the Society for Experimental Biologyand Medicine.Proceedings of the Technical Section, Paper Makers’Association of Great Britain and Ireland.Proceedings of the University of Durham Philo-sophical Society.Przemysf Chemiczny.Recueil des travaux chimiques des Pays-Bas et dela Belgique.Reports of the British Association for the Advance-ment of Science.Reviews of Modern Physics.Review of Scientific Instruments.Roczniki Chemji organ Polskiego TowarzystwrtScientific PaDers of the Institute of Phvsical andsterdam.Chemicznego.Chemicai Research, Tokyo.Scientific Proceedings of the Royal Dublin Society.Scientific Reports, Tdhoku Imperial University.Sitzungsberichte der Akademie der WissenschafteTABLE OF ABBREVIATIONS EMPLOYED IN THE REFERENCES. xiAbbreviated Title.Sitzungsber.preuss. Akad.Strahlenther. . . . .Wiss. BerlinSoil sci.. . . 4 Svensk Kem. Tidskr. . .Tidsskr. Kjemi Berg. . .Tottori Agric. Coll. Sci.Trans. Amer. Electrochem.Tram. Faraday Soc. . .Trans. Opt. Soc. .wiss. Ver58. Siemens- jionz'.Z.anal.Chem. . . .Kangew. Chem. . . .Z. anorg. Chem. . . .2. Biol. . . . . .2. Elektrochem. . . .PapersSOC.2. Krist. . . . .2. Metallk. . . .2. PJlanz. Dung. . . .2. Physik . . . .2. physikal. Chem. . .Z.physio1. Chern. . .Z. Unters. Lebensnz. . .Zentr. Bakt. Par. . . .FULL TITLE.Sitzungsberichte der preussischen Akademie derWissenschaften zu Berlin.Strahlentherapie : Mitteilungen aus dem Gebiete derBehandlung mit Roentgenstrahlen, Licht undRadioaktiven Substanzen.Soil Science.Svensk Kemisk Tidskrift.Tidsskrift for Kjemi og Bergvaesen.Tottori Agricultural College, Scientific Papers.Transactions of the American ElectrochemicalTransactions of the Faraday Society.Transactions of the Optical Society.Wissenschaftliche Veroff entlichungen aus dem Sie-Zeitschrift fur analytische Chemie.Zeitschrift fur angewandte Chemie (now AngewandteZeitschrift fur anorganische und allgemeine Chemie.Zeitschrift fur Biologie.Zeitschrift fur Elektrochemie (und angewandtephysikalische Chemie).Zeitschrift fur Kristallographie.Zeitschrift fur Metallkunde.Zeitschrift fur Pflanzenernahrung und Diingung.Zeitschrift fur Physik.Zeitschrift fur physikalische Chemie, StochiometrieHoppe-Seyler's Zeitschrift fur physiologische Chemie.Zeitschrift fur Untersuchung der Lebensmittel.Zentralblatt fur Bakteriologie, Parasitenkunde undSociety.mens-Konzern.Chemie).und Verwandtschaftslehre.Infektionskrankheiten

ISSN:0365-6217    

DOI:10.1039/AR9333000001

出版商:RSC    

年代:1933

数据来源: RSC

摘要:

TABLE OF ABBREVIATIONS EMPLOYED IN THE REFERENCES. xiERRATA.VOL. 29, 1932.Page Line269 ‘i-lO* SOT “The presence of glubathione in plant organs duringactive vegetative growth is also recorded in an examination ofObelia by E. J. Lund,*S who reports maximum accumulations in,the active tips of stems and roots.” read “The presence ofglutathione in plant organs during active vegetative growth isrecorded by E. J. L ~ n d . ~ ~ Also in Obelia an accumulation ofglutathione is indicated in the apical ends of the stems.”287 3* for ‘‘ has suggested ” read ‘‘ criticises the suggestion.”_______ __ ___ * From bottom

ISSN:0365-6217    

DOI:10.1039/AR9333000011

出版商:RSC    

年代:1933

数据来源: RSC

摘要:

ANNUAL REPORTSON THEPROGRESS OF CHEMISTRY.GENERAL AND PHYSICAL CHEMISTRY.THE plan of dealing with a limited number of special subjects hasagain been followed this year. Regret must therefore be expressedthat many important contributions t o physical chemistry gounmentioned. The advantage of a certain continuity of treatmentseems, however, too great to be sacrificed even to an attempt a tcomprehensiveness. The great achievements in sub-atomic physicsare not dealt with in this section since this year there is a specialsection devoted to such matters, but work on the heavier isotope ofhydrogen is dealt with under the heading electrochemistry, inconnexion with the electrochemical method of concentrating theheavier isotope in water. The discussions of solubility phenomenaand of spectroscopic terms from a general point of view are mattersalready perhaps overdue.Some pages have once again been devotedto chemical kinetics, but since the scope of the subject has beeninterpreted rather elastically, it may be hoped that its representa-tion is not out of proportion to the amount of activity which hasbeen displayed in that field. C. N. H.1. SOLUBILITY AND RELATED PHENOMENA.The experimental work treated in this section deals in generalwith the equilibrium of a given molecular species between twophases, i.e., partition coefficients. In the particular case when oneof the phases is a pure solid or liquid, we are dealing with solubilityin the usual sense of the word, and if one of the phases is gaseousthe problem is one of partial or total vapour pressure.From a,theoretical point of view there is no distinction between these cases.The chief interest of such measurement lies in their relation to theforces of attraction and repulsion between molecules, which aretheoretically predictable in some simple cases. Tho fact that aspecies i is in equilibrium between two phases at constant temperatur14 GENERAL AND PHYSICAL CHEMISTRY.and pressure is most conveniently expressed by the equality ofthe partial molal free energies (or chemical potentials) B’i of thespecies in the two phases. For any dilute fluid phase we can writewhere piOO refers to an arbitrary standard state of the pure com-ponent i , and hence depends only on T , while pio depends onlyupon T and the constituents of the phase; ci is the concentrationand fi the activity coefficient.It is convenient to divide theproblem into two parts : (a) The value of Fi0 in different media, i.e.,the distribution of a solute between two infinitely dilute phases,and ( b ) the value of fi in relatively dilute solutions. The twoproblems become inseparable for concentrated solutions, but canbe treated separately for many purposes.We shall treat first recent work on the so-called solubility ofvapours in compressed gases. The activity of any component ina condensed phase varies with the total pressure according to theequation a log aAlaP = sl,/BT, where ijA is the partial molal volumeof A in the condensed phase. If the pressure is applied by meansof a second gas By the mole fraction of A in the gas phase will beproportional to its activity only as long as the interaction betweenA and B is negligible. For all real cases the attractions betweenA and B lead to a concentration of A much higher than that calcu-lated from the gas laws.This effect, which depends upon thenature of B, can be considered as a solubility of A in compressedB, and may prove of great value for obtaining information aboutforces between unlike molecules. The first work of this kind isdue to F. Pollitzer and E. Strebe1,l who measured the concentrationof vapour in equilibrium with liquid water and carbon dioxide inpresence of various gases up to 200 atmospheres pressure. Sub-sequent investigations have been carried out by A.T. Larson andC. A. Black2 (liquid ammonia in presence of nitrogen mixtures upto 1000 atmospheres), E. P. Bartlett 3 (water in presence of N, + 3H,up to 1000 atmospheres), E. Lurie and L. J. Gillespie (BaC12,8NH,in nitrogen up to 60 atmospheres), A. Eucken and F. Bresler5(carbon disulphide in various gases up to 80 atmospheres), and H.Braune and F. Strassmanns (iodine in carbon dioxide up to 50atmospheres).1 2. physikal. Chenz., 1924, 110, 768; A., 1925, ii, 104; see also I. R.McHaffie, Phil. Mag., 1926, [vii], 1, 561 ; A., 1926, 365.a J . Amer. Chem. SOC., 1925, 47, 1015; A., 1925, ii, 501.3 Ibid., 1927, 49,65; A,, 1927, 207.4 Ibid., p. 1146; A., 1927, 616.f Ibid., 1929,143, 225; A., 1929, 1229.2. phyaikd. Ohm., 1928,134,230; A., 1828, 828BELL SOLUBILITY AND RELATED PHENOMENA.15M. Randall and B. Sosnick have shown how activity coefficientscan be calculated from results of this kind, and A. Eucken andF. Bresler have treated their results in terms of the van der Waalsequation. They find approximate confirmation of Berthelot’srelation a12 = 2/alla,, between the attraction constants. F.Londong has recently given a theoretical treatment of the dis-persion forces between unlike molecules and arrives a t a similarrelation.The molecular model which has attracted most attention in thetheory of solution is that of the ion considered as a sphere with acharge at its centre. The corresponding simplification for thesolvent is to consider it as a continuous medium possessingthe macroscopic dielectric constant of the solvent.M. Born lo wasthe first to consider the electrostatic energy of such a system ; heshowed that in transferring an ion from a medium A to a mediumB, the electrical work is given byFe = (~~/2r)(l/DA - l/DB) . . . . (2)where E and r are the charge and radius of the ion, and D A and D,the dielectric constants of the media.11 This result is obviouslyallied to the well-known fact that salts are more soluble in liquidsof higher dielectric constant, and that this effect is more markedfor multiply charged ions. Considerable difficulties arise, however,in relating Fe to the quantities in equation (1). For equilibriumdistribution of a solute between two phases A and B, TA = FB,and equation (1) givesF A 0 - F B O = - RT log cA/cB .. . . (3)It is natural at first sight to equate this difference to Ye in equation(2). However, if A and B are different solvents the value of cA/cBdepends upon the concentration scale used l2 (e.g., whether we choosemole fractions or volume concentrations), and hence equation (3)needs further justification. Attempts have been made 13 to justify7 J. Amer. Chem. Soc., 1928, 50, 967; A,, 1928, 688.8 LOC. cit. ; see also J. 5. van Law, 2. physikal. Chern., 1929, 145, 20’7 ; A,,9 2. physikal. Chcm., 1930, 11, [B], 221 ; A,, 1931, 149.10 2. Physik, 1920, 1, 45; A., 1921, ii, 166; see Ann. Reports, 1926, 23, 29.11 It should be noted that both Born and Fajans interpret the value of theabove expression as the “ heat of solution of the ions,” i.e., the total energychange.The nature of the process involved makes it, however, certain thatthe quantity obtained is the free energy-see, e.g., N. Bjerrum and E. Larsson,2. physikal. Chem., 1927, 127, 358; A., 1927, 828.1930, 161.12 See R. P. Bell, J., 1932, 2906; A., 127.13 See, e.g., 0. Gatty and A. Macfarlane, Phil. Mag., 1932, 13, 292 ; A., 1932,22716 GENERAL AND PHYSICAL CHEMISTRY.the use of a particular concentration scale in equation (3), but itis necessary to depart from strict thermodynamics, and theassumptions made are of doubtful validity. (Considerations ofthis kind will, of course, apply to any comparison between thermo-dynamic data and molecular models.)We can, however, proceed as follows in evaluating the electricalcontribution to the work of transfer.Consider distribution of theion deprived of its charge between the two phases, and let theequilibrium concentrations be (cA) and (cB). Then as beforeand hence( P A o ) - (FBO) = - BT log (CA)/(cB)For dilute solutions this expression is independent of the unitsin which cA and cB are expressed, and can be legitimately equatedt o Fe in equation (2). The concept of " an ion without its charge "is of course somewhat artificial, and it is obviously not in generalpossible to obtain experimentally (cA) and (c,) in equation (4).There is, however, sufficient similarity between some classes ofions and uncharged molecules to make this method of value.N. Bjerrum and E.Larssonl* calculated the distributioncoefficients of a number of ions between ethyl alcohol and water,partly from existing solubility data, and partly from measurementsof E.M.P. by E. Larsson.15 They showed that equations (2) and(3) lead in many cases t o reasonable values for the ionic radii.They found, however, that for large organic ions the effect of thehydrocarbon part of the ion was of great importance, and thatsolvation must be taken into account in other cases. F. K. V.Koch 16 found that a similar treatment of the solubilities of the silverhalides in water and the alcohols gave ionic radii agreeing with thecrystal lattice values, but that in other solvents there was norelation between the dielectric constant and the free energies oftransfer.A. Macfarlane and (Sir) H. Hartley17 found a generalqualitative agreement between the Born equation and the freeenergies of transfer (calculated from E.M.P. measurements) inwater and methyl and ethyl alcohols, and similar results were14 2. physikal. Chem., 1927, 127, 358; A., 1927, 828.15 Diss., Lund, 1927.17 Phil. Mag., 1932, 13, 425; A., 1032, 230.J., 1930, 1551; A., 1930, 1107; J . , 1928, 209; A., 1928, 230.Cf'. also A. E. Broilsky,Phyaikal. Z., 1929, 30, 665; A., 1929, 1391BELL : SOLUBILITY AND RELATED PHENOMENA. 17obtained by Bronsted for the solubilities of a number of cobalt-ammines in water and methyl alcohol.ls Analogous conclusionshave been reached from a study of the effect of added non-electro-lyte upon the solubility of salts.lg It is found that the resultslead to feasible values for the ionic radii, but that the effect dependsupon the nature of the added electrolyte as well as upon its dielectricconstant.All the authors mentioned in this paragraph haveneglected (deliberately or implicitly) all " non-electrical " con-tributions to the work of transfer and used equation (3) withvolume concentrations.A. Lannung 2* has measured the solubility of the rare gases in anumber of solvents, and the solubility of the alkali halides in acetone.Bjerrum21 has made use of the great similarity between the raregases and the alkali and halogen ions to deduce from Lannung'sresults the electrical part of the free energy of transfer, and tocompare it with equation (2).The alkali halides are appreciablysoluble in only a very limited number of solvents : in this respectthe tetraethylammonium halides offer greater possibilities.Bjerrum and E. J6zefowiczZ2 have measured the solubility ofthese salts in a number of solvents ranging from benzene to water.As an uncharged analogue of the tetraethylammonium ion theyhave measured the distribution of tetraethylsilane 23' between thevapour phase and the solvents employed.has employed crystal-latticeenergies, the Nernst heat theorem, and the Born equation to derivean expression for the solubility of the solid in terms of the ionicradii. By equating this expression to the observed solubility,values are obtained for the ionic radii, which may be comparedwith the crystal lattice radii. The values obtained are all greaterthan the lattice radii, and vary from solvent to solvent.Thegeneral conclusion is that the Born equation gives a qualitativepicture of the results in solvents of similar chemical types (e.g.,water and the alcohols), but fails when applied to a large rangeIn treating these results, Bjerrum18 J. N. Bronsted, A. Delbanco, and K. Volqvartz, 2. physikal. Chem., 1932,162, [A], 1932 ; A., 26.19 See V. I(. LaMer and F. H. Goldman, J. Amer. Chem. SOC., 1931,53, 473;A., 1931, 419; B. B. Owen, ibid., 1933, 55, 1922; A., 670; C. F. Failey, ibid.,1933, 55,4374.20 J . Amer. Chem. SOC., 1930, 52, 68; A., 1930, 406; 2. physikal. Chem.,1932, 161, [A], 235; A., 1932, 1197.21 Trans. Faraday SOC., 1927, 23, 446; A., 1927, 1028; cf.also ref. (24).22 2. physikal. Chem., 1932, 159, [A], 194; A., 1932, 457.23 The true analogue of the N(C,H,),' ion is C(C2HS)4 : this is, howevor, very24 Chemistry at the Centenary Meeting of the British Association, 1931, p. 34.t l i fficult i; o prepare18 GENERAL AND PHYSICAL CHEMISTRY,of solvents. This failure is obviously due to the inadequacy ofthe picture employed-in particular, the replacement of the solventby a continuous dielectric, and the neglect of solvation.25 Thediscrepancies are particularly marked when water is comparedwith solvents of very different chemical type, e.g., hydrogen fluoride,hydrogen cyanide, and liquid ammonia.26 This has been emphasisedespecially by K. Fredenhagen,27 who considers that the electrostaticpicture should be entirely abandoned in favour of the conceptionof the chemical solvation of ions.Fredenhagen's own treatmentof solution equilibrium is, however, open to serious criticism.28A particularly ingenious attempt to eliminate the "non-electrical" factor is due to J. N. BrOn~ted,~~ who measured thesolubility in nine solvents of the saltand of the isomeric non-electrolyte Co(NO&(N&)3. The ratio ofthe solubility of these two substances depends qualitatively onthe dielectric constant in the manner predicted by the Born equation,but there is nothing like quantitative agreement.Since there is a thermodynamic connexion between the chemicalpotentials of the components of a mixture, the Born picture of anelectrolytic solution can also be employed to calculate the effectof electrolytes upon the solubility of non-electrolytes. This isknown as the salting-out effect, and was last dealt with in theseReports in 1926,30 since when the theory has been worked out ingreater detail by P.Debye31 and P. Gross,32 and a large number25 Cf. the evidence obtained from ionic mobilities, Ann. Reports, 1930, 27,326.26 For recent solubility measurements in these solvents, see K. Freden-hagen, 2. plbysikal. Chem., 1927, 128, 1 ; A , , 1927, 936; Fredenhagen and G.Cadenbach, ibid., 1930,146, [A], 254; A., 1930,421 ; Fredenhagen, ibid., 1933,164, [A], 176; A., 566; 2. anorg. Chem., 1930,186, 1 ; A., 1930, 637; P. A.Bond and M. V. Stowe, J . Amer. Chem. Soc., 1931, 53, 30; A., 1931, 297;A.J. Schattenstein and A. Monossohn, 2. anorg. Chem., 1932, 20'7, 204; A.,1932, 990; H. Hunt, J . Amer. Chem. SOC., 1932, 54, 3509; A., 1932, 1197;Hunt and L. Boncyk, ibid., 1933, 55, 3528; A., 1112; M. Linhard and M.Stephan, 2. physikal. Chem., 1933,163, [A], 185; 167, [A], 87; A., 456.2 7 2. physikal. Chem., 1927, 128, 1 ; A., 1927, 936; ibid., 1928, 134, 33;A., 1928, 1316; ibid., 1929, 140, [A], 65, 435; A., 1928, 397, 513; ibid., 1929,141, [ A ] , 195; A,, 1929, 648; ibid., 1931, 152, [A], 321; A., 1931, 430; 2.anorg. Chem., 1930,186, 1; A., 1930, 537.28 See H. Hammerschmid and E. Lange, 2. physikal. Chem., 1931, 155, 85;A., 1931, 1010; ibid., 1932,159, 100; A., 1932, 467.39 Chemistry at the Centenary Meeting of the British Association, 1931, p.39.31 2. physikal. Chem., 1927, 130, 56; A., 1927, 1141.32 Monatsh., 1929, 53, 445; A., 1930, 150.Ann. Reports, 1926, 23, 28BELL : SOLUBILITY AND RELATED PHENOMENA. 19of experimental investigations have been carried out .33 Reviewsand discussions of experimental data have been given by M. Randalland C. F. Failey,3* G. S ~ a t c h a r d , ~ ~ and P. M. Gross.36If the dielectric constant D of the solution is a linear functionof the concentration rn of the non-electrolyte, i.e.,D=DO(l - pm) .(where Do is the dielectric constant of the pure solvent, and p aconstant), then Debye’s theory leads to the expressions = so(l - 07) . . . .where so is the solubility of the non-electrolyte in pure solvent, ands its solubility in salt solution of concentration y.The constant 0is a complicated function of the valencies of the ions, the dielectricconstant, the temperature, the constant p in equation (5), and themean radius of the ions. It is again found that when the experi-mental data are inserted in this equation, feasible values areobtained for the ionic radii, but that both the non-electrolytes andthe electrolytes exhibit individualities not predicted by the simpletheory.37 Most electrolytes decrease the dielectric constant ofwater, so that both p and CJ in equations ( 5 ) and (6) are positivequantities, and the solubilities should decrease on addition ofelectrolyte. This is, of course, the usual salting-out effect, and itis of particular interest to note that Gross, Schwarz, and Iser 38found that the solubility of hydrogen cyanide (which increases thedielectric constant of water) is increased by the addition of salt,is., it is “ salted in.”In the absence of a net electric charge on a molecule, the simplest83 See, e.g., K.Endo, J . Chem. SOC. Japan, 1926, 47, 374; A., 1926, 729(phenol); W. Herz and E. Stanner, 2. physikal. Chem., 1927, 128, 399; A.,1927, 1020 (various substances) ; S. Glasstone and W. It. Hodgson, J., 1927,635; A., 1927, 416 (aniline); J. S. Carter and R. K. Hardy, J., 1928, 127;A., 1928, 243 (m-cresol) ; N. Schlesinger and W. Kubasowa, 2.physikol. Chem.,1929, 142, 25 ; A., 1929, 874 (ethyl acetate) ; G. Claxton and H. M. Dawson,Proc. Leeds PhiZ. SOC., 1929,1, 416; A., 1929, 996 (phenol); K.Linderstrom-Lang, Compt.rend. Trav. Lab. Carkberg, 1929,17,No. 13; A., 1929, 1139 (quinoland benzoquinone); E. F. Chase and M. Kilpatrick (jun.), J . Amer. Chem.SOC., 1931, 53, 2889; A., 1931, 1010; E. Larsson, 2. physikal. Chem., 1930,148, [A], 148, 307; A., 1930, 995; ibid., 1931, 153, [ A ] , 299; A., 1931, 431(benzoic acid); P. Gross, ibid., 1929, 6, [B], 215; A., 1930, 989 (dichloro-ethane and -propane); Gross, K. Schwarz, and M. Iser, Monatsh., 1930, 55,287, 329; A., 1930, 989 (acetone and hydrogen cyanide).34 Chem. Reviews, 1927, 4, 271.35 Ibid., 3, 383; Trans. Faraday Xoc., 1927, 23, 455; A., 1927, 1028.36 Chern. Reviews, 1933, 13, 91.See specially P. M. Gross, Zoc. cit., ref. (36).38 L O C . cit., ref. (33)20 GENERAL AND PHYSICAL CHEMISTRY.distribution is that of a dipole, and it i s also possible in this case tocalculate the electrostatic work of transfer of a molecule from onemedium to another by making simplifying assumption^.^^ Thistheory has been applied to the heat of solution of water in differentsolvents (calculated from solubility measurements at differenttemperatures) and agrees approximately with experiment Allother contributions to the work of transfer are neglected, which isonly permissible when the dipole energy is large, Le., when thedipole moment is high and the radius of the molecule small.Martinand his collaborators 41 have measured the partial vapour pressuresof mixtures of benzene with some of its polar derivatives, and haveinterpreted their results in terms of the dipole energy of the mole-cules in different media.The effects observed are, however, small,and (as the authors themselves realise) it is not legitimate to neglectother factors in this case.In the majority of liquid mixtures the most important forces areno doubt the so-called van der Waals or dispersion forces, which donot depend upon any permanent charge configurations in themolecules.42 These forces can be calculated theoretically in simplecases, and the results of such calculations have been applied withsome success t o the equation of state of ga~es.4~ It is not yetpossible to apply the theory quantitatively to solutions, on accountof the complex spatial distribution of the molecules, but in somecases interesting comparisons can be made between pairs of mole-cules or series of similar compounds.For instance, J. N. Bronsted 44has introduced the conception of " isochemical " series of molecules,i.e., molecules which differ essentially in size but not in chemical'nature. Examples of this are the higher normal hydrocarbons,ethyl valerate and ethyl sebacate, etc. It is reasonable to supposethat in such a series the work of transfer between two given phasesis additive for different parts of the molecule, e.g., the value forC16H34 should be twice that for C8H1,; similarly the value forC02Eta (CH2),*C02Et should be twice that for CH,*( CH,),*CO,Et.3S Ann. Reports, 1931, 28, 39; see also A. R. Martin, Phil. Mag., 1929, 8,550; A,, 1929, 1389; Nature, 1931,128,456; A., 1931,1222; J.N. Bronsted,loc. cit., ref. (29).40 R. P. Bell, J., 1932, 2905; A., 127.4 1 A. R. Martin and B. Collie, J., 1932, 2658; A., 1932, 1197; Martin and42 See Ann. Reports, 1930, 27, 17.49 See F. London, Zoc. cit., ref. (9) ; J. C. Slater and J. G. Kirkwood, PhysicalRev., 1931, [ii], 37,682 ; A., 1931,675 ; K. Wohl, Z. physikal. Chem,, 1928, 133,305; A., 1928,827 ; ibid., 1929,2, [B], 77 ; A., 1929,251 ; &bid., 1931, Boden-steinFestband, 807; A., 1931, 1222; i b a . , 1931,14, [B], 36; A., 1931, 1216;G. Briegleb, ibid., 1933, 23, [B], 105; A., 1231.C. M. George, J., 1933, 1413.44 Z. physlsikal. Chem., 1931, Bodenstein Pestband, 257; A., 1031, 1221BELL : SOLUBILITY AND RELATED PHENOMENA.21Bronsted 45 adduces a certain amount of experimental support forthis principle from measurements of solubility and vapour pressure.The principle should be more rigidly true in the extreme case ofcolloid particles of different sizes. Similar ideas underlie I. Lang-muir’s “ principle of independent surface action,” 46 according towhich the potential energy of a molecule in a liquid is composedadditively of the interactions at the different parts of the “inter-face ’’ between the molecule and its surroundings. J. A. V. Butler 4’has recently found an approximate confirmation of Langmuir’stheory in his results for the vapour pressures and solubilities of thenormal aliphatic alcohols in water. There appears to be a constantincrement in the work of transfer for each CH, group added.So far, we have only considered dilute solutions, and haveneglected the solute-solute intermolecular forces which account forthe term RT log fi in equation (1).In principle, it should bepossible to deduce this term from the theory of different types ofintermolecular forces, and hence to develop a general theory of thedeparture of more concentrated solutions from their behaviour a tinfinite dilution. Actually, however, it is only for long-range forcesthat it is possible to separate the terms solute-solute, solvent-solvent, and solute-solvent interaction without a precise knowledgeof the statistical distribution of the molecules in the liquid. Theonly long-range forces which we meet with in practice are inter-ionic forces, and hence it is only for electrolyte solutions that thisproblem has been treated with much success. The original theoryof the thermodynamics of electrolyte solutions was that of Debyeand Huckel, last treated in these reports in 1927.48 Since then agreat deal of work has been done in examining the theoretical basisof the original theory, in extending it to more concentrated solutions,and in testing the theory experimentally.The original derivation of the Debye-Huckel expression involved(besides the idealised model of an ionic solution employed) thefollowing approximations or assumptions.(a) The potential of the ionic atmosphere round any ion obeysthe Poisson-Boltzmann differential equation,where Y? is the mean potential at any point and ni is the meannumber of ions of charge ei per C.C.45 J.N. Bronsted and E. Warming, 2. physikal. Chem., 1931, 155, [A], 343;A., 1931, 1119.4 6 Colt. Syrnp. Monographs, 1925, 3, 48.4 7 Proc. Roy. SOC., 1932, [A], 135, 366; A., 1932, 459; J. A. V. Butler,48 Ann. Reports, 1927, 24, 22 ; 1926, 23, 21 ; 1925, 22, 27.D. W. Thornson, and W. H. Maclennan, J., 1933, 674; A., 77222 GENERAL AND PHYSICAL CHEMISTRY.( b ) The distribution of charge round any two ions is the sum of(c) In solving the differential equation (7), it is assumed thatthe charges which would be induced by the ions separately.PiY/kT<l . . . . . * (8)Much work has been done in examining these fundamental pointsin the Debye-Huckel theory. R. H. Fowler 49 was the first to pointout that equation (7) might be only an approximation, but con-cluded that the " fluctuation terms '' could be neglected in dilutesolution.L. Onsager has shown that assumptions (a) and (b)above are not in general self-sonsistent except in very symmetricalcases. However, these difficulties disappear in solutions so dilutethat equation (8) is true and the original Debye-Huckel limitinglaw holds. Valuable confirmation of this is afforded by the workof H. A. Kramer~,~l who has treated the problem from the stand-point of general statistical mechanics without introducing thePoisson-Boltzmann relation. I n order to obtain convergent seriesin the mathematical treatment, it is necessary to assign finite radiito the ions. Kramers shows, however, that in sufficiently dilutesolution the activity coefficient is independent of the ionic radii,and at infinite dilution his expression becomes identical with thesimple Debye-Huckel equation.The correctness of the latterequation as a limiting law at infinite dilution is thus confirmed byseveral lines of critical e~amination.5~Matters are much more complicated for concentrated solutions,since the assumptions of the original derivation no longer hold inthis case. In particular, assumptions (a) and (b) above becomemutually exclusive (except in certain symmetrical cases).53 Froma purely theoretical point of view it is therefore doubtful how muchsignificance can be attached to attempts to obtain a completesolution of the Poisson-Boltzmann equation (7).54 (Failure to49 Trans.Faraday Soc., 1927, 23, 434; A., 1927, 1028. L. Onsager (Chenz.Reviews, 1933, 13, 73) has recently shown that Fowler's " fluctuation terms "are not actually negligible, but are cancelled out by a.nother term omitted byFowler.50 Phygikal. Z., 1927, 28, 277; A., 1927, 157.5 1 Proc. K . Akad. Wetensch. Amsterdam, 1927,30, 145; 4., 1927, 626. TheSame conclusion is arrived a t by P. van Rysselberghe, J . Chem. Physics, 1933,1, 206.52 See, further, G. Scatchard, Physikal. Z., 1932, 33, 22; A., 1932, 127;Scatchard and J. G. Kirkwood, ibid., p. 297; A., 1932, 467.68 For an excellent account of these difficulties, see L. Onsager, ref. (49).54 Such attempts are dealt with in Ann. Reports, 1932,29,24. The completesolution of the Poisson-Boltzmann equation has also been obtained forunsymmetrical electrolytes, see V.K. LaMer, T. H. Gronwall, and L. J. Greiff,J . Phyaical Chem., 1931,35,2245; A., 1931, 1127BELL : SOLUBILITY AND RELATED PHENOMENA. 23realise these fundamental difficulties had previously led to dis-crepancies between different methods of deriving the electricalfree energy of the solution from the average potential of an ion.) 55The error involved in such elaborations can only be estimated byan extension of Kramers’s statistical treatment to more concentratedsolutions : this involves great mathematical difficulties, and hasnot yet been attempted. It may be mentioned that Bjerrum’s theoryof ion association 56 is probably less open to such criticism, since hedoes not apply the Poisson-Boltzmann equation t o pairs of ions closetogether, which is just when the usual assumptions are least valid.The Debye-Huckel limiting law (which is unaffected by theabove considerations) is readily applicable to mixtures of electro-lytes; for the activity coefficient of the salt AB in any diluteelectrolyte mixture it gives the expressionwhere zA and zB are the valencies of the ions A and B, and p isdefined by t,c = +&yi2, the summation being extended over allions present in the solution; cc is a constant depending on thetemperature and the dielectric constant of the medium.Since inthe presence of solid salt the chemical potential of AB is constant,determinations of the solubility of salts in presence of other electro-lytes constitute a simple experimental test of equation (9), and thefirst reliable evidence for the Debye-Hiickel theory was obtainedin this way.57 Later work has in many cases led to confirmationof the theory both in aqueous 68 and in non-aqueous 59 solvents.The experimental results appear in some cases to be at variancewith equation (9) even at very great dilutions, notably for saltsof unsymmetrical higher-valency types 6o and solvents of lowi5 5 See L.Onsager, loc. cit., ref. (53).5 6 See Ann. Reports, 1932, 29, 32.67 Ann. Reports, 1925, 22, 32.5 8 See V. K. LaMer, C. V. King, and C. F. Mason, J . Amer. Chein. SOC., 1927,49, 363; A., 1927, 314 (cobaltammines); S. Popov and E. U‘. Nouman, J.Physical Chem., 1930, 34, 1853 ; A., 1930, 1107 (silver chloride) ; B.H. Peter-son and E. L. Meyers, J . Amer. Chem. SOC., 1930, 52, 4853; A., 1931, 309(cupric iodate) ; E. Larsson and B. Adell, 2. anoTg. Chern., 1931, 198, 344 ; A.,1931, 566 (silver benzoate and acetate).59 A. L. Robinson, J. Physical Chern., 1928,32,1089 ; A., 1928, 944 (potassium bromide in acetone); J. W. Williams, J . Amer. Chern. SOC., 1929, 51,1112; A., 1929, 649 (cobaltammines in methyl alcohol); L, A. Hamen andJ. W. Williams, ibd., 1930, 52, 2759; A., 1930, 1121 (cobaltammines in ethylalcohol-water mixtures) ; J. N. Bronsted, Zoc. cit., ref. (18).60 V. K. LaMer, C. V. King, and C. F. Mason, J . Amer. Chern. Soc., 1927,M410; A., 1927, 314; LaMer and F. H. Goldman, ibid., 1929, 51, 2632; A .1929, 138624 UEXIRBL AND PHYSICAL CHEMISTRY.dielectric constant.61 It is, however, just in these cases thatapproximations (a), (b), and (c), above, break down outside anextremely small concentration range, and it can, in fact, be shownthat the range in which the limiting law will be obeyed is in manycases too dilute to be studied experimentally.We shall not dealfurther here with attempts to explain the behaviour of electrolytesoutside this range. As mentioned above, considerable uncertaintyattends any more elaborate solution of equation (7), and theevidence for and against assuming incomplete dissociation (in anysense) has been recently dealt with in these reports.62 It may,however, be mentioned that solubility data in aqueous and in non-aqueous solution have been interpreted successfully both by theGronwall-La&-Sandved extended theory 63 and by assumingincomplete dissociation.84A principle of very general application in correlating activitycoefficients in mixtures of electrolytes is the " principle of specificinteraction," first stated by J.N. Bronsted 65 and recently derivedin a more lucid manner by E. A. Guggenheim.66 If we consider asalt AB present in very small amount in a solution of a salt XY,then the activity coefficient of AB (written faB(xu)) depends essenti-ally upon the interactions between A and X, A and Y, B and X,and B and Y. The principle of specific interaction assumes thatthese interactions are determined entirely by the electric chargeswhen the two ions are of the same sign, but that when the two ionsare of opposite signs, specific effects (e.g., the radii of the ions, shortrange forces, etc.) will be of importance. This obviously dependsupon the fact that similarly charged ions are very rarely close toone another, while oppositely charged ions tend to spend anappreciable time in close proximity.Many of the observed devia-tions from the Debye-Hiickel theory may be explained in thisway. For instance, LaMer, King, and Mason 67 found that althoughthe solubility of the salt [Co( NH3) .*' [Co (NH3),( NO,),C,OJ,' obeysthe Debye-Hiickel limiting law in solutions of potassium nitrate61 C. A. ~ R U S and R. P. Seward, J. Physical Chem., 1928, 32, 1294; A.,1928, 1182; J.W. Williams, Chern. Reviews, 1931, 8, 303.84 Ann. Reports, 1932, 2s. 21-29.63 R. P. Seward and W. C. Schumb, J . Amer. Chern. SOC., 1930, 52, 3962;A., 1930,1612 ; Seward and C. H. Hamblet, ibid., 1932,54,554; A., 1932,339 ;A. W. Scholl, A. W. Hutchison, and G . C. Chandlee, ibid., 1933, 66, 3081; A.,1014; E. W. Neuman, ibid., 1932, 54, 2195; A., 1932, 801; ibid., 1933, 55,879; A., 466; LaMer, Gronwall, and GreifF, Zoc. cit., ref. (54).84 H. E. Blayden and C. W. Davies, J., 1930, 949; C. W. Davies, ibid.,p. 2410; A., 1930, 860; 1931, 40.118 J . Amer. Chern. Soc., 1922, 44, 877; A., 1922, ii, 481.Oti Report of the 18th Scandinavian Naturalist Congrew, Coponhagen, 192967 LOG. cit., ref. (60)BELL : SOLUBILITY AND RELATED PHENOMENA. 25and barium chloride, yet there are large deviations in potassium ormagnesium sulphate and potassium ferricyanide, i.e., when the twomultiply charged ions present are of opposite sign.The theoryalso leads to the following quantitative predictions :(a) Zor two solute salts AB and A‘B with a common anion B,the ratio faB(xy)lfa.B(xy) is independent of the nature of both B and X.(b) Similarly, for two solvent salts XY and XY, the ratioJpAB(x.p) is independent of both A and Y.( c ) For a salt AB present in small amount in a mixture of x partsof XY and 1 - x parts of X’Y’ (the total salt concentration beingconstant), log fAB is a linear function of x. This rule was firstadvanced independently by B r o n ~ t e d , ~ ~ but its relation to thespecific interaction principle was later shown by E.Guntelberg 68and by Guggenheim.69These predictions (in which, of course, the r8les of anion and cationare interchangeable) were shown by Bronsted to be in accordancewith a large mass of data for the solubility of cobaltammines insalt solutions up to 0.1N. Subsequent solubility measurementsby LaMer and his collaborators 70 have further confirmed theprinciple, and extended it to include salts of higher valency types.Only for cadmium chloride 71 does the principle fail, probablyowing to the formation of complex ions.The prediction of the thermodynamic properties of non-electrolytemixtures over the whole concentration range must be regarded asone of the most difficult problems in the theory of solutions.Inspite of a great deal of theoretical and experimental work, littlereal progress has been made since the matter was last dealt with inthese reports.72 The subject is still generally treated from thepoint of view of deviations from the perfect solution, which is bestdefined by the equationFi- FiO= RTlog Nd . . . (10)where Ni is the mole fraction, and pio refers to pure component(and not to the dilute solution). There seems, however, to be agrowing suspicion that equation (10) may not represent in anySense a (‘ normal ” behaviour of a liquid mixture, but may ratherbe due to an unusual and complex compensation of several effects.It is well known that if equation (10) is true for a binary mixtureover a certain temperature range, then the two components will6 5 E.Giintelberg, 2. physikccl. Chem., 1926,123, 199; A., 1926, 1207.6s Id. A. Guggenheim, bc. cit., ref. (66).7 0 LaMer, King, and Mason; LaMer and Goldman, Zocc. cit., ref. (60);LctMer and R. 0. Cook, J . Amer. Chrn. SOC., 1929, 51,2622; A., 1929, 1386.71 H. B. Friedmttn and V. K. LaMer, ibid., 1931, 53, 103 ; A., 1931, 300.I2 Ann. Reports, 1925, 22, 3726 GENERAL AND PHYSICAL CHEMISTRY.mix without energy or volume change over the same temperaturerange. E. A. Guggenheim 73 has pointed out that the converse ofthis proposition is not in general true, and has shown that equation(10) can only be deduced by making much more drastic assumptionsabout the energy and volume relations. He has also attemptedto derive the same equation by statistical methods,74 but histrwtment would only appear to be valid for the whole concentrationrange when the two components have equal molecular volumes,and the same applies to W.Heitler’s treatment of the liquid as acubic lattice.75In attempting to account for deviations from the laws of perfectsolution, it is certainly necessary at present to consider only solutionsin which the constituents are as far as possible non-polar, and themolecular attractions are essentially due to the van der Waals forces.For such systems J. H. Hildebrand76 has proposed the equationwhere p is a constant for a given mixture and temperature. Thisequation is in good agreement with the experimental data forsolubilities and vapour pressures in a large number of solutions,which Hildebrand terms “ regular solutions.” He suggests thatsuch solutions are characterised by an absence of chemical andorienting effects, and that the entropy of mixing is the same as fora perfect solution of the same concentration.The last assumptionis by no means self-evident unless the molecules are quite sym-metrical, but it (or its equivalent) is made in every theory of devia-tions from perfection. It then follows immediately that the termpN,2 should represent g2, i.e., the partial molal heat of mixing ofcomponent 2; this is in agreement with the experimental data insome cases, especially when the components are n~n-polar.’~ J. J.van Laar and R. Lorenz 78 have derived equation (11) on the basisof the van der Waals equation of state; this, however, leads to avalue of p independent of temperature, which is not in accordancewith experiment.W. Heitler 79 has arrived at the same equationby treating the structure of the liquid as a cubic lattice, but hisdeduction only applies to the case of equal molecular volumes.No entirely satisfactory derivation of equation (11) has been given;moreover, since the term expressing deviations from ideality mustF2 -F2”= RT logN,+ pNI2 . . . (11)73 J . Physical Ohm., 1930, 34, 1761; A., 1930, 1120.74 Proc. Roy. SOC., 1932, 135, [A], 181; A., 1932, 338.75 Ann. Physik, 1926, 80, 630; A., 1926, 1006.76 J . Amm. Chem. SOC., 1929,51,66; A , , 1929, 266.77 See Ann. Reports, 1925, 22, 38.78 2. anorg. Chein., 1925, 145, 239; d., 1925, ii, 866.LOC.cit., ref. (76); cf. J. H. Hildebrand and E. J . Salstrom, J . Amer.Chern. Soc., 1932, 54, 4257; A,, 26BELL : SOLUBILITY AND RELATED PHENOMENA. 37depend on the forces between the molecules, we shouid expect itto be a function of the volume of the solution. Any exact treatmentof the problem is exceedingly diEcult owing to lack of knowledgeof the spatial distribution of the molecules in liquids. Our onlyreal knowledge of this last problem is derived from X-ray studies,80and Hildebrand and S. E. Wood81 have recently made a veryinteresting attempt to apply the results of H. Menke 82 on thestructure of mercury to calculate the thermodynamic properties ofbinary mixtures of symmetrical molecules. The deduction involvesseveral approximations, but is noteworthy as the first attempt toemploy our knowledge of the actual structure of liquids in this field.The final equation isF2 - F2* = RT log N , + A,,vp .. . (12)where A12 = V,((E,IV,)* - (E11~1)*l2where v1 is the volume fraction of component 1, E l , E2 are thecohesive energies, and V,, V2 the volumes per mol. of the purecomponents. Hildebrand shows that this equation agrees excel-lently with his data for the solubility of iodine in titanium, silicon,and carbon tetrachlorides and carbon disulphide. Exactly thesame equation has been derived by G. Scatchard 83 by making asomewhat arbitrary assumption about the energy of mixing. Hefound that the equation agreed better than equation (11) with theform of the curves for s, - F2, but that the actual magnitude ofthe deviation term was often in error.It should, however, benoted that the systems he has chosen involve less symmetricalmolecules than those treated by Hildebrand. Scatchard foundfair agreement for the observed and the calculated solubilities ofnaphthalene in a number of solvents.If the molecular volumes are equal, equation (12) reduces to (ll),and it is difficult to distinguish between them experimentally exceptin the case of widely different molecular volumes. The constantsp and A,, in these equations are a measure of how far the systemdeparts from ideal behaviour. Hildebrand has previously suggestedthat 8 should be approximately proportional to the difference ininternal but equation (12) suggests that the difference80 See Ann.Reports, 1929, 26, 306.81 J . Chern. Physics, 1933,1, 817.82 Physikal. Z., 1932, 33,593; A., 1932, 986; cf. also 0. Kratky, ibid., 1933,83 Chern. Reviews, 1931, 8, 321.84 See Ann. Reports, 1925, 22, 38; J. H. Hildebrand and M. E. Dorfman,J , AWT. Chem. SOC., 1927, 49, 729; A., 1927, 406; (Miss) M. E. Dice andRildebrand, ibid., 1928, 50, 3023; A., 1929, 131; W. Westwater, H. W.b’rantz, and Hildebrand, Phpical Rev., 1928, 31, 135 ; A., 1928, 228 ; Hilde-brand axid. J. M. Carter, J . Amer. Chem. SOC., 1932, 54, 3592; A., 1932, 1197.34,482; A., 76828 GENERAL AND PHYSICAL CHXMXSTRY.of the energy densities is a, more reliable criterion. Por a fluidobeying the van der Waals equation, the two are identical,indicating at least a qualitative truth in Hildebrand’s originaltheory of solubility.R. P. B.2. ELECTROCHEMISTRY.The Heavy Isotope of Hydrogen.The hydrogen isotope of mass 2 has been the subject of a con-tinually increasing volume of research during the year. A varietyof methods of concentrating the isotope has been explored andthe uncertainty as to the proportion of the isotope in ordinaryhydrogen has been largely removed. A preliminary survey of thephysical properties of the oxide (“ heavy water ”) has been carriedout and a few extremely interesting observations of the chemicalbehaviour of the isotope have already been recorded. As thequantity available for investigation multiplies (and water contain-ing 0.5% of the heavy isotope is already available commercially),results of the greatest chemical importance are likely to accumulaterapidly.An isotope of such marked individuality is most convenientlytreated as a new element and deserves a name of its own.Itsdiscoverers have suggested the name “ deuterium,” the correspond-ing ion being called the “ deuton ” ; there are, however, variousdifliculties connected with these names and Lord Rutherford hassuggested “ diplogen ” for the atom and “ diplon ” for the ion, andit seems likely that, among British chemists at any rate, the latternomenclature will find more favour. Fortunately both alternativeslead to the symbol “ D ” for the new atom.The following processes have so far been shown to lead t o anappreciable concentration by fractionation of the heavy isotope :(1) Distillation of liquid hydrogen (the original method) ;(2) electrolysis of water in acid and in alkaline solution; 2, 3, 4 (3)adsorption of hydrogen on charcoal ; (4) distillation of water ;(5) adsorption of water on charcoal; (6) reduction of water by1 H.C. Urey, F. G. Brickwedde, and G. M. Murphy, PhyBicaZ Rev., 1932, [ii],* E. W. Washburn and H. C. Urey, Proc. Nat. Acad. Sci., 1932,18,496 ; A.,8 G. N. Lewis, J. AWT. Chm. SOC., 1933, 55, 1297; A., 442.* C . H. Collie, Nature, 1933, 132, 568.5 H. S. Taylor, A. J. Gould, and W. Bleakney, Physical Rev., 1933, 43,6 G. N. Lewis and R. E. Cornish, J. Amer. C h . BOG., 1933,5$2616; A.,40,l; A., 1932, 554.1932, 894.496.793.E. W. Washburn and E.R. Smith, J . Ohem. PityeiCS, 1933, 1,426WOLFENDEN : ELECTROCHEMISTRY. 29hot iron ; a$ (7) diffusion of hydrogen through palladium ; 10 (8)displacement of hydrogen in acids by zinc."Of these methods, only the second has so far been used forthe systematic concentration of large quantities of diplogen. Thetechnique G. N. Lewis and his collaborators employed is describedin a paper by him and R. T. Macdonald.12 Their raw material waswater from an old electrolytic bath containing about twice the normalconcentration of the heavy isotope. A solution of approximatelyM/B-sodium hydroxide was electrolysed with nickel electrodesuntil the volume was reduced ten-fold; the solution was carbonatedand distilled, and the process repeated. From 20 litres of originalsolution, Lewis obtained 1.5 C.C.of specific gravity 1.073 (correspond-ing to 66% D,O), and ultimately from several series of electrolyseshe prepared 1.3 C.C. of practically pure D,O.The electrolytic separation is not perfect, and the gas liberateda t the cathode contains an appreciable proportion of diplogen.According to Lewis and Macdonald, when the amount of H,O hasbeen reduced by electrolysis by a factor n, the amount of D20 isreduced by a factor na, where a is approximately 0.2 for electrolysisa t 35'. D. H. Rank l3 has pointed out that, owing to evaporation,a rise in temperature will increase the apparent value of a with acorresponding fall in the efficiency of separation. Two importantcorollaries to the imperfection of the electrolytic separation are(i) that technical electrolytic baths continually replenished by watertend to reach a relatively small equilibrium concentration of D20,and (ii) that it is expedient in the later stages of the electrolyticconcentration to recombine the mixture of hydrogen and diplogenevolved.Not much has been done so far to investigate the conditionsaffecting the efficiency of the electrolytic separation, and accord-ingly the mechanism is still in doubt.Since it has been estimatedthat the difference in the reversible electrode potentials of the twoions must be small,14 the principal alternatives seem to be: (1)The greater speed with which the lighter ion picks up an electronfrom the cathode to form a neutral atom; (2) the greater speed ofunion of hydrogen atoms on the cathode to form molecular hydrogen,8 W.Bleakney and A. J. Gould, Physical Rev., 1933, 44, 265; A., 994.9 J. Horiuti and M. Polanyi, Nature, 1933, 132, 819.10 A. and L. Farkas, Nature, 1933,132,892.11 A. and L. Farkas ; communicated to the Royal Society by Professor E .12 G. N. Lewis and R. T. Macdonald, J . Chem. Physios, 193$1, 341.18 IW., p. 760.14 H. C. Urey and D. Rittenberg, {bid., p. 137.K. Rideal on Dec. 16th, 193330 GENERAL AND PHYSICAL CHEMISTRY.compared with the rate of formation of I>, from diplogen atoms;(3) the greater mobility of the lighter ion in solution.J. Horiuti and M. Polanyi l6 conclude that hydrogen over-voltage is due to the inertia associated with the first of the aboveprocesses, since the rate of establishment of equilibrium betweengaseous diplogen and its ions in aqueous solution at a platinumsurface is very sensitive to the composition of the solution.R. P.Bell and J. H. Wolfenden,16 who have investigated the influenceof various factors on the efficiency of the electrolytic separation?also conclude that the separation in alkaline solution dependsprimarily on the first of the three factors enumerated above. B.Topley and H. Eyring 16a have independently arrived at somewhatsimilar conclusions.The proportion of the heavy isotope in ordinary hydrogen hasbeen variously estimated. Some of the original estimates wereundoubtedly too low because they depended on spectroscopicmeasurements of electrolytic hydrogen, which inevitably containsless than the ordinary amount of diplogen.G. N. Lewis andR. T. Macdonald,12 from a pyknometric comparison of ordinarywater with pure water, have suggested the proportion 1 : 6,500,but in a later communication Lewis makes the (as yet un-amplified)remark that this estimate is " altogether too high." W. Bleakneyand A. J. Gould have examined in the mass-spectrograph a,specimen of hydrogen obtained by the complete decomposition ofrain water with hot iron; they estimate the proportion as 1 : 5,000.The most sensitive test for the heavy isotope is the bombardmentmethod of M. L. E. Oliphant, B. B. Kinsey, and Lord Rutherf0rd.l'When lithium is bombarded by accelerated diplons, cw-particlesare produced whose range is 50% greater than those observed whenprotons are used as projectiles.This method is likely not onlyto lead to an independent estimate of the I) : H ratio in ordinaryhydrogen but also to provide a rapid and accurate method of estimat-ing the relative value of various natural and artificial sources ofwater rich in the heavy isotope. Two other methods of evaluatingthe D : H ratio in minute quantities of material have been described ;E. S. Gilfillan and M. Polanyi l8 have devised a flotation methodapplicable to specimens of water less than 0.01 C.C. in volume;A. and L. Farkas l9 have applied the thermal conductivity method,l6 Nature, 1933, 132, 931.l6 Ibid., 1934, 133, 25.lBO J . A w r . Chem. SOC., 1933, 55, 5058.l7 Proc. Roy.SOC., 1933, [A], 141, 722; A., 1100; cf. G. N. Lewis, M. S .18 2. physikcsl. Uhm., 1933,166, [A], 255.lD Nature, 1933,132, 892.Livingaton, and E. 0. Lawrence, Physical Rev., 1933, 44, 55WOLFENDZK : ELECTROCHEMISTRY. 3 1devised for the analysis of mixtures of ortho- and para-hydrogen,to determine the isotopic ratio in diplogen-hydrogen mixturesoccupying as little as 0.002 C.C. at N.T.P.The physical properties of pure D20 have been examined by G. N.Lewis and R. T. Macdonald20 and by P. W. Selwood and A. A.Frost.21 The principal results sothe following table :Melting point ..............................Boiling point .................................Temperature of maximum density.. ....Viscosity a t 20" (millipoisea) ............Surface tension at 20" (dynea/cm.) ...Specific gravity at 25" .....................Refractive index ( n r ) ..................a, ,, (n?') ..................Molar susceptibility ( x lo(;) .........far obtained are summarised inD,O .3.8"101.42"1.105611-6"14.2; 21 12.60 2267.81.32811.3265- 13OrdinaryH,O .0"100"1.04"10-8772-751.332931.33094- 13In addition, the ratio of the dielectric constant of D20 at infinitewave-length to that of H,O is estimated by Lewis23 and his co-workers as 0.990 a t 25"; the vapour-pressure ratio D20/H20 risessteadily from 0.87 at 20" to 0.949 a t 100°.20 The solubilities ofsodium chloride and of barium chloride in pure D20 are of the orderof 15% less than in ordinary water.24Lewis and T.C. Doody 25 have measured the mobility of the ionsof the heavy isotope, of chlorine, and of potassium in pure D,O.Their results at 18" are as follows :H . D'. K'. Cl'.In D,O .................. - 213-7 64.5 55.3In H,O .................. 315.2 - 64.2 65.2In spite of the present discrepancy in the values for the viscosityof D20, it seems probable that the mobility of I)' in D20, whencorrected for the increased viscosity of the solvent, is not more than30% less than that of H' in H20, in contrast to the predictions ofBernal and Fowler.* The measurements of Lewis and Doodydo not, however, definitely exclude the possibility that mobilitydifferences play a part in the electrolytic separation, at least inacid solution, since this may depend on the experimentally in-accessible mobility of D' in H20.The chemical behaviour of the heavy isotope has been discussedzo J . Amer.Chem. Soc., 1933, 55, 3057; A., 894.al Ibid., p. 4335; A., 1233.z2 G. N. Lewis and R. T. Macdonald, ibid., p. 4730.23 G. N. Lewis, A. R. Olson, and W. Maroney, ibid., p. 4371.34 H. S. Taylor, E. R. Caley, and H. Eyring, ibid., p. 4334; A., 1240.2 5 Ibid., p. 3609.* Seep. 3432 GENERAL AND PHYSICAL CHEMISTRY.theoretically in several papers. Chemical differences betweenhydrogen and diplogen are to be expected on theoretical groundsfor several reasons, but the dominating factor is likely to be thesmaller half-quantum of zero-point energy possessed by diplogenin virtue of its greater mass. Except in the special case where theenergy levels in the activated complex are displaced by a com-pensating amount, the lower zero-point energy will lead to a higherheat of activation for reactions involving diplogen and thereforeto a lower reaction velocity.E. Cremer and M. Polanyi26 audH. Eyring and A. Sherman27 both anticipate that the velocityof reaction (and velocity of desorption) will be substantially smallerfor the heavy isotope. H. C . Urey and D. Rittenberg14 havecalculated the free energy of the molecules D,, DH, DC1, and DI,and thence the equilibrium constants for reactions involving thesemolecules and their hydrogen analogues. They estimate thatthe equilibrium constant [HDI2/[H,] [D2]'will fall away substantiallyfrom the value of 4 (to be expected on a random pairing-off ofthe two kinds of atom) as low temperatures are reached.Theyalso conclude that the ratio of the equilibrium constants (KJK,)for the two parallel reactions :(1) H2+12=2HI . . . . . .D,+I2=2DI (2) . . . . . .will not be unity but 1.234 at 575" K. and 1.222 at 700" K. Thecorresponding ratio for the two parallel reactions with chlorineis estimated at 0.807 at 575" K. and 0.874 a t 700" K. They furtherconclude, more tentatively, that the free-energy difference betweenthe two isotopes is not likely to lead to a difference in reversibleelectrode potential adequate to account for the electrolyticseparation.Experimental observations of the chemical behaviour of diplogenare so far scanty but of the greatest interest. K.3'. Bonhoefferand G. W. Brown28 have shown that, if ammonium chloride isdissolved in D-rich water, the D : H ratio in the salt after recoveryapproximates to that in the water ; similar observations were madeby Lewis in the case of ammonia.29 Bonhoeffer and Brown foundthat if the same experiment is performed with sugar in D-rich water,the equipartition of D between water and sugar extends to onlyhalf the hydrogen atoms in the sugar SO that exchange can pre-sumably only take plaee with the hydroxylic hydrogen.2 6 Z. physilcal. Ohem., 1932, [B], 19,443; A., 235.27 J . Chem. Physics, 1933, 1, 435.28 2. phy8ikal. Ohm., 1933, [B], 23, 171; A., 1242.2g J . Arner. C M . SOC., 1933, 55, 3602; A., 1020WOLFENDEN : ELECTROCHEMISTRY. 33A.and L. Parkas l9 have studied the equilibrium H, + D, = 2HDas set up in contact with a hot nickel wire ; they find the equilibriumconstant [HD]2/[H2][D2] to have a value of about 3 above roomtemperature, in accordance with the predictions of Urey and Ritten-berg,l* and to be independent of temperature in this temperaturerange, pointing to a negligible heat of reaction. They have alsoinvestigated the same equilibrium as established in the homogeneousgas phase above 600".A more unexpected observation is that of M. L. O l i ~ h a n t , ~ ~ whofound that hydrogen gas rich in diplogen lost more than 95% ofits diplogen content after storage over ordinary water for six weeks.Since the volume of gas was unchanged, this points to exchangebetween gaseous diplogen and the combined hydrogen of the water.J.Horiuti and M. Polanyi have since shown that the same changeproceeds more rapidly in the presence of catalytically active plati-num; they suggest that the platinum assists the formation anddischarge of diplogen ions as it does at a hydrogen electrode. I nfurther experiments l5 they have shown that the rate of exchangeis very sensitive to the composition of the solution, being highestin pure water and almost negligible in a mixture of alcohol (98%),water (2%) and N/4-potassium hydroxide ; this leads Horiutiand Polanyi to conclude that the slow process determining therate of exchange is the discharge of the ion and not the union ofatoms t o form molecules, a conclusion of significance for the generaltheory of overvoltage.Differences between the rate of reaction of diplogen and ofhydrogen in corresponding reactions have now been definitelyestablished, and instances are likely to multiply rapidly.W.Bleakney and A. J. Gould,8 when decomposing rain water bypassage over hot iron, found that the first fraction of hydrogen,examined in the mass-spectrograph, showed a D : H ratio of 1 : 6,000whereas the last fraction showed a ratio of 1 : 4,500. This has beenconfirmed by Horiuti and P01anyi.~ A. and L. Parkas l1 have foundthat, if acidified water containing 50% of diplogen is treated withzinc, the residual solution contains a greatly increased proportionof the heavy isotope. E. Pacsu 31 has shown that the mutarotationof cc-glucose dissolved in water containing about 50% of diplogenis only about half as rapid as the corresponding reaction takingplace in ordinary water.On the other hand, E. Cremer and M.Polanyi26 were unable to find any increased concentration ofdiplogen in the residual hydrogen remaining from the catalytichydrogenation of styrol on palladium. They concluded from their30 Nature, 1933, 132, 765; A,, 1223.31 J . Amr. Chein. Soc., 1933, 55, 6056.REP.-VOL. XXX. 34 GENERAL AND PHYSICAL CHEMISTRY.experiments that hydrogen does not react more than 2.5 times asfast as diplogen with styrol under the conditions of their experiment.If the reaction proceeded by the ‘‘ tunnel ” mechanism of M. Bornand J. Franck’s32 theory of contact catalysis, a velocity ratio ofthe order of lo6 might have been expected.This observation ofCremer and Polanyi is primarily of interest in so far as it foreshadowsthe application of diplogen to the solution of problems of generalchemical interest.Mention must be made of the biological behaviour of heavywater. Lewis33 has shown that tobacco seeds will not germinatein nearly pure D,O and only grow half as fast in 50% D,O as inpure water. H. S. Taylor and his collaborators 34 have shown thatpure D,O kills a variety of small organisms with a speed roughlyproportional to the degree of organisation of the animal : forinstance, tadpoles are killed in one hour, protozoa in 48 hours;30% D,O seemed to be without effect on the organisms studied.Our knowledge of the chemistry of diplogen is still extremelyslender, but the possibilities of future development are innumerable.Apart from the interest attaching to the individual behaviour ofa new atom of such general chemical importance, it is highly probablethat diplogen and its compounds will prove a versatile weaponin attacking many problems of general interest. The behaviourof D, in gas reactions and the catalytic activity of the diplon insolution are only two out of many possible directions in whichexperiments with the new atom may prove illuminating in physicalchemistry.In organic chemistry the new possibilities of isomerismare almost inexhaustible; the new isotope is likely to prove evenmore valuable in affording a new approach to many problems ofstructure and reaction mechanism (such as those involving hydrogenmigration) by a method somewhat akin to that of radioactiveindicators.The possibilities of the diplon in effecting artificial disintegrationare discussed in the report on Radioactivity and Sub-atomicPhenomena.A Theory of the Structzcre of Water and Ionic Solutions.A recent paper by J.D. Bernal and R. H. Fowler35 describesan unusual approach to a variety of problems of considerable33 Nach. Ges. Wiss. Gdttingen, 1930, 77; A., 1931, 318.33 J. Amer. Chem. SOC., 1933, 55, 3503; A., 1093.34 R. S. Taylor, W. W. Swingle, H. Eyring, and A. H. Frost, J . Ohm.35 J . Chern. Physics, 1933,1,615. A shortened form of the paper is given inPhysics, 1.933, 1, 751.Trans. Faraday Xoc., 1933,29, 1049; A., 1106WOLFENDEN : .ELECTROCHEMLSTRY.35physicochemical interest. Not only does it suggest a " pseudo-crystalline " structure for liquid water, but it applies the postulatedstructure to account qualitatively and sometimes quantitativelyfor phenomena as apparently unrelated as the heat of sublimationof ice, the variation of dielectric constant of water with temperature,the degree of hydration of ions and their heats of hydration, theviscosity of electrolytic solutions, and the abnormal mobilities ofhydrogen and hydroxyl ions in water.From the density of water, which is much less khan would beexpected for a close-packed liquid with the known molecular di-mensions of water, and from the X-ray diffraction curves of liquidwater, Bernal and Fowler infer that the molecules are for the mostpart arranged in a four-co-ordinated structure similar to the dis-tribution of SiO, molecules in quartz.Such an arrangement(" Water-I1 ") predominates in the liquid between 0" and 100";supercooled water contains a certain proportion of a tridymite-like arrangement of molecules (" Water-I ") corresponding to thcstructure of ice itself; above 100", and to an increasing extent asthe critical point is approached, water passes over into the close-packed form (" Water-I11 ") characteristic of an ideal liquid. Thetetrahedral co-ordination of both Water-I and Water-I1 harmoniseswith the wave-functions of the water molecule, which lead to ELpicture of electrical-density distribution resembling a tetrahedronwith two corners of positive charge and two of negative charge.The " pseudocrystalline " structure of both forms differs fromthat of a true crystal in the irregularity of the co-ordination andthe continual fluctuations of the arrangement.Among the manifold applications of this picture of water, thoserelating to ionic solutions are of the greatest interest to chemists.The hydration of an ion will occur when the potential energy of awater molecule co-ordinated with the ion (in the crystallographicsense) is less than that of the molecule in free water, where it isco-ordinated with other water molecules.Both potential energiesare calculated in terms of simple Coulomb forces from the modelof the water molccule inferred from spectroscopic and dipole dataand from the ritdius and charge of the ion.On this basis it isshown that all univalent ions whose radius is less than 1.6 8. andall polyvalent monatomic ions must be hydrated. I f it is assumedthat those ions are hydrated whose apparent volume in solutionis much less than their volume in the solid state, the above con-clusion is in harmony with the facts.I n calculating heats of hydration, it is assumed that each ion issurrounded by a sphere of co-ordinated water molecules outsidewhich the water behaves as a continuous dielectric. The hea36 GENERAL AND PHYSICAL CHEMISTRY.developed in the formation of the co-ordination sphere is calculatedin terms of the mutual potential energy of ion and dipole as indicatedabove; that developed in the dielectric is calculated by the well-known Born formula for the work of electrostatic solvation, theradius of the co-ordination sphere being used instead of that of theion.A number of complicating factors, such as the mutual in-fluence of the co-ordinated water molecules on one another, makesit impossible to calculate heats of solvation in this way on purelya primi grounds, but Bernal and Fowler have shown that theexperimental data lead to plausible values for the size of the co-ordination sphere. Their calculation thus accounts for beats ofhydration in purely electrostatic terms without ignoring, like theBorn-Bjerrum equation, the molecular structure of the solvent.The application of the Bernal-Fowler model to the viscosityof electrolytes is still in a tentative and qualitative stage.It issupposed that, in addition to the simple electrostatic effect due totheir charges3s and the Brownian effect, the addition of ions towater may change its viscosity by “loosening or tightening thestructure” of the liquid. Ions such as potassium or rubidium,which produce “ negative viscosity ” and a strong lowering of thetemperature of maximum density, are regarded as being particularlyeffective in breaking down the structure of water. Since such anion produces the same effect on the structure of water as a rise intemperature, it is said t o raise the “ structural temperature,” whichis defined as that temperature at which pure water would haveeffectively the same inner structure.The final application of the structure of water made by Bernaland Fowler is to the abnormal mobilities of the hydrogen and thehydroxyl ion.The Grotthuss mechanism first advanced by E.Hiickel37 is developed along quantum-mechanical lines. Withplausible assumptions concerning the height of the potential barriersinvolved, it is possible to account for the abnormal mobility ofthe two ions by means of a succession of jumps of protons from watermolecule to water molecule or from one hydroxyl ion to another.The authors anticipate that the increased mass of the diplon (D‘)will diminish the probability of the quantum-mechanical jumpso greatly as to wipe out almost completely the extra mobility of D’in D20. The mobility of the diplon in D20 should therefore beabout one-fifth of that of the hydrion in ordinary water.Thisexpectation has not been realised.”36 See H. Falkenhagen and M. Dole, 2. phys-ikal. Chern., 1929, [B], 6, 159;A., 1930, 156.37 2. Elektrochem., 1925, 34, 546; A . , 1920, 143.* Seep. 31WOLFENDEN : ELECTROCHEMISTRY. 37The Mechanism of the Hydrogen and Oxygen ElectrodeProcesses.It is only within the last few years that progress has been madetowards the elucidation of the detailed mechanism of electrodeprocesses. On the one hand, the nature of the electrical doublelayer and the phenomena of electro-capillarity are being extensivelyinvestigated,s8 and on the other hand, by kinetic studies of the estab-lishment and decay of electrode potentials, notably by F. P.Bowden 39and by J. A. V. Butler,4O some insight is being gained into the suc-cessive stages involved in the discharge of ions during electrolysis.*In the latter work interest centres largely round the problemof hydrogen and oxygen over-voltages and, although the phenomenaare often very complicated and their interpretation controversial,one conclusion a t least seems to be emerging. It is that, in mostcircumstances, the rate-controlling factor which determines theover-voltage is the neutralisation of the ion and not the furtherstages whereby the discharged ion is transformed into the gaseousproduct of electrolysis. In the case of hydrogen over-voltage thecrucial step is H + E ---+ H and not H + H -+ H,. Over-voltage is thus due, not to the accumulation of atomic hydrogen,but to the slowness with which the hydrogen ions pick up electronsfrom the cathode. This postulate is common to the over-voltagetheories both of T.Erdey-Griiz and M. Volmer41 and of R. W.Gurney; 42 Gurney is, however, more specific as to the nature ofthe energy barrier between ion and metal which determines therate of neutralisation of the ions.Recent work suggests that at very small over-voltages at aplatinum cathode complicating factors are present. G. Armstrongand J. A. V. Butler 43 have measured the rate of decay of hydrogenover-voltages a t mercury and platinum cathodes. Although theresults for mercury were consistent with the Gurney theory, yet38 See, e.g., 0. Stern, 2. Elektrochem., 1924, 30, 508; K.Beimewitz and J.Schulz, 2. phypikal. Chern., 1926,124, [A], 116; A., 1926, 1212; A. Frumkin,Ergeb. exakt. Naturwiss., 1928, 7, 240; F. 0. Koenig, 2. physikal. Chem.,1931,154, [A], 421, 454; A., 1931,801; J. Philpot, Phil. Mag., 1932,13,776;A,., 1932, 470; S. R. Craxford, ibid., 1933, 16, 849.39 Proc. Roy. SOC., 1929, [A], 125, 446; A., 1929, 1391 ; ibid., 126, 107 ; A.,1930, 169.40 Trans. Faraday SOC., 1924, 19, 734; A., 1924, ii, 598; ibid., 1932, 28,379; A., 1932, 700; Proc. Roy. Soc., 1932, 137, [A], 604; A., 1932, 1092.41 2. phy8ikal. Chem., 1930,150, [A], 203; A., 1930, 1376.42 Proc. Roy. Soc., 1931, [A], 134, 137 ; A., 1931,25 ; cf. Ann. Reports, 1932,45 Trans. Faraday SOC., 1933, 29, 1261.* A paper by A. Frumkin (2. physikal. Chern., 1933,164, [A], 121 ; A., 468)29, 34.makes some progress towards correlating these two fields of investigation38 GENERAL AND PHYSICAL CHEMISTRY.the rate of decay a t platinum, where the over-voltage is muchsmaller, was less than if it were governed solely by the transfer ofelectrons from the metal to the solution, and the authors concludethat in these circumstances there was an appreciable accumulationof electromotively active hydrogen at the electrode.Somewhatsimilarly, L. P. Hammett 44 concludes from his current-over-voltage curves for hydrogen at platinum electrodes that at highover-voltages the rate of hydrogen evolution is determined by therate of neutralisation of the ions, but that, as the over-voltagediminishes, the rate of combination of hydrogen atoms becomesthe dominant factor.It is well known that the “ oxygen electrode ” does not behavein a thermodynamically reversible manner, the E.M.F.of thehydrogen-oxygen cell being far from reproducible and fallingshort of the theoretical value by about 0.2 volt. This is generallyattributed to the presence of oxides of platinum, which preventsthe saturation of the electrode with oxygen. Support for this viewand a further insight into the mechanism of the oxygen electrodeare afforded by a recent paper by T. P. according to whoma, platinum electrode in oxygen-saturated electrolyte is coveredby an oxide film containing cracks and pores. The abnormallylow potential of the electrode is due to a continuous flow of currentbetween the film and the relatively anodic metal, causing an irre-versible removal of electromotively active material from the film.If, however, the current necessary for this irreversible oxide form-ation is supplied by a cathode other than the film itself, Le., if theelectrode is anodically polarised by an external source of current,the reversible potential may be obtained.Hoar measured thecurrent necessary to keep a bright platinum electrode in oxygen-saturated electrolyte at its theoretical reversible potential byanodic polarisation. This current falls off a t first rapidly withtime and then slowly a t a rate quantitatively consistent with thehypothesis that a progressive closing-up of the pores in an oxidefilm is taking place.Cathodic and anodic polarisation curves were then measured atconsiderably higher current densities, where the irreversible processwould be relatively insignificant. In both cases the logarithm ofthe current density varied linearly with the voltage except at verylow current densities.By extrapolating both anodic and cathodiccurves back to the potential at which the rates of charge and dis-charge of ions would be equal, the reversible potential of the oxygenelectrode was found to be 1.20, in fair agreement with the theoreticalvalue. J. H. W.45 Proc. Roy. Soc., 1933, [A], 142,628. 44 l’runs. E’wi*udccySoc., 1933,243, 770HINSHELWOOD : CHEMICAL KINETICS. 393, CHEMICAL KINETICS.In this section some special problems have been selected fordiscussion.Energy Transfer Relations.It is in connexion with problems of energy transfer that reactionkinetics comes into closest relation with theoretical physics on theone hand and with experimental spectroscopy on the other.Fromthe chemical point of view the specificity of such changes is wellrecognised, and there are now innumerable examples of theirimportance in molecular activation and deactivation.An interesting physical method of approaching the problem isby the study of the variation of the velocity of sound withfrequency.l At very high frequencies the adiabatic changesaccompanying the passage of sound through a gas are too rapid toallow the establishment of equilibrium between the translationaland vibrational energy of the molecules : thus the effective valueof y will change at frequencies where the failure of complete energyequilibrium sets in.have studied thedispersion of sound in carbon dioxide and in chlorine, in the absenceand in the presence of foreign gases. The latter help to maintainthe equilibrium between translational and vibrational energy, butvary very much in effectiveness. Hydrogen, hydrogen chloride,and methane were found to be effective, while argon was ineffective.The contrast between hydrogen and argon recalls the relativeefficiency of these two gases in maintaining the rates of unimolecularreactions at low pressures.In discussing the theoretical aspect of the matter, Franck andEucken express the view that the important factor in energytransfers is the extent to which the linking of the molecule receivingthe vibrational energy is modified by the proximity of the secondcolliding molecule.In other words, there must be mutual disturb-ance of the potential curves. Molecules can be set in vibrationby the impact of electrons, because the electrons penetrate themolecules and cause such a disturbance of the bindings that thenuclei find themselves with a mutual potential energy and beginto vibrate. The inefficacy of helium in communicating energyin unimolecular reactions is attributed to the fact that it will berepelled by saturated molecules; the diatomic hydrogen is notsimilarly repelled because its own potential-energy curve can suffermodification as it approaches. Particularly great disturbances byJ.Franck and A. Eucken, 2. physikal. Chem., 1933, [B], 20, 460, wherereferences to the work of Pierce; Kneser; Herzfeld and Rice; Richardsand Reid ; Euclcen, Mucke, and Becker ; and Henry are given.A . Eucken and R. Becker2. physikal. Chem., 1933, [B], 20, 467; A,, 55440 GENERA& AND PHYSICAL CHEMISTRY.exchange forces are to be expected when radicals or free reactiveatoms approach a molecule, and dipoles and ions should favourtransfers because of the disturbance which they create in theelectronic system.In this connexion, the efficiency of free atoms of halogens incatalysing the decomposition of nitrous oxide may be re~alled.~The decomposition of various organic substances possessing a mobilehydrogen and a breakable C-C linkage is strongly catalysed by thehighly polarisable molecule of iodine, and it is signscant that allthe substances whose reactions are promoted in this way possessdipole moments in or near the bond broken in the chemical ~ h a n g e .~The influence of ion pairs in catalysing the isomeric change ofoximes is also noteworthy.4aAccording to the theoretical considerations of 0. K. Rice,5probability of energy transfer decreases with increasing mass of thecolliding particle. In the decomposition of nitrous oxide, therelative probability of activation by collision of nitrous oxidemolecules with molecules of the rare gases appears to decrease fromhelium to xenon.gIf internal rearrangements of energy are difficult, it is possible toconceive the existence of different kinds of activated state.Whenthe molecule has received energy in a collision, and this energy islocated in the molecule in a, particular way, there may be oneprobability of chemical decomposition, whereas when the energyislocated in another way there may be another probability. If themolecule is of moderate complexity, there may be, quite apart fromcontinuous variations of transformation probability with totalenergy, a limited number of activated states which can be regardedas qualitatively different. If this is so, there will appear to takeplace a number of simultaneous reactions, differing in physicalmechanism, but chemically identical. It can be shown that thesereactions will be differently influenced by pressure. Investigation,over a wide range, of the influence of pressure on the decompositionof acetaldehyde shows that the physical mechanism of the reactionmust be regarded as a composite one in the above sense.' The* F.F. Musgrave and C. N. Hinshelwood, Proc. Roy. Soc., 1932, [A], 137,25; A., 1932,917; M. Volmer and M. Bogdan, 2. physikal. Chem,, 1933, [B],21, 257; A., 680.S . Bairstow and C. N. Hinshelwood, Proc. Roy. SOC., 1933, [A], 142, 77;A., 1251.Oa T. W. J. Taylor and D. C. V. Roberts, J . , 1933, 1439.J . Amer. Chem. SOC., 1932, 54, 4558; A., 129.M. Volmer and M. Bogdan, Zoc. cit., ref. (3). ' C. J. M. Fletcher and C. N. Hinshelwood, Proc. Roy. SOC., 1933, [A],141, 41 ; A., 910HINSHELWOOD : CHEMICAL KINETICS. 41idea of a balance between activating and deactivating collisions,and the recognition of the ability of almost any type of foreignmolecule to contribute to these processes in a highly specific manner,make it possible t o correlate the most diverse relations betweenrate of reaction and concentration, and simplify the classificationof chemical changes.It is now usually not profitable to attemptto specify the influence of various concentrations in terms of an'' order of reaction." Instead, it is better to write down an equationexpressing the rate of formation and destruction of active mole-culesYs or of an activated complex of a reactant molecule and acatalyst molecule, and solve it for the concentration of activatedmolecules. The rate of reaction will in general be proportional t othe latter.Every substance present contributes potentially t oactivation and to deactivation, and according to the value of thespecific constants expressing this, the most varied relations appearas special cases.9 I n this way the well-known " quasi-unimole-cular " reaction which changes from the first order t o the secondwith diminishing initial concentration is accounted for, and theinfluence of various foreign substances is expressed. Other specialcases are catalytic reactions the rate of which is proportional tothe concentration of the catalyst, catalytic reactions with a ratenearly independent of the catalyst concentration, and either of thefirst order with respect t o the reacting substance itself, or evenapparently impeded by increase in the reactant concentration.Examples of all these types are now known.Catalytic Actions of Magnetic Substances.The conversion of para- into ortho-hydrogen is catalysed bymagnetic molecules.I n the gas phase, oxygen, nitric oxide, andnitrogen peroxide are active.1° The reaction is bimolecular, beingof the first order with respect to the para-hydrogen and having a rateproportional to the catalyst concentration. At 293" K. the velocityconstants for the three catalysts are 9.16, 34.9, and 12.5 litre/mol./min. respectively. The effectiveness of the collisions in causingtransformation varies little with temperature. The interconversionof para- and ortho-hydrogen is also catalysed by various solvents,llthe velocity constants being about 5 x times those of the gasreaction.The action of the solvent is attributed to the influenceof the nuclear magnetic moment of the hydrogen atoms which itcontains. The reaction is also catalysed by various paramagneticAnn. Reports, 1931, 28, 41.9 Proc. Roy. Soc., 1933, [A], 142, 77; A., 1251.10 L. Farkas and H. Sachsse, 2. physikal. Chem., 1933, [B], 23, 1 ; A., 1251.l1 Idem, ibid., p. 19.B 43 GENERAL AND PHYSICAL CHEMISTRY.ions, the rate being proportional to the square of the magneticmoment of the ion. Catalysis by dissolved oxygen occurs at a ratecorresponding to the rate in the gas phase.Interconversion of the two forms of hydrogen in contact withsolid substances a t low temperatures also depends upon the para-magnetism of the solid.12The theory of the magnetic conversion has been given by E.Wigner.13In$uence of Internal Electronic -Displacementson Reaction Velocity.If there is an appreciable dipole moment located in a particulal.bond of a molecule, then in accordance with the sign of the momentthere will be a certain displacement of electrons throughout themolecule either towards or away from the bond containing thedipole.In many chemical reactions one atom or group can beregarded as the principal seat of the change ( e . g . , in the hydrolysisof an alkyl halide the behaviour of the halogen atom is clearly thefactor of primary importance), and tlhe drift of electrons towards oraway from it may modify the reactivity profoundly. It is far fromeasy always to predict the sense in which the reactivity will bemodified.If, for example, chlorine is being replaced by hydroxyl,the tendency of the chlorine to come away as a negative ion isincreased by a drift of electrons towards the part of the moleculeto which it is attached. On the other hand, the approach of anattacking hydroxyl ion is hindered. Some specific idea of thereaction mechanism may be necessary before the effect can bepredicted. Nevertheless, organic reactions can be classified intotwo types : those which are facilitated by a displacement of electronsin the direction of the reacting atom or group (type A), and thosewhich are facilitated by a corresponding displacement away fromthe reactive centre (type B).14 The conception of an order ofrelative polarity of substituents has long been familiar in organicchemistry: it has lately become more precise on account of thepossibility of measuring dipole moments.Thus, quantitativerelations between " polarity '' influences on reaction velocity andthe magnitude of dipoles can be looked for.In the reaction of a ketone CH,*CO*C,H,*X with a halogen, thevelocity is determined by the enolisation of the ketone, and is12 K. F. Bonhoeffer, A. Farkas, and K. W. Rammel, 2. physikal. Chem.,1933, [B], 21, 225; A., 680; H. S. Taylor and H. Diamond, J . Amer. Chem.SOC., 1933, 55, 2613; A., 788.2. physikal. Chem., 1933, [B], 23, 28 ; A,, 1236.l4 C. K. Ingold and E. Rothstein, J., 1928, 1217HINSHELWOOD : CHEMICAL RTNETICS. 43proportional to the hydrion concentration. It has therefore beenconcluded that this velocity represents the speed of associationof protons with molecules of ketone.If the substituent X causesa drift of electrons away from itself, then the ease of addition ofprotons elsewhere in the molecule should be increased, and viceversa. I n accordance wibh expectation, the rate of reaction decreasesin the following order for various substituents (all in the p-position) :Me, H, I, Br, C1, NO,. By making the assumption, not yet justifiedby direct experiment, that the reaction rate is connected with theenergy of activation by the simplest possible relation : rate =collision number x e-E’RT, changes in velocity from substance tosubstance can be recalculated as changes in energy of activation.It then appears l5 that the difference in energy of activation caused bya substituent X is nearly linearly proportional to the dipole momentof the compound C,H,X.I n many other reactions, especially side-chain reactions of aromaticcompounds, the same order of substituents is found, whether wearrange them according to decreasing velocity of reactions of typeA or increasing velocity of reactions of type B.This order is CH,,H, Halogens, NO,. Recalculating differences in speed in termsof a hypothetical difference in energy of activation, we find a similarquantitative relation with p, the dipole moment of C,H,X, vix.,E = E, & c(p -E, is the heat of activation when there is no substituent in thebenzene nucleus, E the corresponding value when the substituentX is present.The alternative signs apply according as the reactionis of the type A or B. When X is in the meta-position the relationapplies rather exactly for all except halogen substituents ; for para-substituents the deviations from a quantitative relation appear tobe greater. The halogens deviate from the quantitative relationin such a way that they could be made to fit it if they were creditedwith larger dipole moments than t,he measured ones. I n thisconnexion it is pointed out l5 that L. E. Sutton’s work 16 shows thearomatic halogen compounds to have dipole moments smaller thanwould be expected from analogies with aliphatic halogen com-pounds: some factor operates to reduce the measured moment.If this factor did not operate in influencing the reaction velocity,the anomalies would be partly accounted for.The halogens are anomalous in another respect, viz,, that,although as a group they come among other substituents in theorder to be expected from the dipole moments, yet they are oftenl5 W.S. Nathan and H. B. Watson, J . , 1933, 890, 1248; A., 1124.16 Proc. Roy. SOC., 1931, [ A ] , 133, 668; A . , 1931, 135444 UENERAL AND PHYSICAL CHEMISTRY.in an inverted or in an apparently random order among them-se1ves.l’ It is well known that to explain the facts of aromaticsubstitution two opposing effects of halogens in producing electrondisplacements have to be postulated; and the suggestion hasbeen made that these opposing effects may operate in differentways on different parts of complex reaction mechanisms, producinga resultant which according t o circumstances may be greatest forany one of the halogens.The inner mechanism of these effects,however, is still far from clear.Resonance and Stability of Molecules.In quantum mechanics the state of a system is described by afunction usually designated #. When the value of 4 is known, theenergy of the system can in principle be calculated. If two valuesof $t are possible, the state is described by the “linear combin-ation ” This expression is, according to the rules, giventhe following somewhat arbitrary meaning : a and b measure theprobability that a given example of the system is in the state where$ equals $tl or +2 respectively.If the system under considerationis a molecule, and the two states are two conceivable electronicstructures, then a and b measure the relative frequency of occurrenceof these forms. A really fundamental justification of the procedureof forming linear combinations of possible proper functions isperhaps too much to expect, but it is as well to be under no mis-apprehension as to its essentially arbitrary nature. If and $2belong to two structures of nearly the same energy, and if thereis the possibility that the two kinds of molecule “ interact ” in thesense of modifying each the potential energy of the other (a possi-bility which must presumably be conceded to most molecularsystems), then it is found by applying Schrodinger’s equation to theexpression a#, + b#2, that a and b vary with time, waxing andwaning in a way which indicates a rapid oscillation of the moleculesbetween the two possible structures.The process is called reson-ance. It also follows from the equations that the energy of eitherstructure involved in the resonating process is less than it wouldhave been had the interchange not been possible. Whatever degreeof satisfaction may be felt with the logical (as distinct from themathematical) basis of the calculation, the result can be clearlystated : when two structures of approximately equal energy arepossible for a molecule, it will fluctuate between them (or, if weprefer to express it so, exist in an indefinite intermediate condition),and moreover, it will be more stable than it would have been had1’ C.W. Shoppee, J., 1933, 1117; J. W. Baker, ibid., p. 1128; G. M.+ b$t2.Bennett, ibid., p. 1112; R. Robinson, ibid., p. 1114; A., 11e51HINSHELWOOD : CHEMICAL KINETICS. 45the alternative forms not been possible. At least this can be usedas an hypothesis in attempts to interpret the facts about thestability of different types of molecule.As an example the hydrogen halides may be considered. L.Paulingl* estimates that the energy of hydrogen fluoride with acovalent, electron-pair link is about the same as that of the ionicmolecule H'F'. Thus, both forms play their part : the moleculeis neither covalent nor ionic, but both (the covalent form being ofgreater importance in this actual case). With the other hydrogenhalides, the potential-energy curves of the covalent and the ionicforms do not cut, and the structure corresponds to the form oflower energy, vix., the covalent form.According to these principles, we should expect bond energiesto be additive only when possibilities of resonance are absent.Energies of individual bonds, such as C-H, C=O, and N-H, canbe calculated from the heat of formation of simple compounds suchas methane, ammonia, and formaldehyde.From the tabulatedvalues, the energy of formation of a compound of any assignedelectronic formula can be calculated. According to Pauling andJ. Sherman,19 if there is a considerable difference between themeasured energy and the value so computed, it is to be ascribed tothe " resonance energy." From an examination of a considerablenumber of examples, they reach the important conclusioii that when-ever the bond energies are not additive the deviations are in factin the direction corresponding to greater stability of the molecule.Resonance energies varying from 0.2 to 10 volt-electrons areinferred in this may.Pauling and Sherman make the comment thatthe bond energy of C=O is surprisingly large, and attribute itsmagnitude to resonance between C : : 0 : and C : 0 : . (It must beadmitted, however, that the possibility of introducing semipolarlinks increases the possibilities of resonance in almost any kind ofmolecule almost indefinitely, and makes the fundamental questionof a correlation between resonance forms and special stabilityrather more difficult to test.)In the present connexion the aromatic compounds are of specialinterest, on account of the large number of possible valency arrange-ments in their molecules.There may be resonance between aconsiderable number of different structures, e.g., two Kekul6 andthree possible Dewar structures?O and the resonance energy may0 . + 0 . - ..18 J. Amer. Chern. SOC., 1932, 54, 988.19 J . Chm. Physics, 1933, 1, 606.L. Pauling and G. W. Whelmd, {bid., p. 362.sy&erns, see L. Pauling and J. Sherman, ibid,, p. 679.For other conjugate46 GENERAL AND PHYSICAL CEEMISTRY.be very great. It is suggested that this has a very importantconnexion with aromatic stability.The theory has also something to say about the existence offree radicals of the triphenylmethyl type.Considering, first, hexa-phenylethane, we have two carbon atoms with the normal tetra-hedral disposition of valencies. These atoms are joined to oneanother and to three phenyl groups. If the bond joining themis broken, considerable energy is absorbed. In the two free radicalsformed, a rearrangement of valencies into a coplanar structure isassumed to take place. The number of resonance possibilities isgreatly increased, and the associated resonance energy is estimatedto be more than enough to compensate for the rupture of the C-Clink. Itis claimed that this method of treating the problem will explaininany of the more special details about differences in stability ofradicals differing slightly in structure.21Now that the kinetics of the dissociation and recombination ofsuch substances can be experimentally studied, as K.Ziegler 22 hasshown, and energies of activation measured, there will be a wideHence the stability of the free radical is accounted for.scope for the application of theory. c. N. H.Reactions of Free Atoms and Radicals.Recent developments in the production of free atoms and freeradicals are opening up new fields in chemistry, particularly in thepreparation of new inorganic compounds and in the measurementof heats of activation of atomic reactions, these measurementsbeing of great importance in theories of molecular structure and inthe interpretation of the complexities of chain reactions in gases.Atomic hydrogen,23 oxygen,2* and chlorine 25 are readily preparedby electric discharges through the gases at low pressures.Themethyl radical, and the less stable ethyl radical, are obtainableby heating the lead tetra-alkyl,26 by reaction of the alkyl halide21 E. Huckel, Trans. Paraday SOC., 1934,30,40 ; compare also C. K. Ingold,ibid., p. 52 ; Pauling and Wheland, Zoc. cit., ref (20).22 Ibid., p. 10.23 R. W. Wood, Phil. Mag., 1921,42, 729; 1922,44,538; K. F. Bonhoefler,Z. physikal. Chem., 1924, 113, 199; E. Boehm and K. F. Bonhoeffer, ibid.,1926,119, 385.24 E. Wreds, 2. Physik, 1929,54,53 ; P. Harteck and U. Iiopsch, Z.physikal.Chern., 1931, [B], 12, 327.25 G. M. Schwab and H. Friess, Naturwiss., 1933, 21, 222; A., 580; 2.Ekktrochern., 1933, 39, 586; A., 1021 ; W.H. Rodebush and W. C. Klingel-hoefer, J. Arner. Chern. SOC., 1933, 55, 130; A., 232.26 F. Paneth and W. Hofeditz, Ber., 1929, 62, 1335; I?. Paneth and W.Lmtsch, ibid., 1931, 64, 2702; R. N. Meinert, J . Amer. Chein, SOC., 1933, 55,979; A., 4'33BOWEN : CHEMlGAL KINETICS. 47with sodium v a p o ~ r , ~ ~ or with hydrogen atoms,28 and by heatingthe vapours of many organic substances.29 The methylene radicalappears t o be formed in the decomposition of methane on hotplatinum,30 and the production of the phenyl and the butyl radicalhas also been described.31 These radicals may be detected andestimated by their reactions with halogens or with thin films ofzinc, antimony, tellurium, etc. The cyanogen radical is producedin the reaction of sodium vapour with cyanogen halides,32 and thehydroxyl radical, which has a very short life, has also been pre-pared.33 The production of new unstable compounds by theinteraction of atomic hydrogen or oxygen with other molecules isdescribed by P.H a r t e c l ~ , ~ ~ and the preparation of mono-, di-, andtri-alkyls of arsenic, antimony, and bismuth by the direct inter-action of free methyl and ethyl with the elements by Paneth.35The most important use of work on free atoms and radicals appearst o be in the interpretation of the mechanism of thermal and photo-chemical reactions. In the photochemical reactions of hydrogenwith chlorine or bromine, for example, the reactionsX + H2-+ HX + H (X = C1 or Br)H + X,-+ HX + XH + HX--+ X + H,have been assumed to explain the experimental results (cf.p. 50).By the use of halogen or hydrogen atoms, each one of these reactionscan be separately examined in detail, its efficiency and temper-ature dependence found, and a direct confirmation of the correctnessof the photochemical assumptions obtained. I n a similar way,by examining the rates of reactions of hydrogen and oxygen atomswith oxygen and hydrogen molecules, and finding their dependence2 7 H. von Hartel, M. Meer, and M. Polanyi, 2. physikal. Chem., 1932, [B],19, 139; Ann. Beports, 1932, 29, 43; E. Horn, M. Polanyi, and D. W. G.Style, Trans. Paraday Soc., 1934, 30, 189; H. von Hartel, ibid., p. 187.28 H. M. Chadwell and T. Titani, J. Amer. Chem. SOC., 1933, 55, 1363; A , ,678.39 F.0. Rice, W. R. Johnson, and B. L. Evering, ibid., 1932, 54, 3529; A.,1932, 1108; F. 0. Rice, Trans. Paraday Xoc., 1934, 30, 152; J. A. Leer-makers, J . Amer. Chem. Xoc., 1933, 55, 3499.30 L. Belchetz, Trans. Paraday SOC., 1934, 30, 170.31 M. F. Dull and J. H. Simons, J . dmer. Chem. SOC., 1933, 55, 3898; F. E.32 J. Curry and M. Polanyi, Z . physikal. Chern., 1933, [B], 20, 276; A., 573.33 K. F. Bonhoeffer and H. Reichardt, ibid., 1928,139, 75; A,, 1928, 1188;K. F. Bonhoeffer and T. G. Pearson, ibid., 1931, [B], 14, 1; A., 1931, 1215;W. H. Rodebush and M. H. Wahl, J . Chem. Physics, 1933,1, 696.Vrey and H. J. Hepp, ibid., p. 3357.34 Trans. Paraday SOC., 1934, 30, 134.3 G F. Peneth, ibid., p. 17948 GENERAL AND PHYSICAL CHEMISTRY.on the temperature, information can be obtained as to the efficienciesof the reactions : 36H, +- 0-+ OH + HH, + 0 + (third body) --> H,O0, + H + (third body) -+ HO,H2O + 0 -+ 20HH, + 0, + H-+H,O + OHwhich may play a part in the thermal reaction between hydrogenand oxygen (cf.p. 51). Rice3' has attempted to explain themechanism of the decomposition of organic compounds (particu-larly hydrocarbons, ethers, and ketones) by heat in terms of dissoci-ation into radicals. On the hypotheses (1) that the primarymechanism is a unimolecular splitting of a link in the molecule,(2) that, as the energy of linking of C-H is greater than that ofC-C, the latter link splits in preference to the former, and (3) thatthis primary formation of radicals leads to a chain reaction, he isable to account semi-quantitatively for the actual products found.The view that a C-C link splits more easily than a C-H link, how-ever, does not accord with many chemical observations. R.Conrad3* has shown, from experiments on the decomposition ofhydrocarbons in a positive-ray discharge tube, that if a hydrocarbonmolecule loses one hydrogen atom the tendency is for anotherC-H link to break to form the link C-C.The mechanism of thephotodecomposition of aldehydes and ketones in ultra-violet light,though not yet clear, also shows that a C-H link is more easilybroken than a C-C link ; e.g., while the molecule CH,:O decomposeswith a quantum efficiency of 1 independently of wave-lengthY39that for the molecule CMeH:O falls off with increasing wave-lengthsfrom 1 to 0,40 and that for the molecule CMe,:O reaches the valueof only 0.17.41 Thermally, too, aldehydes are more easily decom-posed than ketones by homogeneous reactions, and at energies ofactivation much less than the heats of linkage of C-H or C-C.In all reactions of the unimolecular type, it appears that a mechanisminvolving a bimolecular interaction of two parts of the same mole-cule, rather than a simple splitting of a link, is more probable; e.g.,Hsrteck, Zoc. cit., ref.(34).152.36 K. H. Geib and P. Harteck, Trans. Paraday SOC., 1934, 30, 131: P.37 F. 0. Rice, Chm. Reviews, 1932, 10, 136.; Trans. Faraday SOC., 1934, 30,38 Trans. paraday SOL, 1934, 30, 215.39 R. G. W. Norrish and F.W. Kirkbride, J., 1932, 1518; A., 1932, 706.*O P. A. Leighton and F. E. Blacet, J. Amer. Chem. SOC., 1933, 55, 1766;41 G. H. Damon and F. Daniels, ibid., p. 2363; A . , 792.A., 682BOWEN : CHEMICAL KINETICS. 49in the decomposition of acetaldehyde, the hydrogen atom beginsto link up with the methvl group before detaching itself from thecarbonyl group ; and while the C-H and C-CH, linkages are weaken-ing, the carbonyl group is simultaneously undergoing reorganisationto give carbon monoxide C&O. In the slow oxidation of ethaneat high pressures, over 60% of the products contain two carbonatoms (chiefly ethyl alcohol and acetic acid), a fact which is difficultto reconcile with the hypothesis of a ready splitting of a C-C link.42The absorption spectra shown by aldehydes, as well as bychromylm and sulphuryl chlorides,44 may well be diffuse owing tooverlapping effects of rotational lines rather than because of aunimolecular decomposition of the excited molecule.If so, thedifficulties which arise in explaining the photochemical reactionsof these substances in terms of '' predissociation " disappear ; thereactions in fact falling into the bimolecular class.According to modern conceptions of valency, no bond is to beexpected to break in a molecule without simultaneous changes inthe energies of all the other bonds, so that comparisons between" heats of linkage " and activation energies or photochemicalthresholds are to be made on the basis of examination of theirdifferences (representing energies of reorganisation) rather than inexpectation of coincidences.The values of the bond energies in amolecule will also be greatly affected by the near approach of areacting atom. On these views, the old accepted mechanisms ofthe Walden inversion and of the Beckmann change, which describedthe effects as an atom approaching and attaching itself to theface of a tetrahedral carbon or other atom, followed by redistribu-tion of the valency links, present no difficulties of electronicinterpretation .45The Hydrogen-Chlorine Reaction.The alleged inhibiting effect of drying on the hydrogen-chlorinephotoreaction 46 now appears to be n~n-existent,~' and thereby42 D. M. Newitt and A. M. Bloch, Proc. Roy. SOC., 1933, [A], 140, 426; A.,678.' 43 R.de L. Kronig, A. Schaafsma, and P. K. Peerlkamp, 2. physikal. Chem.,1933, [B), 22, 323; A., 997; Kantzer, Cornpt. rend., 1933, 196, 1882; A., 791.44 H. W. Thompson, Nature, 1933, 132, 896.45 A. R. Olson, J . Chm. Phys;.S, 1933, 1, 418; H. Eyring and M. Polanyi,46 Cf. Ann. Reports, 1931, 28, 46.4' A. J. Allmand and H. C. Craggs, Nature, 1932, 130, 927; A., 34; G. K.Rollefson and J. C. Potts, J . Arner. Chem. Soc., 1933, 55, 860; A., 368; F.Bernreuther and M. Bodenstein, Sitzungsber preuss. Akad. Wiss. Berlin, Phys.-Math. Klasse, 1933, 6; A., 676.2. physikal. Chem., 1931, [B], 12, 279 ; A , , 1931, 68850 GENERAL AND PHYSICAL CHEMISTRY.one of the greatest obstacles to the acceptance of the Nernst chainmechanism has been removed.The mechanism of the reactionnow appears to be : 48Cl, + hv --3 c1 + c1,followed by the repeating stepsC1+ H, -+ HC1+ H,H + c1, -+ HCl + c1.Such chains can also be set up by the artificial introduction ofchlorine or hydrogen atoms into a mixture of hydrogen andchlorine.49 I n spite of the simplicity of the chain mechanism, thereaction rate depends in a very complex way on the light intensityand the reactant, resultant, and inhibitor (e.g., oxygen) concentra-tions ; but very complete reaction schemes covering the experi-mental data can now be f~rmulated.~O I n the absence of oxygen(a difficult impurity to remove), the reaction chains may end bythe pracesses :(1) H + HCl-+ H, + C1 51(2) c1 + c1-+- c1, 52(3) c1 + wall -+ ic1,When reactions (1) and (2) are the controlling ones the reactionrate varies as the square root of the light intensity, just as in thehydrogen-bromine photoreaction, and this has been confirmedby experiment.53 When reaction (3) is important, the size of thevessel has an influence on the rate.54 In the presence of oxygenas an inhibitor, new chain-ending processes arise involving theremoval of chlorine atoms (as oxides of chlorine) and of hydrogenatoms (ultimately t o give water) by oxygen molecules.Experi-mentally it is found that the removal of the hydrogen atoms is48 W. Nernst, 2. Elektrochem., 1918, 24, 335; A., 1919, ii, 208.49 W. H. Rodebush and W. C. Klingelhoefer, J. Amer. Chem. SOC., 1933,55, 130; A., 232; P. Harteck, Trans. Faraday Soc., 1934, 30, 134.R.G. W. Norrish and M. Ritchie, Proc. Roy. SOC., 1933, [A], 140, 99, 112,713; A., 576,791 ; M. Bodenstein and P. W. Schenk, 2. phylsikal. Chem., 1934,B], 20, 420; D. L. Chapman and J. S. Watkins, J., 1933, 743; A., 915.5 1 Norrish and Ritchie, Zoc. cit., ref. (50).52 The C1 atoms may form C1, molecules without altering the kinetics;(G. K. Rollefson and H. Eyring, J . Amer. Chern. SOC., 1932, 54,170; A., 1932,348; G. E. Kimball and H. Eyring, ibid., p. 3876).53 D. L. Chapman and 3'. B. Gibbs, Nature, 1931,127, 854; A., 1931, 806;Norrish and Ritchie, Zoc. cit.34 D. L. Chapman and P. P. Grigg, J., 1928, 3233; 1929, 2426; A., 1929,164; 1930, 46ROWXN : CX€EBDCATJ KTNETTCS. 51the more important of the two, and it may occur by a t least twomechanisms :These processes are identical with those which are likely to occurin the thermal and the mercury-sensitised photochemical reactionof hydrogen and oxygen.56Miscellaneous Reactions.It has already been pointed out (p. 48) that experiments on thedirect introduction of hydrogen and of oxygen atoms into hydrogen-oxygen mixtures may throw light on the mechanism of the hydrogen-oxygen reaction. Such atoms may also be introduced by photo-sensitisation : hydrogen atoms by using mercury atoms as sensitiser,or by photodecomposition of ammonia or of hydrogen iodide,67and oxygen atoms by the photodecomposition of nitrogenperoxide. 58The results, though not yet capable of a full interpretation, showsuch a similarity to the thermal reaction that it is very probablethat atoms are concerned in the chain mechanism of the latter.59The behaviour of several other chain reactions has been investigated,wix., the oxidation of phosphine,60 of triethylphosphine,61 and ofcarbonyl sulphideJ62 and the formation of peroxides as chain carriersin the oxidation of a~etaldehyde,~~ and of dimethyl- and diethyl-zinc.64 The relative values of the inhibiting effects in solution of alarge number of inhibitors on the photopolymerisation of vinylacetate and on the thermal autoxidation of sodium sulphite have5 5 The process H + HCl + O,+H,O + C10 may also occur (R.G. W.5 6 J. R. Bates, Proc. N u t . Acad. Sci., 1933, 19, 81; A., 358.5 7 J. R. Bates and D. J. Salley, J . Ajmer.Chem. Soc., 1933, 55, 110; J. R.Bates and G. I. Lavin, ibid., p. 81 ; H. S. Taylor and D. J. Salley, ibid., p. 96,A., 236, 237.58 R. G. W. Norrish and J. A. G. Griffiths, Proc. Roy. Soc., 1933, [A], 139,147 ; A., 236.69 For wall effects in the hydrogen-oxygen reaction, see : C. N. Hinshelwood,E. A. Moelwyn-Hughes, and A. C. Rolfe, ibid., p. 521 ; A., 469; A. T. William-son, J. Amel.. Chern. SOC., 1933, 55, 1437; A., 572.Norrish and M . Ritchie, Zoc. cit.).80 H. W. Melville and H. L. Roxburgh, Nature, 1933, 131, 690; A., 678.61 H. W. Thompson and N. S. Kelland, J., 1933, 1231.62 C. E. H. Bawn, J., 1933,145 ; H. W. Thompson, F. L. Hovde, and A. C. K.63 R. N. Pease, J . Amer. Chem. SOC., 1933, 55, 2753; A., 913.64 H. W. Thompson and N. S. Kelland, J ., 1933, 746, 756; A , , 917 .Cairns, ibid., p. 208; A., 35562 GENERAL AND PHYSICAL CHEMISTRY.been shown to be approximately equal, in agreement with thechain-reaction theory of H. J. L. B a ~ k s t r o m . ~ ~Several new homogeneous unimolecular gas reactions have beeninvestigated : the decompositions of ethyl bromide,66 of methyl andethyl a ~ i d e s , ~ ~ and of trichloromethyl chloroformate.68 As expected,they all tend to become bimolecular at low pressures.69 The thermaldecomposition of ozone and the effect of foreign gases have beenfurther examined.'OCareful investigations of the kinetics of the reactions of halogenswith reducing agents in aqueous solution show that the reactantmay be either the halogen molecule, e.g., chlorine or bromine withphosphorous a ~ i d , ~ 1 bromine or iodine with formic or thehypohalous acid formed by hydrolysis, e.q., bromine and the oxalateion? The trihalide ion does not appear to be a reactant in oxid-ation reactions, but is the reacting species in the addition of iodineto the p-phenylpropiolate ion.74 I n the photoreactions of bromineand of iodine with the oxalate ion, the free halogen acts as an innerlight-filter, the trihalide ion being the photoactive species.75 Fromthis result it may be concluded that the free halogen atom isineffective in the reaction, which occurs between X,' ions formedby photodissociation of the trihalide ion and the oxalate ionsHC,04' and C,Oi'.The rates and modes of addition of halogen acids to unsaturatedorganic compounds have been of great interest to the organic chemiston account of the light they throw on electronic configurations andsteric effects.'6 An attempt to interpret the observed phenomena65 C f .Ann. Reports, 1928, 25, 329; K. K. Jeu and H. N. Alyeiz, J . Amer.6 6 E. L. Vernon and F. Daniels, ibid., p. 922; A., 469.6 7 J . A. Leermakers, ibid., p. 2719, 3098; A., 910, 1017.6 8 H. C. Ramsperger and G. Waddington, ibid., p. 214; A., 232.69 Cf. Proc. Roy. SOC., 1927, [A], 114, 84; 115, 215; 116, 163.70 A. Glissmann and H. J. Schumacher, 2. physikal. Chem., 1933, [B], 21,71 R. 0. Griffithand A. McKeown, Trans. Faraday Soc., 1933,29,6ll; A., 574.7 2 D. L1. Hammick, W. K. Hutchison, and F. R. Snell, J., 1926, 127¶ 2715;A 1920, 32; D.L1. Hammick and M. Zvegintzov, J., 1926, 1105; A , , 1926,691 ; A. von Kiss and A. Urmbnczy, 2. anorg. Chern., 1933,213,353 ; A., 1017.73 R. 0. Griffith, A. McKeown, and A. G. Winn, Trans. Paraday SOC., 1932,28, 107; A., 1932, 344; B. Makower and H. A. Liebhafsky, ibid., 1933, 29,597 ; A., 573.C'hem. SOC., 1933, 55, 575; A., 357.323; A,, 677; B. Lewis, J . Physical Chem., 1933, 37, 633.74 E. A. Moelwyn-Hughes and A. R. Legard, J., 1933,424; A., 573.7 5 R. 0. Griffith, A. McKeown, and A. G. Winn, Tram. Paraday SOC., 1933,29, 309; A., 237; E. Abel, H. Schmidt, and K. Retter, 2. phy8ikaZ. Chem.,1932,163,53 ; A., 237.7 6 H. Burton and C. K. Ingold, J., 1928, 904; 1929, 2022; A., 1928, 634;1929, 1270; cf. Ann. Reports, 1928, 26, 121, 128; 1929, 28, 119BOWEN : CHEMICAL KINETICS.53in conjugated systems by means of wave-mechanics has been made.77It now appears that much of the experimental data on modes ofaddition must be revised because of the unexpected influence ofoxygen. Kharasch and his co-workers 78 have shown that, inabsence of oxygen or of added peroxides, hydrogen bromide slowlycombines with molecules of the type CH,=CHX (X = Br, CH,, etc.)to give CH,*CHBrX. A trace of dissolved oxygen (which formsa peroxide of the unsaturated substance) or of added peroxidescauses the addition to take place more rapidly, giving CH,Br*CH,X.79The decomposition of nitroamine dissolved in isoamyl alcohol,as in water, is catalysed by basic ions.80 From the specific catalyticeffect of the HSO,' ion in the hydrolysis of ethyl acetate, the seconddissociation constant of sulphuric acid is calculated to be 10-2.81In aqueous solution the thermal decomposition of ozone isbimolecular, and between pH 5.3 and 8 it is catalysed by the hydrogenion.The rates of change and the equilibria in the reactions CO, +H,O H,CO, have been measured by methods specially devisedto cope with the high velocities in both dire~tions.8~ This equi-librium affects the theory of the rate of dissolution of solid carbonatesin acids, as freshly liberated carbonic acid contains more hydrogenion than the equilibrium mixture.**In addition to €he photoreactions already described, mentionmay be made of the following reactions.I n ultra-violet light ofwave-length shorter than 2750 A., chlorine monoxide decomposeswith a quantum efficiency of 4-5 at 20°, while a t longer wave-lengths the efficiency is about 2.It seems likely therefore that inthe two regions two types of decomposition occur : 85 C1,O -+2C1+ 0 and C1,O --+ C1+ C10.The photodecomposition and oxidation of phosphine 86 and ofNeutral salts also exert a catalytic effect.g27 7 H. Eyring, A. Sherman, and G. E. Kimball, J . Chem. Physics, 1933,1,586.78 M. S. Kharasch and F. R. Mayo, J . Amer. Chem. SOC., 1933, 55, 2468;M. S. Kharaseh and M. C. McNab, ibid., p. 2531 ; A,, 805; J. C. Smith, Nature,1933, 132, 447.79 See also R. P. Linstead and H. N. Rydon, ibid., p. 643.80 J. N. Bronsted and J. E. Vance, 8. physikal. Chem., 1933, 163, 240; A.,81 H.M. Dawson, E. R. Pycock, and E. Spivey, J . , 1933, 291; A., 471.88 K. Sennewald, 8. physikal. Chem., 1933, 164, 305; A., 678.83 R. Brinkman, R. Margaria, and F. J. W. Roughton, Phil. Tram., 1933,04 C. V. King and Cheng Ling Liu, J . Amer. Chem. SOC., 1933,55,1928; A.,85 H. J. Schumacher and R. V. Townend, Z. physikal. Chem., 1933, [B], 20,86 )I. W. Melville, Proc. Roy. SOC., 1933, [A], 189, 541; A,, 473.471.[A], 232, 65; A., 678.679.375; A., 57754 GENERAL AND PHYSICAL CHEMISTRYcarbonyl chloride 87 are chain reactions initiated by a process ofpredissociation. Diazomethane is decomposed by light with aquantum efficiency of about 4, the primary products being probablyCH, and N,.88 Solutions of N-chloroacetanilide, when illuminated,give, besides p-chloroacetanilide, diacetylhydrazobenzene, chlorin-ated solvent, and other products, indicating that the primarychange may be the liberation of a chlorine akom, as in the case ofnitrogen chloride itself .89The coloured solutions of alkali metals in liquid ammonia arephotoactive, and decompose in ultra-violet light to give hydrogenand the metallic amide.90 The discovery of a method of intensifica-tion of weak photographic images, using silver sulphide, has beenannounced, and may prove of great value in scientific work.91A number of experiments have been made on the effect of super-sonic waves, and of sound waves of audible frequency, and ofhigh energy, on chemical reactions.Colloids such as gelatin aredepolymerised ; but other effects such as the formation of hydrogenperoxide from water and air are claimed.Many of the resultsobtained are probably due to local heating of the solutionsemployed .92Preliminary experiments have been described on reaction rates inunimolecular films on water, such as the oxidation of unsaturatedacids and ester formation in the layer.93All chemical changes involving solids liberating or reacting withgases show “ interface effect,” noticed long ago by Faraday in theexample of the loss of water from hydrated salts. The reactiontakes place wholly a t the interfaces between solid reactant and eithersolid resultant or other suitable solids,94 and the rate is determinedby the rate of growth of the interfaces. These rates of growth may8 7 F.Almasy and T. Wagner-Jauregg, 2. physikal. Chem., 1932, [B], 19,405; G. I<. Rollefson and C. W. Montgomery, J . Amer. Chem. SOC., 1933, 55,142; G. K. Rollefson, ibid., p. 148; A., 237.8 8 F. W. Kirkbride and R. G. W. Norrish, J . , 1933, 119; A., 359.89 F. W. Hodges, J . , 1933, 240; A., 473.90 It. A. Ogg, P. A. Leighton, and F. W. Bergstrom, J . Amer. Chem. SOC.,91 K. Hickman and W. Weigerts, Nature, 1933, 132, 134; A., 915.92 A. Szalay, 2. physikal. Chem., 1933,164, 234; A., 578; H. Beuthe, ibid.,163, 161; A., 473; A. Szent-Gyorgyi, Nature, 1933,131,278; A., 359; E. W.Flosdorf and L. A. Chambers, J . Amer. Chem. SOC., 1933, 55, 3051; A., 915.93 A. H. Hughes and E. K. Rideal, Proc. Roy. Soc., 1933, [A], 140, 253; A.,679; S.E. Bresler, W. W. Druschinin, and D. L. Talmud, 2. physikal. Chem.,1933, 164, 389; A., 672.94 G, N. Lewis, ibid., 1905,52, 310; I. Langmuir, J . Amer. Chem. Soc., 1916,38,2263; C. N. Hinshelwood and E. J. Bowen, Phil. Mag., 1920, [vi], 40,569;A., 1920, ii, 743; Proc. Roy. SOC., 1921, [A], 99,203; A., 1921, ii, 443.1933, 55, 1754; A . , 682ROWEN : CHEMICAL KINETICS. 55thus be measured from the reaction A serious experimentaldifficulty, however, is often encountered in measurements on solid-gas reactions, for, if they are reversible, the equilibrium (owing tocxcessive slowing down of the reaction after it has proceeded someway) often appears to differ considerably if approached from thetwo sides. Possible reasons for this have been advanced.96There appears to be no reason to doubt that interface action is anexample of an increase in rate due to the provision of a new pathfor the reaction with a lower heat of a~tivation.~’ The reactions ofmetallic oxides with hydrogen, which take place at the interfacemetal-oxide, have been theoretically examined from this point ofview.98 For the direct attack by hydrogen of cupric oxide, forexample, a high heat of activation is necessary because the copperatoms produced must first exist in the adsorbed state on the oxidelattice. At a metal-oxide interface the metal atoms can pass easilyto a metallic lattice, and the heat of activation of the reaction islowered.This view represents only one of many possibilities,however, for in a reactionthe relative values of the heats of adsorption of the gases B and Don the solids A, C, and X, as well as those of atoms of C on A and Xmust be considered in any theoretical analysis of problems ofreactions of gases with solids (including catalysis and promoteraction).The enumeration of the possibilities of different paths forthe reaction involving adsorption of the resultants in different waysis easy; the difficulties arise when any attempt is made to discoverprecisely which particular possibility applies to any one chemicalexample.An elaborate investigation into the question of the limits ofapplicability of the differential aeration theory of metallic corrosionhas been carried out.99 Corrosion is often observed to occur wherethe simple theory would not predict it, Le., in places where oxygenis in excess.Dissolved oxygen “ passivates” iron, making itcathodic to othcr iron by forming a film. Under certain conditions,however, the film, particularly when thick and not exposed to freeaccess of oxygen, may apparcntly crack away, leaving anodic areas9 5 IV. E. Garner, A. S. Gomm, and H. R. Halos, J., 1933, 1393. :’I, J. Zawadzki and S. Bretsznajder, 2. physikal. Chem., 1933, [B], 22, 60,97 Cf. C. N. Hinshelwood, “Kinetics of Chemical Change in Gaseous98 A. R. Ubbelohde, Trans. Furuday SOC., 1933,24,532.9 9 Cf. U. R. Evens, “ The Corrosion of Metals,” E. Arnold & Co., London,1926, chap. 6 ; G. D. Bengough and F. Wormwell, Proc. Roy. Soc., 1933, [ A ] ,140, 399; A., 679.Asolid + Bgas + Xsolid -+ csolid + Dgas + Xsolid79; A., 911.Systems,” 3rd edition, Oxford, 1933, pp. 220, 30156 UENERAL AND PHYSICAL CHEMISTRY.where vigorous corrosion may occur, more vigorous even than that onparts of the iron not protected by so thick a film.The phenomenonof '' water-line " corrosion is an example of this type of effect. Acomplete elucidation of all the factors involved is a peculiarlydifficult problem. E. J. B.4. SPECTROSCOPY.The increasing importance of spectroscopic data in chemistry a i dthe prolific accumulation of new facts and theories since the lastreport on this subject 1 warrant a somewhat more detailed accountof several matters which have hitherto been regarded as purelyphysical. It is impossible in the length of t,his report to mention allthe relevant topics, or even enter exhaustively into those given;those subjects chosen have not been more than cursorily discussedbefore.The derivation from spectroscopic material of chemicaldata, e.g., energies of binding, interatomic distances, angles betweenvalency bonds, theories with regard to the mechanism of chemicalchange, and the determination of activation energies, has beengiven elsewhere.2The Energy Levels of Atoms and Molecules.-In contrast t o theenergy of an atom, which is entirely electronic, the energy of amolecule is of three kinds, electronic, vibrational, and rotational.To a first approximation these can be regarded as additive, althoughthe interaction of rotational motion with that of electrons in theirorbits is fundamental in elucidating the fine structure of the bandsin the molecular spectrum.3 It is very largely from a study of thevibration and rotation levels that the chemical data just referredto are obtained.Vibration and rotation of a molecule, however,tend to instability, and the existence of stable molecules is con-ditioned by electronic energy. An understanding of the electronicconfiguration of molecules is therefore essential for the correctinterpretation of valency. In the sense of the Lewis-Langmuir-Sidgwick theories of electron-sharing this has long been realised.On the other hand, a satisfactory reason for electron-sharing hasnever been given, and there are strong reasons for believing thatits interpretation as a pairing of two electrons antisymmetric inspin is inadequate, and although often fortuitously appli~able,~not universally valid.5Ann Reports, 1926, 23, 296; 1929, 26, 16.Bid., 1932,29,46; 1931,28,19; N.V. Sidgwiclc and E. J. Bowen, ibicl.,1931,28, 367; R. S. Mulliken, Chem. Reviews, 1929,6, 526; E. Rabinowitsch,2. Ekktrochem., 1931, 37, 91.See also this vol., p. 79.Mulliken, Rev. Mod. Physics, 1930, 2, 87; 1931, 3, 89.Ann. Reports, 1930, 2'7, 14.R. S. Mulliken, Chem. Reviews, 1931, 9, 347; 1929, 6, 503; Rev. Mod.Physics, 1932, 4, 1 ; Phyakal Rev., 1932, 41, 49THOMPSON : SPECTROSCOPY. 57Several summarising papers 6 and monographs have recentlyappeared on the energy levels and electron configurations of atomswhich are a necessary preliminary to the understanding of molecularelectronic levels.The energy levels-so-called terms-of hydrogenand ionised helium were found empirically to be of the form R/n2,and line frequencies were given by the difference of two such terms.Applying the quantum theory Bohr and Sommerfeld satisfactorilyexplained this result, n being the principal quantum number ofthe electronic orbit. The terms calculated by the semi-Keplerianmethod agreed with those observed. Regarding alkalis as built upfrom a core with one “ emitting electron,” similar terms shouldresult, but in general the values found are B/n*2, n* not beingintegral. This is best interpreted by supposing that n* is compositeand equal to n + a, the quantity cx arising from a penetration ofthe core by the orbit of the outer emitting electron. When penetra-tion does not occur n* becomes integral and the term is said to behydrogen-like.Some terms differ only slightly from those ofhydrogen; in such cases there inay be a slight polarisation of theatom core by the emitting electron. With non-hydrogen-liketerms, penetration causes the emitting electron to become morefirmly bound during part of its orbit, the corresponding increasein the term value implies that n, exceeds n*, and the quantity a,which i s then negative, is called the quantum defect. It is clearthat the shape of the orbit of the emitting electron will decide theextent of penetration. But the Bohr-Sommerfeld theory intro-duced a second quantum number Icy which, besides measuring theangular momentum of the electron in its orbit, essentially deter-mined the eccentricity of the orbit, and it is consequently to beexpected that the value of the quantum defect a will be relatedto this azimuthal or subsidiary quantum number.It is possibleto verify this by determining on the one hand the mean radiusof the atomic core, i.e., the radius of the atomic ion, and on theother, the nearest distance of approach of the emitting electron tothe nucleus. If the former exceeds the latter, penetration willoccur. Mean ionic radii have been calculated for the alkaliswave-mechanically by L. Pauling and by D. R. Hartree,*using the electron distribution function X41cr2t@ for all the6 R. C. Gibbs, Rev. Mod. Physics, 1932,4,278 ; S.Dushman, Chem. Reviews,1928, 5, 109 ; Hund, “ Linienspektren,” Springer, 1927 ; Pauling and Gouds-mit, “ Structure of Line Spectra,” McGraw-Hill, 1930; Jevons, “ BandSpectra,” Cambridge, 1932, Chap. V ; A. Sommerfeld, “ Atombau,” 5thEdn.7 Pm. Roy. SOC., 1927, [A], 114, 1S1; L4., 1027, 394; J. Amer. C‘hem. SOC.,1927,49, 766; A., 1927, 399.Proc. Camb. Phil. Xoc., 1928, 24, 89, 111; A., 1928, 21668 GENERAL AND PHYSICAL CHEMISTRY.electrons t (see below). Further inspection shows that the empiricalvalues of n* fall into groups, in which the members differ successivelyby unity, the defects being constant within a group, but differingfrom one group to another. The quantum defects were expressedas letters s, p , d . . . . , and a term I?/(% + was called an nsterm.From the above it i s clear that to s, p , d, must be assigneddefinite values of the subsidiary quantum number E . In theoriginal theory k = 0, 1, 2 . . . n, the zero value being forbiddensince it would imply a linear orbit passing through the nucleus;E is now replaced by a new quantum number I = k - 1, the changebeing fully justified in the wave mechanics; and in order to complywith the singlet nature of the s levels and doublet p , d . . . found foralkalis, s electrons are assigned a value I = 0, p electrons haveI == 1, d have I = 2, etc. Thus a 4p alkali electron means onein a 4 principal quantum orbit with subsidiary quantum numberunity, the term being R/(4 + p)2, p being the particular correctionvalue for penetration in this state.Again, it is found that s termsonly combine with p terms, p with s and d, d with p and f . . .Expressed differently, this means that there is a selection principlelimiting changes of the angular momentum quantum number I tounity, ke., AZ = This result would be predicted by considera-tions of the correspondence principle and is indicated in the wavemechanics. The doublet nature of p , d, f . . . alkali levels, thesinglet and triplet alkaline earkhs, and other systems of multipletlevels imply the presence of a further factor, which was denotedby Sommerfeld as the inner quantum number. G. E. Uhlenbeckand S. Goudsmit first attributed a spin on its own axis to theelectron, the spin quantum number s being taken empirically as &,and the spin angular momentum 5 4.h/2n.$ Values of the totalangular momentum j are then to be obtained by compounding1 with s. For alkali s ( I = 0) electrons, j = & 4; for p electrons,j = 14 or ; for d electrons j = 2 i or l+, etc. . . . Energetically thetwo s states are then indistinct, in agreement with fact. Theadditional selection rule for transitions, A j = 0 or 9 1, is alsoexplained theoretically.In order to comply with the nomenclature to be used for othertypes of atom, the states given by alkali s, p , d . . . electrons areT 4 = $(s,y,z)eZnivt is a wave function, i.e., solution of the appropriateSchrodinger equation, and I,@ is an electric density, independent of time anddependent only on space co-ordinates.1.Nature, 1926, 117, 264; A., 1926, 215.h - h h$ The replacement of angular niorricnts 1 .by dZ(2 + 1) . and 8 .hby % s(a + 1 ) . 5 is neglected in this reportTHOMPSON SPECTROSCOPY. 59now designated as 8, P, D . . . terms. The multiplicity of thelevel is written above and to the left, and the value of j as a subscriptto the right, and the principal quantum number may precede thewhole. To retain uniformity, the doublet superscript is retainedfor the s levels although they are really singlet. Thus 4,0, impliesthat for the emitting electron n = 4, I = 2, s = + 8, j = $.Assuming the core-model, alkali terms can be calculated approxi-mately by the methods of Bohr and Sommerfeld. But the diffi-culties of the many- body problem-helium or higher niany-electronatoms-are so far insuperable.Whilst Ca+ or Al++ can be regardedas alkali-like, neutral Ca or A1 cannot be treated at all. However,the introduction of electron spin makes the whole system verycomplicated. The consequent failure of the core model led to theintroduction of the vector model in which the various angularmomenta of the particles in the system are compounded togetherquantum vectorially to give the resultant state of the whole. Thisvector model is invaluable in interpreting the magnetic propertiesof atoms and in predicting the possible states for a given electronicconfiguration. As Hund has shown, however, different ways ofcombining a given set of quantum vectors lead to different results.The most common type of " coupling " found is that named afterH. N.Russell and F. A. Saunders,lo in which the I values of allthe electrons are combined, then all the s values, and finally thetwo resultants L and 8. For example, in an atom with two outerelectrons not in closed shells * (for which XI = 0, Zcs = 0), e.g.,calcium, if the I quantum numbers are respectively I,, I,, and thespins sl, s2 (& i), the resultant orbital angular momentum L isone of (I, + I,), ( I I + I , - 1) . . . . . (II - Z2), and the net spin1 or 0. With the latter spin value each L value will give onetotal angular momentum J = L ; with the former there will bethree values L + 1, L, L - 1, and the tcrms will be singlets andtriplets. With three-electron atoms X = Z s = 8 or -2 and J =L + Q, L - 4, or L + 4, L + Q, L - fry L - Q giving doublets andA closed shell contains groups ofelectrons with equal 1. In a magnetic field the E level which is (21 + 1)-folddegenerate gives (21 + 1) values of the magnetic quantum number wz, i.e.,I, 1 - 1, .. . . - 1. Since the spin may be f 3, there will according to Pauli'sprinciple be 2(21 + 1) possibilities for any E group. Since E = 0, 1, . . . (n - l),l = n - 1there will be f2(21 + 1) = 2n2 electrons in a closed n shell. Closed shells areE = O[Is2] [ ( 2 ~ ) ~ ( 2 ~ ) ' 1 [ ( ~ S ) ~ ( ~ P ) ' ( ~ V ~ IK L MIt is obvious how this result applies t o the structure of the periodic classifica-tion.10 Astrophys. J., 1925, 61, 38; A., 1925, ii, 911.* Electrons of equal n form a shell60 GENERAL AND PHYSICAL CHEMISTRY.quartets.I n the same way, four-electron atoms give rise to singlets,triplets and quintets. In this form of coupling the only differencebetween the singlet and triplet states of given L is difference inspin orientation, which should not be considerable ; the rathermarked difference’in the terms is due to a resoiiance phenomenoncharacteristic of the quantum and wave mechanics.ll Expressionsgiving the term energies as functions of the quantum numbers arediscussed by Hund and by Mulliken.The atomic terms are called 8, P, D . . . according asL =: Xi! = 0 , 1 , 2 . . . For example, the ground state of carbonls22s22p2 (Le., two 1s electrons, two 28, two 2p) will have I, = 1,I, = 1, and L = 0, 1, or 2, giving X, P, or D terms; S will be0 or 1, giving singlets and triplets. Actually the ground term is3.2’.Similarly, the ground state of nitrogen (ls22s22p3) is 4X, andthat of sodium (ls22s22p63s) is 2S. The total angular momentumquantum number of the entire atom, J , is obtained as before bycompounding L and X, i.e., L + 8, L + X - 1, . . . L - 8, andmay be written as a suffix. A further distinction is that atomicstates with even values of XI for all the electrons in the atom are“ even” ( 9 ) and with XZ odd are “ odd” (u). The significanceof this can only be satisfactorily understood in terms of the wavemechanics; it can be shown that an even state is such that if itswave function is +(x,ylxlx,y,x, . . .), then+(-xl - yI- - x1 . . .) =#(xlylzl . . .), and for an odd state #(- x1 - y1 - x1 .. .) = -+(x,y,xl . . .). The ground state of carbon is even (Z! = 2)3Pg;of nitrogen odd (XI = 3)4X,; of sodium even (22 = 6)2Ss,, and thefirst excited state ls22s22p23s of nitrogen even (ZZ = 2)2Xs,.The selection rules for transitions are in general that the total spinshall be unaltered, g terms combine only with u terms, AL = 0 or & 1and AJ = 0 or & 1 with no transition J = 0 + J = 0.Just as the ls22s22p2 state of carbon should give rise to theterms IS, lP, lD, 3S, 3P, 3D, so, in general, a given electronicconfiguration will provide many terms. F. Hund12 and J. C.Slater 13 have given rules for determining their relative depths.Usually the greater the multiplicity the lower the term, and forterms of equal multiplicity that one is the lower which has thegreater L value.That the ground term of carbon is 3P ratherthan 3D is due to the fact that the configuration leading to 3 0is forbidden by the Pauli principle.*11 W. Heisenberg, 2. Physilc, 1926, 38, 411; 1926, 39, 499; 41, 239; seealso Ann. Reports, 1933, 29, 17; 1931, 28, 21 ; 1930, 27, 14.l2 2. Physik, 1926, 33, 346.l1 Physical Rev., 1926, 28, 291 ; 1929, 34, 1293.* As a further example helium may be considered. The ground state isAssuming complete decoupling of the vectors (strong magnetic field), la2THOMPSON : SPECTROSCOPY. 61The splitting of atomic energy levels in a magnetic field is wvery important means of deciding their nature, i.e., L value andmultiplicity. If the field is “ weak,” the vectors L and X remaincoupled together and precess around their resdtant J .The latterprecesses around the field axis with a frequency g . 0, where g isthe so-called Land6 splitting factor and o the Larmor precessionfrequency; g depends only on the quantum numbers of the state.The J vector is oriented so as to assume quantised values, M = J ,J - 1, . . . - J , along the field direction; 31 is the magneticquantum number. Each of these orientations has a definiteenergy value which can be calculated (Zeemaizii effect).* Thus, alS0 level does not split, a ID, gives five levels, and a six. Thereis an additional selection rule AM = 0 or & 1 for transitions.In a “ strong ” field, the L,X coupling breaks down, each vectoritself precessing with its own frequency about the field directionand giving quantised components ML and Ms ; M = ML + Ma, butthe energy relationships are different from those in a “weak ”field (Paschen-Back effect).New selection rules, AML = 0 or & 1,AM, = 0, replace these for J and M .I n an applied symmetrical electric field the plane of each electronicorbit precesses around the field axis, but the spin vector of theelectron is not directly affected by the field. The L vector iscontrolled by the applied electric field, the S vector only by themagnetic field due to the orbit,al motion. The details and splittingenergies are given in BIulliken’s r e p ~ r t . ~The essential feature of the wave mechanics is Schrodinger’swave equation,l*1 8x2m h2 C- .A2+ -+ - (E - U)+ == 0 ;U is the potential energy of the system, E the total energy, and$ the wave function, depending on space co-ordinates and time.The imposition of boundary conditions for t,b (it must be finite,continuous, and single-valued at any point, and vanish at infinity)makes it possible to solve the Schrodinger equation only for definifel4 Ann. Reports, 1930, 27, 13 ; see also Hinshelwood, op. cit., pp. 36 et sep.there will be five quantum numbers n,, Z, s, mz, ma. According to Pauli, notwo electrons can be identical inn, I , m,, m,. In the given case L = 0, S = 0 or 1,and the possible terms are lS and 3S. In the strong field each electron wouldhave n = 1, I = 0, mi = 0, ma = - 4 or + $. With the first three quantumnumbers identical, the spins must be antiparallel, Le., the state is singlet IS.For the configuration 1929, however, the triplet state is also possible since theelectrons me now not “ eqnivalent.”Thesplitting of levels actually observed here is determined by Mb.* The Stern-Gerlach experiment involves essentially the same process62 GENERAL AND PHYSICAL CHEMISTRY.discrete values of E. If the equation is expressed in sphericalpolar co-ordinates r, 8, +, and for a hydrogen atom U = e2/r besubstituted, the E values are simply the stationary states of thehydrogen atom.Writing + = R(r) q e ) a(+), where R(r), 0(8), a(+)are functions of r, 8, and + only, the equation can be “ separated ”into three complete differential equations each of which only hassolutions if certain parameters have proper values.These para-meters are the quantum numbers n, E = 0, 1, 2, . . . (n - l),and m = 0, & 1, -J= 2 . , . & 1. The q,’s for different states cannow be expressed as functions for the co-ordinates and of thequantum numbers. I n general $ passes through several maximaand minima in passing from the nucleus to infinity. Places inwhich t,h = 0 are called nodes, and the frequency of their occurrenceis intimately connected with the quantum n~rnbers.1~ Actually,in Cartesians # = $ (x,y,z)e2xivt, where i = drl so that @ isindependent of time. tb represents the electron density a t anypoint specified by the co-ordinates, or the probability that theelectron will be found at this point.$xr2 . ,,@ is called the electrondistribution function and if this quantity is plotted against T (Le.,nucleus a t the origin) the distribution of electricity in the electroncloud is obtained.* For s electrons, E = 0 and o(e) O(+) turnsout to be constant ; in this case $ depends only on r, i.e., is sphericallysymmetrical. The electron-distribution function will also bespherically symmetrical in this case. For other, e.g., p , levels,this does not hold, and the particular type of symmetry obtainedis important in bond formation.I n cases other than those of hydrogen or ionised helium (themany-body problem again) the wave equation cannot be solved.By using approximations helium has been studied by J. A. Gaunt,16E. A. H~dleraas,~~ and I. Waller,18 lithium by V.Guillemin, jun.,and C. Zener,lg and some other atoms by Recently amethod of solving the equation to obtain atomic terms and wavefunctions has been achieved by means of the Hartree “self-con-1 5 Darrow, “ Einfiihrung in die Welleiunechanik,” Hirzel, 1929 ; Caatel-l6 Proc. Camb. Phil. Soc., 1925, 24, 332; A., 1928, 685.1 7 2. Physik, 1929, 54, 347; A., 1929, 616; 1930, 65, 209; 1930, 66, 453;A., 1931, 143; E. A. Hylleratts and B. Undheim, ibid., 1930, 65, 759; A.,1931, 4.fraiichi, ‘‘ Modern Physics,” Churchill, 1932, Vol. 11, Chap. VIII.Ibid., 1926,38,635; A., 1926, 987.l9 Ibid., 1930,61, 199; A., 1930, 649.2o Physical Rev., 1930, 36, 51; A., 1930, 1234.* This indicates what is meant by saying that an electron can simultaneouslyThe “ mean ” radius of the orbit agrees with that be a t more than one point.of the old Bohr theory.See Pauling and G-oudsmit, op. citTHOMPSON : SPECTROSCOPY. 63sistent ” field,21 also used by Slater,22 V. F0ckJZ3 and J. E. Lennard-Jones.% The self-consistent field for an atom or ion is defined asbeing such that “ the wave functions for the electrons in that fieldgive a distribution of charge which with the nucleus reproducesthe field.” Hartree replaces the many-body problem by a one-body problem with a central field for each electron. The Schrodingerequation written as a sum for all the electrons in the system isreplaced by a sum of separate equations in which the potential isa function of the ith electron only. Wave functions of the latterequation can be obtained, and the total atomic wave functionexpressed as a combination of these one-electron wave functions,so chosen to satisfy Pauli’s principle and be antisymmetric in theco-ordinates of all pairs of electrons.J. McDougal125 has appliedthe method to Si+f++- which is a spherically symmetrical core.A certain field is assumed and the simplified Schrodinger equationfor each core electron (n,Z) is solved. In each case the equationhas, for any E value, one solution which is zero at the origin andone which is zero a t infinity. For certain values of E these solutionsare identical, i.e., they join smoothly a t intermediate radii. Theapproximately correct E value is found by trial and error. Byusing this, the one-electron wave function (~ri follows.The sumof the charge densities given by && is thus found. This quantityis added for all electrons except one, the potential due to the resultingcharge calculated and compared with the field assumed for oneelectron. Approximations are made again by trial and error untilthe assumed and the calculated field agree. The sum densitydistribution then gives the “ self-consistent field.” The separatecharacteristic energy values of the one-electron wave equations arevery close to the spectral terms. To get the optical terms ofSi++’- t , the entire calculations should now be repeated with introduc-tion of the series electron 38, or 423. . . . It is, however, possibleto regard this series electron as moving in the field of the coreSi - - + I , and solve the Schrodinger equation using the potentialgiven by the self-consistent field.Estimations for thirteen termsagree well with fact ; slight discrepancies may be due to a polarisationof the atom core not allowed for. It is interesting to find that forterms of the same value of I the energy values vary widely, but thequantum defects remain the same.Hartree’s method involves certain approximations in the way the21 D. R. Hartree, Proc. Camb. Phil. SOC., 1928, 24, 89; A., 1928, 216.22 Physical Rev., 1929, 34, 1293.23 2. Physik, 1930, 61, 126.x Proc. Gamb. Phil. Soc., 1931, 27, 469; A,, 1931, 1109.25 Proc. Roy. SOC., 1932, [A], 138, 550; A., 10764 GENERAL AND PHYSICAL CIIErnSTRY.wave functions of the individual electrons are combined to give theentire atomic wave function. Slater and Pock used more accurateforms but their equations cannot be manipulated mathematically.D.R. Hartree and (Miss) M. M. Black 26 have extended the work tocalculate the terms, ionisation potentials and mulfiplet separationsof 0, O+, O++, O+++. Lack of spherical symmetry introduces com-plications, and when a 2p electron is removed from Of to give Of+the wave functions of the remaining 2s and 2p electrons are noticeablychanged; the O+ is not an unperturbed 0++ core with added 2pelectron. D. R. Hartree 2’ has now given the atomic wave functionsand self-consistent fields for C1- and Cu+. Others soon to be publishedinclude Al+++, K+, Rb+, Cs+. It is interesting to compare theresults for Cu+ (Group I b ) with those for Na+ (Group Ia).One point not usually made obvious is the exact connexionbetween the quantum numbers which appear as characteristicparameters of the wave equation and those taken as vector angularmomenta in the model used above.The characteristic parametersare connected with a number of nodes. It is, however, possible toformulate the solutions of the wave equation so that the vectorialnature of I is apparent.28 In the relativistic wave theory the precisenature of s a n d j should appear too.If the number of stable electronic levels of a molecule was com-parable with the number often found €or atoms, the analysis of aband spectrum would be in many cases impossible. Many expectedmolecule states are, however, unstable, i.e., with purely repulsivepotential-energy curves or curves with shallow minima at largere ; they are nevertheless important in explaining certain con-tinuous spectra * phenomena involving collisions of the secondkind, and the pressure broadening of spectral lines.The excitationof a molecule to many states (and their consequent detection) isalso to some extent limited by its dissociation under the conditionsby vibration into some other state. The observed electron levelsof H2 and He, closely resemble the He term diagram; there aresinglets and triplets, the relative depths and multiplet separationsbeing similar. A. Fowler 29 first noticed that some of the levelscould be expressed as Rydberg terms R/n*2 and band heads given26 Proc.Roy. SOC., 1932, [A], 139, 311; A., 202.27 Ibid., 1933, [A], 141, 282; see also J. C. Slater, Physical Rev., -1929, 35,210; 1932,42,33; A., 1932, 1187.28 E. Rabinowitsch, 2. Elektrochem., 1931, 38, 370 ; W. Weizel, “ Banden-spektren,” Handb. expt. Physik (Erganzungsband).39 Proc. Roy. SOC., 1915, 91, [A], 208.* The hydrogen continuum produced in a Iow-pressure discharge tubs isdue to a transition from the stable excited (lsa)(2s~)~Z,+ state to the unstable(ls~~)(2pa)~82 (W. Finkelnburg and W. Weizel, 2. Fhysik, 1931, 88, 577)THOMPSON : SPECTROSCOPY. 65by difference of terms as for atoms. Moreover, as Mulliken firstreali~ed,~O the few lowest known electron levels of CN, N,+, CO+,BO, BeF, MgH, each containing thirteen electrons, are closely similarto those of Na and Mg+.This might be expected if the atoms eachretained their two ?z = 1 (K electrons) and shared an octet of eightn = 2 electrons, leaving one n = 3 outer or valency electron.I n these circumstances the 21-electron molecules SiN, A10, MgFshould have similar levels to K or Caf. Some analogy is found,but there is ample evidence to indicate that the close parallelbetween Rydberg atom levels and molecule terms cannot be SUS-tained.* Mulliken 31 has explained the matter fully, indicatingwhy Rydberg-like terms arise for H, and He,.The vector model is fundamental for an understanding of theelectronic levels of molecules.32 An outer electron may be regardedas moving in the electric field due to the core.I n diatomic mole-cules this field is axially (not spherically) symmetrical along theinternuclear axis. The term values should then be analogous tothose obtained from the “ united atom ” in a strong electric field(Stark effect). There should be groups of levels, as in the atomgroups, of different ML, each with sub-levels of different Ms. Briefly,the L vector is resolved along the internuclear axis to give a newquantum number A. In place of ML, Ms, and M , the symbolsA, Z and SZ are used. Ah/2x and Zh/2x are the components ofthe orbital and spin angular momenta along the internuclear axis,and a . hl2x is the component of the total angular momentum alongthis axis. The unresolved orbital and spin angular momenta aredetermined by L and 8.A = ML can take the values0, 1, 2 . . . L, but in the absence of perturbations every levelwith A>O will be doubly degenerate, owing to the two oppositedirections of revolution (& ML). The presence of molecularrotation will remove this degeneracy ( A-type doubling ; compareJevons, op. cit., p. 126). = A +C = (A + 8) ( A 3- 8 - 1) . . . . ( A - 8). really re-place L, 8, and J in the notation for atomic levels; terms withA = 0, 1, 2 . . . are 2, 11, A . . . states like the 8, P, D . . . levelsof atoms. 2rIh means A = 1 , 8 = 4, X = - 9, R = 8. Two otherclassifications are necessary: + and - terms are distinguishedaccording to the symmetry or asymmetry of the wave functionwith respect to a reflexion in any plane passing through the nuclei;C = 8, X - 1, .. . . -8, andA, Z,3O R. S. Mulliken, Physical Rev., 1925, 26, 561.3 1 Idem, Rev. Mod. Physics, 1932, 4, 59.32 Hund, L L Linienspektren ” ; Mulliken, Rev. Mod. Physics, 1930, 2, 60.* E.g., G. Herzberg, 2. Physik, 1929, 52, 815; R. C. Johnson and R. K.Asundi, Proc. Roy. Soc.. 1929,123, [A], 560.REP.-VOL. XXX. 66 GENERAL AND PHYSICAL CHEMISTRY.even (9) and odd (u) levels are distinguished when the nuclei of adiatomic molecule are similar according to the symmetry or asym-metry of the wave function with respect to a reflexion through themid-point of the line joining the nuclei. As in atoms, states withCZ even are g ; those with CZ odd are u. The selection rules fortransitions are similar to these in the Stark effect, A S = 0 or rarely& 1, i.e., change in multiplicity zero or rarely 2, also AA = 0 or& 1, AX = 0.The analogy with atomic nomenclature is further preserved indescribing the individual electronic-quantum numbers.33 To anyelectron are assigned four quantum numbers n, I, A, and s, wheren, I , and s have the usual meaning, and A is the measure of thecomponent of orbital angular momentum along the internuclearaxis.Electrons with Z = 0, 1, 2, . . . . are s, p , d electrons,and those with A = 0, 1, 2 . . . are G, x , 6 . . , electrons." A4po electron means one with n = 4, I = 1, A = 0. Very often,however, the n,Z quantum numbers of molecular electrons cannotbe specified and have no real meaning. This corresponds withthe inability to " separate " the appropriate Schrodinger equationinto elliptical co-0rdinates.3~ The A quantum number in suchcases nevertheless remains distinct.Mulliken refers to such electronsas xo, yo, wx, . . . etc. Lennard-Jones 35 has used 013, o*ls,x2p, . . . etc., where the symbols Is, 2p . . . following the o orx imply definite genetic connexions with electrons of the separateatoms. Often the symbols KK are used to represent the four mostfirmly bound electrons in a molecule, these being essentially un-changed inner K electrons of the combining atoms. As in atoms,closed shells exist in molecules. According to Pauli, no two electronscan occupy identical n, I , A orbits (be equivalent) unless their spinsare antiparallel. It follows that two equivalent o electrons willform a closed shell c2, four x electrons (two + A, two - A) x4,four 6 (two + A, two - A) s4.It is also clear that in any closedshell A = 0 and S = 0, i.e., a l X state.The molecule terms arising from a given electron configuration arenow obtained in the same way as those for atoms. One G electrongives a 2Z+ term; one x electron gives 21T; three equivalent xelectrons ( x 3 ) give 2n. Two equivalent x electrons (9) givel A (mil= m12= &1, ML = & 2, Le., A = 2 but S = 0 by Pauli'sprinciple because mzl= m12), lC-I and "c- (mzl= + 1, mL2= - 1,In addition g G+= u and + z$ + or - _--1 -.33 R. S. Mulliken, Chem. Reciews, 1929, 6, 503.34 W. Weizel, op. cit., ref. (28).3 5 J. E. Lennard-Jones, Trans. Faraday SOC., 1929, 25, 668 ; A., 1929, 1360.* With equal nuclei even and odd electrons must be distinguished, e.g.,ub and a,, according to symmetry propertiesTHOMPSON : SPECTROSCOPY.67A = 0, and S = 1, as well as X = 0 since mil =I= mi2). Otherexamples are given by M ~ l l i k e n , ~ ~ F. H ~ n d , ~ ’ and G. Her~berg.~*Hund has discussed the dependence of the energy on A and Sfor a given electron configuration; as in atoms, for a given valueof A terms of highest multiplicity lie lowest.It is impossible here to enter into the application of the waveequation to the study of diatomic molecules. Pauling39 has givenan account of the calculations for H, and H2t. An excellent anddetailed treatment can be found in Weizel’s book. The resultsgive full substantiation t o the vector model described above.Itmay also be observed that the method of the Hartree self-consistentfield may prove very valuable in this connexion.The splitting of a level under the influence of a magnetic field,often used to determine the nature of atomic levels, is not verysuitable for deciding experimentally the type to which a molecularstate belongs, for the splitting is as a rule very complicated andthe spectrum analysed with difficulty. The simplest and mostusual criterion employed is to study the rotation fine structure ina band involving transitions to or from the level. Other propertiesof a state, e.g., magnetic character, are sometimes used.The theoretical prediction and discussion of the electronic levelsof diatomic molecules in terms of the vector model has been con-tributed mainly by F.Hund,4O R. s. M ~ l l i k e n , ~ ~ and W. Heitlerand G. Her~berg.~, Three lines of approach have been followed.First, the “united atom” isoelectronic with the molecule to beconsidered may be hypothetically subjected to an axially symmetricelectric field, the nucleus being split in the desired manner. Theterms and configurations of the resulting molecule should followfrom the original atomic terms and configurations. For example,the lowest levels of the nitrogen atom (ls22s22p3) are 4S, ,D, 2P.It can be shown * (Hund, Mulliken, Zocc. cit.) that in the field (Starkeffect) these terms respectively give rise to the molecular statesfor CEI 4C-; 2C-, , A ; %+, 211. The known states of the CHaG Rev.Mod. Physics, 1932, 4, 10.3 7 2. Physik, 1930, 63, 727; A., 1930, 1226.38 “ Molekulstruktur,” Hirzel, 1929, p. 175.39 Chern. Reviews, 1928, 5, 173.40 2. Physik, 1926, 36, 657; A., 1926, 657; 37, 742; 1927, 40, 742; A.,1927, 183; 42, 93; A., 1927, 495; 1930, 63, 719; A., 1930, 1226; 2. Elelctro-chern., 1928, 34, 441 ; Ergeb. exakt. Naturwiss., 1929,S.41 Physical Rev., 1928, 32, 186, 761; A., 1928, 1067; 1929, 226; 1929, 33,730; A., 1929, 740; Rev. Mod. Physics, 1932, 4, 1.42 2. Physik, 1929, 53, 52 ; A,, 1929, 629 ; G. Herzberg, ibid., 57, 601 ; A . ,1929, 1367 ; also “ Molekiilstruktur,” p. 167.* 8 terms ( L = 0) only give B(h = 0); D terms (L = 2) give 2(A = O ) ,II(A = l), A(A = 2), and P terms ( L = 1) give X(A = 0 ) and n ( A = 1).68 GENERAL AND PHYSICAL CHEMISTRY.molecule are 2X-, 213, 2A, 2Zi;+, and the absence of 4Xc- is probably tobe explained by the weak transitions from this quartet to the otherdoublets.The second 211 state is probably unstable. If, insteadof considering what happens to the terms themselves, i.e., to Aand 8, the changes in the electron configuration are followed underthe field’s action, three possibilities arise, the ( l ~ ) ~ ( 2 ~ ) 2 ( 2 p ) 3 be-coming( lsc~)~( 2s0)~( 2p~)~(2rpx)or (ls.)2(2sO)2(2pa)(2rpx)2or ( is0)2(2sQ)2(2px)3(An s electron can only become Q, a p either o or x . ) The first stategives a 211 term ; the second 4X-, 2C-, 2A, 2C+ ; and the third 2II, i.e.,the same result is obtained as before. For further details Hundand Mulliken’s papers should be consulted.The method is fairlysuccessful for hydrides of the type given, but not so good for othermolecules.The second procedure is to bring together the neutral atoms andfind the molecule terms arising from given atomic terms. Hund(Zoc. cit.) and E. Wigner and E. E. Witmer 43 have given the rulesinvolved.X2); the orbital angular momentum vectors L, and L2 quantisethemselves along the internuclear axis and then combine to giveA ; 44 e.g., O(3.P) + excited N(2D) give one each of 2iD, *GI, twoeach of 2A, 4A, three each of 211, 411, three each of 2Z, 4Z. Again,normal nitrogen (4S) + O(3P) give one each of 211, 411, 611 and oneeach of 2Z, 4C, 6C.If the approaching atoms have equal nuclei and are not in thesame state, twice as many levels result from the exchange degeneracy,each of the above occurring as an even and an odd term.Withequal nuclei in the same state the degeneracy disappears but theterms are alternately even and odd.It is interesting to observe that two normal oxygen atoms (3P) giverise to 3&-, 3&-, 3Cg+ terms which do not combine with each other ;consequently, the upper level involved in the atmospheric oxygenbands must be formed from one normal and one excited atom, animportant fact in estimating the energy of dissociation of oxygen.45It i s not possible, however, to give general rules for determining thequantum numbers of the individual molecular electrons resultingfrom the union of two atoms in specified L, 8, J states.The resultant spin vector X = (8, + 8,) .. . . (8, -43 2. Physik, 1928, 51, 859; A., 1929, 117.44 See Kronig, “ Band Spectra and Moleoular Structure,” Cambridge, 1930,46 G. Herzberg, 2. physikal. Chem., 1029, [B], 4, 223; A,, 1930, 277; 1930,pp. 24-25.[B], 10, 189; A,, 1930, 1524THOMPSON : SPECTROSCOPY. 69The third line of approach is essentially that of molecule buildingused by Bohr for atoms. The wave functions and energy levelsof one electron in the presence of the nuclear skeleton are firstdetermined, and the electrons are then added one a t a time, goinginto the lowest available orbits, mutual electronic interactionsbeing ignored. If the electron configuration can be determined inthis way the terms are deduced as above.The relationships are,however, more complicated than in atoms, since the orbit typesare more varied, and the order of binding depends, not only onthe atomic numbers of the nuclei, but also on internuclear separ-ations and other factors.46 The advantage of the method, especiallyover the previous ones, is that it explains the similarity in termschemes and physical properties of different isoelectronic molecules.Moreover, in deducing the terms of nitrogen and carbon monoxidefrom N + N and C + 0 by the second method, there is no indicationof which are the molecular ground states and why they are identicalin the two cases. J. E. Lennard-Jones4' first applied this methodof electron configuration and called it the method of molecularorbitals.Molecular orbits (one-electron wave functions) werecalculated for such molecules as nitrogen, oxygen, and fluorine ;it was concluded that some electrons should be shared whilst othersremain localised round a particular nucleus (cf. 01.8, 0 ~ 1 8 , 7r2p above).The extent of sharing was one of degree only. Localised electronswere in atomic orbitals or levels. E. 13iicke148 used a similarmethod to study the properties of a C=C double bond. Heformulated the latter as consisting of two electron pairs [aI2[7rl2,where o has a rotational symmetry about the C--C axis and xhas an angular distribution proportional to sin 4 (4 = 0 in thecommon plane of all nuclei). The stability of the double bond torotation is connected with the properties of the [ ~ ] , j , 6 orbital.E'or nuclei which are not very different, Hund, Herzberg, Mulliken,and Weizel49 have estimated the order of electron levels and theapproximate way in which it changes with changing internuclearseparation. I n this connexion ionisation potentials are useful.They have given rules (assignment or correlation rules) for statingthe atomic level which will give rise t o any particular molecularlevel when the atoms are brought together.Although predicting the possible molecular terms, none of theprocedures above outlined originally indicated whether the levels413 R.S. Mulliken, Rev. Mod. P h y s k , 1932, 4, 40, 41.4 7 Trans. Paraday Xoc., 1929,25, 668; A., 1929, 1360.48 Z. Physik, 1930,60,423; A., 1930,525; Ann. Reports, 1930,27,17; 1931,49 2.Physik, 1930, 59, 320; A., 1930, 394.28, 1770 GENERAL AND PHYSICAL CHEMISTRY.corresponded to stable states or not. Heitler and London firstgave a quantum-mechanical explanation of binding with specialreference to hydrogen, the interacting atoms being in S states.50They also calculated the energy of interaction. Band-spectrumdata in the normal 1 2 states of H,, Li,, Na,, each formed fromtwo normal 2 8 atoms, are in good agreement with theory; thepredict,ed ,I: repulsion state of H, is also known. But there arevery many disagreements, e.g., according to theory two normal ISatoms should only give rise to a repulsive lCI state, whereas casesare known (Ca,, Zn,, Hg,) of stable molecules. Again the theoryfails to explain the stability of oxygen and the 3C ground state ofthis molecule.Extension of the theory by Heitler 51 and othersso as to include interaction of P terms has not proved very success-ful, and the attempts to generalise it into a theory of spin valencyin which valency bonds involve a pairing of two electrons anti-symmetric in spin must be regarded as unacceptable. 52 Neverthelessi t is important to remember the Heitler and London method ofcalculating energies of molecule formation.Herzberg has attempted to explain the stability or instabilityof a molecule by enquiring whether in its formation the atomicelectrons are moved to a higher or a lower level. The potential-energy curve of a molecule can then be regarded approximatelyas the superposition of all the separate potential-energy curves ofthe individual electrons, measured as the atoms approach.Electronswill then be of two types, bonding and antibonding, according towhether they contribute a positive or a negative amount of energyto the binding. A stable molecule will be formed when the numberof bonding electrons exceeds the number of antibonding ones.Herzberg suggests that the number of bonding pairs minus that ofantibonding pairs gives the valency of a bond, and explains satis-factorily the facts for simple molecules such as 13,l-, €I,, He,, Li,,Mulliken has arrived a t the idea of bonding and antibondingelectrons too. In the formation of 211 CH from 3P carbon and 2 shydrogen,( 1 ~ ) ~ ( 2 ~ ) ~ ( 2 p ) ~ , 3P + ~ s , ~ X -+= ( l~o)~(2so)~(2po)~(2p7c),~II or in othersimilar processes, i t is seen that 1s hydrogen electron hasbecome " promoted," Le., its principal quantum number hasincreased.This promotion, which can be regarded as a continuousN,, 0,.50 Ann. Reports, 1930, 27, 14; R. H. Fowler, Rep. Brit. ASSOC., 1931, p. 226.61 W. Heitler, Physikal. Z., 1930, 31, 185; A., 1930, 525; J. H. Bartlett,52 R. S. Mulliken, Chern. Reviews, 1932, 9, 347; Herzberg, " Molokul-Physical Rev., 1931, 37, 507; A., 1931, 548.6tXTIktUr. ' THOMPSON : SPECTROSCOPY. 71process as the atoms approach, occurs with a number of electronsin the formation of CN, CO, N,, etc. Mulliken points out fromexamples that, in general, promoted electrons tend to give repulsivepotential-energy curves.He regards a promoted electron as anti-bonding, and attributes in general to any electron a bonding powerwhich may be positive, negative, or zero. Mulliken's terminologydiffers from that of Herzberg in the sense that a promoted electroncould, according to the latter, in certain circumstances, be bonding ;but their essential ideas are similar. In a later paper Mullikendiscusses this point and introduces the term premotion.Mulliken has attempted to generalise the idea of bonding powerof electrons with a theory of v a l e n ~ y . ~ ~ For example, in theformation of H2 two processes are possible :H( IS) + H( IS) -+ H2(( 1 ~ 5 ) ~ 'C,+)H( IS) + H( IS) -+ H2( 180, 2p0 3Zui). orI n the former case the two electrons are bonding and a stablemolecular state results; in the second case the promotion energyof the IS electron to the 2130 state is so large that it more thanbalances that gained by the firmer binding of the Is0 electron,hence no stable molecule i s formed.I n the union of two helium atoms there would similarly be fourIS electrons; in accordance with the Pauli principle, two of thesemust be promoted and the antibonding effect prevents moleculeformation.With one excited helium atom this difficulty would,however, not necessarily arise.Mulliken stresses the fortuitous nature of antisymmetric spinarrangements in the Heitler-London theory and points out thatsome molecules (e.g., NaH) have states in which the pair of electronsforming the bond have parallel spins, whilst other cases exist inwhich one electron appears to bring about a bond.He discussesin detail the structure of many molecules and explains the saturationof valencies in terms of closed shells. He prefers to regard valency,not as a whole-number property of atoms, but as a more flexiblequantity depending upon variable electronic bonding powers, theoccurrence of bonding electrons in pairs being incidental as far aschemical binding is concerned.In a critical review of the various methods used in the theoryof valency, Hund 53 also apparently concludes that the propertiesof molecules are more likely to be interpreted in terms of molecularorbitals than in terms of electron-pair bonds. G. N. Lewis hasrecently reviewed the theory of the electron-pair bond.5453 Z.Physik, 1931, 73, 1; ibid., 1932, '73, 565; A., 1932, 10, 215.54 J . Chem. Physics, 1933, 1, 17; A., 21172 GENERAL AND PHYSICAL CHEMISTRY.Mulliken’s report 55 contains a valuable and exhaustive computa-tion of the potential-energy curves of most of the electron statesof many diatomic molecules.The first important wave-mechanical calculations on the structureof polyatomic molecules were made by Pauling 56 and J. C. slate^-,^'using what has been called the method of electron-pair bonds.Their work consists essentially in a generalisation of the Heitler-London theory, but instead of considering the interaction of atomsin definite states they suggest that in the presence of the strongperturbation such as occurs in molecule formation the atomic statesshould not be treated separately but merged together.Theydetermine new atomic wave functions which are hybrids of theusual ones and are able to associate definite directions with each.This leads to the idea of directed (orbital) valency. The inter-action of two wave functions, belonging severally to an electronin each atom, is considered, and the maximum degree of bondingis assumed to result when the wave functions overlap to the greatestextent. The theory has been very successful in interpretation ofthe covalent link. One objection often raised to it is that it involvesa hybridisation of the s, p wave functions.Mulliken and others have developed the alternative method ofmolecular orbitals or one-electron wave €unctions for polyatomicmolecules.Both atomic and molecular orbitals may be thought ofas defined in accordance with the Hartree method of the self-consistent field. Unshared electrons are described in terms ofatomic orbitals, and shared electrons in terms of molecular orbitals.The underlying idea comes from the work of H. Bethe,5* whoexamined the perturbation of the orbitals of an atom when sur-rounded by other atoms in a crystal. By methods of group theory,he showed how the s, p , d . . . levels of an atom split under per-turbations of different symmetry. E. Wigner 59 extended the workto include all the 32 classes of crystal symmetry. G. Placzek 60has further elaborated the result with special reference to nuclearvibrations in polyatomic molecules.The novelty lies in the con-sideration of the effect of symmetry properties on orbitals in mole-cules belonging to one of the 32 symmetry point groups, the~ 5 5 Rev. Mod. Physics, 1932, 4, 1.5 6 J . Amer. Ohem. Sac., 1931,53,1367,3225; A., 1931, 670, 1356; 1932, 54,988, 3570; A., 1932, 561, 1191; Ann. Reports, 1931, 28, 287, 367; Fowler,Rep. Brit. ASSOC., loc. cit., ref. (50).57 Physical Rev., 1931, 3’7, 481; A., 1931, 548; 1931, 38, 325, 1109; A , ,1931, 1113, 1356.58 Ann. Physik, 1929, [v], 3, 133 ; A., 1929, 1367.59 Nach. Qes. Wiss. Cfcittingen, Math. Klassa, 1931.6O Marx’s “ Handb. der Rdologio,” 1933THOMPSON : SPECTROSCOPY. 73symmetry being given by the nuclear skeleton. This leads to theelectronic configurations of the molecules.In his first paper, Mulliken 61 considers four types of symmetry :regular tetrahedral (CH,), in which p levels do not split; linear(OX', HF), in which p splits into Q and n; trigonal pyramidal(NH,, OH,'), in which p splits into [GI and [XI, a somewhat modifiedG and x ; and finally isosceles triangular (H,O), in which p splitsinto three types designated u, b, c.The amounts of splitting canbe estimated by group theory. The (r orbitals are non-degenerate,x two-fold degenerate. The appropriate molecular structures arethen as follows :OH-, HP : ls22~22p22px4 ; NH,-, H,O : ls22s22pa22pb22pc2 ;H,O+, NH, : ls22~22p[n]~2p[~]2; NH4+, CH, : 1s22s22p6.A comparison of the structures of carbon tetrafluoride and tetra-iodide is interesting. In the former case the electrons around thecarbon atom probably constitute from the latter's point of viewclosed shells, ls22s22p6 ; the other electrons in the molecule, i.e.,those especially attached to the fluorine, are probably very remotefrom the carbon.From the point of view of the fluorine nucleusthere are ls22s22pG, but the rest of the molecule sets up a field ofthe trigonal type and splits up the 2p6 into 2p[0]~, 2p[nl4, of which2p[0I2 are shared with the carbon nucleus. The four pairs 2p[aI2of the fluorine atoms are then all shared with the carbon, whichregards them as constituting its outer shell 2s22p6. Comparisonof this result with that for the tetraiodide leads, when estimates of thewave functions are made, to the idea that about six electrons only(not eight) hold the four halogen atoms in the iodide. Thus thistetraiodide is unstable.The structures of BF,-, SO4--, C104-, SO,--, and C10,- are alsodiscussed on the basis of assumed symmetry.With the last, un-shared electrons are found, in agreement with the work of W. H.Zachariasen.62 The linear molecule of carbon dioxide is shown to be0 ~ 0 ~ ~ 2 2 ~ ~ ~ ~ ~ ~ x ~ n ~ , which agrees with its diamagnetic character. Int,hc case of PtCl,-- the splitting of orbitals is considered.In a more detailed paper, Mulliken G3 reviews critically the Lewistheory of valency, comparing the properties of molecular orbitalswith those of electron-pair bonds. He points out that many rulesknown about valency find interpretation in the occurrence andbehaviour of non-degenerate and degenerate orbitals : e.g.,0, is .. . . n2, the x molecular orbital is two-fold degenerate, so61 Physical Rev., 1932, 40, 56; A., 1932, 562.62 J. Amer. Chem. Xoc., 1931, 53, 2123; A., 1931, 897.63 Physical Rev., 1932, 41, 49; A., 1932, 902.c 74 GENERAL AND PHYSICAL CHEMISTRY.that the electrons may have parallel spin, S = 1, giving the known3C- state. It is not necessary to have two electrons to get a bond :one electron in a bonding orbital has a bonding effect.*The molecular orbitals are expressed in zeroth approximations aslinear combinations of atomic orbitals. If in the union of H(1s)and H+, the 1s atomic orbitals for the respective nuclei are $A and+B, the molecular orbitals will be$0 = cO($A + $B) ; $1 = c l ( $ A -where co,cl are normalising factors depending on the nuclear separa-tion.The energy change in the binding can be understood byconsidering the mean charge density e2$2 corresponding to eachmolecular orbital :$02 = co2($A2 + +a2 + 2 4 A 4 B ) ; $I2= cl2(+A2 -k $B2 - 2$A+B)*This shows that e2+$ is relatively more concentrated in the regionbetween the nuclei than if a mere overlapping of e2(q5A2+ +B2)occurred ; on the other hand, +12 is relatively much less concentrated.This added or diminished concentration is decisive in making $*bonding and $1 antibonding. The procedure can be applied tomore complicated symmetrical and to unsymmetrical molecules,the appropriate linear combinations of each atomic orbital beingtaken.It is clear that in order to have a large bonding energy$A$B should be large, which implies a large overlapping of atomicorbitals. This is Pauling and Slater’s criterion of bond formation;the point here is that overlapping orbitals are demanded only,regardless of whether the resulting molecular orbital is occupiedby two electrons, by one, or by none. A molecular orbital whichis occupied by two electrons might be regarded as being essentiallythe same as an electron-pair bond.F. Hund 64 has arrived at similar conclusions to the above.M. Dunkel’s work 65 is also important. In two further papersMulliken G6 has explained in detail the method of applying Bethe’sresults to determine the number, spacing, and degeneracy of electronstates of a molecule, and the transition possibilities.It is necessaryin the language of group theory to know the different “ irreduciblerepresentations ” of the symmetry group for the given nuclearskeleton. Applying the theory to CH, and hence to ethylene,two cases are considered; in one the CH, radicals are “ planar ”84 2. Physik, 1931, 73, 1, 565; A . , 1932, 10, 215; 1932, 74, 429; A., 1932,65 2. physikcal. Chem., 1930, [B], 7 , 81; [B], 10, 434; A., 1930, 525.6 6 Physical Rev., 1932, 41, 751; A., 1932, 1190; 1933, 43, 279; A , , 339.* Pauling has also suggested the possibility of one-electron links.449THOMPSON : SPECTROSCOPY. 76to each other, in the other they are perpendicular. The normal stateof “ planar ’’ ethylene turns out to beand the first excited state isl~2~~2[s]2[s]2[y]2[[y]2[x + 2]2[x + 2732 singlet,1s21s2[s]2[~]2[y]2[y]2[z + xI2[2 + x][x - x] singlet and triplet;[s], [y], [XI, and [XI are molecular orbitals built up as linear combin-ations of those of the 1s H and 28, 2ps, 2py, Zp, C orbitals.I n“ perpendicular ” ethylene, the x orbitals overlap, giving a bond[X -+ 212, but the [XI do not. The levels in this case fall very nearto each other. The relative term depths calculated indicate thatthe lowest enera It is, however, shown that theexcited “ plans$ state lies higher than, and can intercombinewith, a ‘‘ perpendicular ” level. This suggests a mechanism forthe photochemical isomerisation of cis-tyans isomerides.In a later work Mulliken 67 constructs molecular orbitals formolecules RX, and discusses the structure, ionisation potentials,and stability of CH,, NH,, H,O, NH,, CH,, etc.The structurels22[sI22[23]6 can be assigned to CH, without introducing a, hybrid-isation of s, p orbitals.C. P. Snow and C. B. Allsopp 68 have discussed the C=C bond andfind experimentally that alteration of the symmetry relationshipsin ethylenic compounds does not materially affect the levels andtransitions noticed. Lennard- Jones 69 has examined the structureof simple radicals. The change from 0 to NH and CH,, whichare isoelectronic, is followed, and similarly F -+ HO -+ NH,.The reorganisation of radicals in dissociation processes is alsodiscussed.70 Huckel 71 has also investigated the electronic arrange-ments in certain organic free radicals.In studying the case of methane, J.H. van Vleck 72 has shownEathematically that there is justification for Mulliken’s idea thats-p hybridisation does not occur. Slater and Pauling’s method oflocalised bonds avoids excessive electronic accumulation on thesame atom, but it is shown by van Vleck that this accumulationcan occur. The Slater-Pauling method does not allow the wavefunctions to be of a symmetry type appropriate to a tetrahedral field.The above can only be regarded as a rather fragmentary accountof a very complicated subject, but it may serve to indicate thegeneral trend of the problems involved. A real understanding ofthe subject matter can only be gained from a more detailed studyof the literature referred to.state is planar.67 J.Chern. Physics, 1933, 1, 412.6% 699 7% 7 1 Trans. Faraday Soc., in press.72 J . Chem. Physics, 1933, 1, 177, 21976 GENERAL AND PHYSICAL CHEMISTRY.Eihtle mention can be made of the vibration and rotation levelsof molecules; nor can the practical consequences of the inter-action of electronic and rotational motion be discussed.* Fordiatomic molecules the elementary theory has been outlined inprevious reports. For polyatomic molecules the levels and theirstatistical weights are summarised from a theoretical standpointby D. M. Dennison 73 and examples are given by D. S. Villar~.~*For rotational levels three cases can be distinguished, a collinearmolecule, a symmetrical spinning top, and an asymmetric top.G.Herzberg and E. Teller 75 have discussed the selection rules forvibrational quanta in electronic transitions. Different types ofvibration depending on symmetry relationships are distinguished.It is shown that in absorption from the ground state only thosebands appear which are associated with the excitation of total-symmetrical vibrations.HyperJine Structure and Nuclear Moments.-It is not quite trueto say that all the energy of an atom is electronic. I n general aquantum number i has to be introduced in addition to n, I , s, j andm, due to the fact that the nucleus spins on its own axis.76 Thisnucleus spin has angular momentum i . h/Zx and a magnetic momentwhich interacts with the extranuclear electrons. I n terms of thevector model i and j must be compounded quantum-vectoriallyto give f, the total atomic angular momentum.Thus f = (j + i),(j + i - l), . . . . (j - i ) or (i + j), (i + j - l), . . . . (i -j)according as j or i is the greater, and there will be, instead of eachoriginal j level, (2; + 1) or (2j + 1) levels. The interaction energiesand relative separation of the levels can be evaluated from thequantum numbers.'' Their very small magnitude gives the namehyperfine structure. Thus, in the bismuth spectrum the 0 2level splits into four, the 2D, into six, the 2P, into two, so thati > I. From the interval relationships the value found isA quite different interpretation of hyperfine structure is, how-ever, possible; this explains the effect as due to the presence ofdifferent isotopes.Not the mass, but the actual structural differenceof the isotopic nuclei is responsible, giving each isotope a, slightlyi == $.7875 Rev. Mod. Physics, 1931, 3, 280.74 Chem. Reviews, 1932, 9, 369; cf. Mecke, " Molekiilstruktur," p. 23.7 5 2. physikal. Chem., 1933, [B], 21, 410; A., 766.7 6 W. Pauli, Naturwiss., 1924, 12, 741.7 7 L. Pauling and S. Goudsmit, " Line Spectra," Chap. XI.7 8 S . Goudsmit and E. Back, 2. Physik, 1927, 43, 254, 321; A., 1927, 706;* Cf. Rabinowitsch, 2. Elektrochem., 1931, 38, 451; also Jevons, Weizel,1928, 47, 174; A., 1928, 340.opp. citTHOMPSON SPECTROSCOPY. 77different non-Coulombian field. Though it is not always possibleto differentiate spin and isotope effects, the Zeemann effect ishelpful.In the application of this subject t o the investigation of nuclearstructure, the magnetic moments of the nuclei and the so-calledg(1) factors (ratios of magnetic to mechanical nuclear moment)are important.‘9 Measurements of these quantities for differentatoms alre multiplying very rapidly and it is impossible to list allthe results here; i values can be found in Pauling and Goudsmit’sbook and especially in the report of Kallmann and Schuler.80 Typicali values are He4, 0 ; Li9, 0 ; Li7, g ; Naa, $; CP5, g; Brsl, 3 ;I n the past yearmany atoms have been studied, including Cu, Al, In, Ga, Rb, As,Zn, N, Kr, E’, Sn, Sb.81 S. Goudsmit 82 and E. Fermi andE. Segr& g3 have examined hyperfine structure theoretically.The spin of the proton can be determined from the alternation inintensity of lines in rotation bands in the hydrogen molecularspectrum.84 As for the electron, the spin i = Bohr units.Onthis basis it should be possible from observed i values to ascribe.nuclear configurations. There is, however, still some doubt whetherintranuclear electrons spin, and this must at present be left anopen question. A discussion of the facts has been given by H.Schuler and H. Westmeyer,85 and by N. S. Grace,86 who concludesthat nearly all nuclei with large magnetic and mechanical momentsare of odd atomic number, whilst those with small or zero momentsare of even atomic number.Another method can be used to determine the magnetic moments7g J. C. McLennan, Proc. Roy. SOC., 1932, 136, [A], 754; A., 1932, 791;8o Ergeb.exakt. Naturwiss., 1932, 11, 134.81 R. Ritschl, 2. Physik, 1932, 79, 1; A., 2 ; G. Breit, Physical Rev., 1932,42, 348; A. G. Shenstone and J. J. Livingood, Physical, Rev., 1931, 37, 1023;R. C. Gibbs and P. G. Kriiger, ibid., p. 656, 1559; A., 1931,992; H. E. White,ibid., p. 1175 ; D. A. Jackson, Z . Physik, 1933,80,59 ; J. S. Campbell, PhysicalRev., 1931, 38, 1906; Nature, 1933, 131, 204; A., 334; D. A. Jackson, Proc.Roy. SOC., 1933, [A], 139, 673; A., 439; Ill. F. Crawford and A. M. Crooker,Nature, 1933, 131, 655 ; A., 547 ; H. Schiiler and H. Westmeyer, 2. Physilc,1933, 81, 565; A., 547; E. F. Bacher, Plqsical Rev., 1933, 43, 1001; A., 767;H. Kopfermann, 2. Physik, 1933, 85, 353; A., 1095; 83, 417; A., 880; J.X.Campbell, ibid., 84, 393; A., 991; S. Tolansky, Nature, 1933, 132, 318; A.,1095; J. S. Badami, 2. Physik, 1932, 79, 206; A., 2.;. An interesting case is that of lead.81G. Gamow, “ Atomic Nuclei and Radioactivity,” Oxford, 1931, p. 23.82 Physical Rev., 1933, 43, 636; A., 552.83 2. Physsik, 1933, 82, 729; A., 759.84 Jevons, “ Band Spectra,” p. 139; Weizel, “ Bandenspektren,” p. 140;S 5 Naturwks., 1933, 21, 674; A., 1101.86 Physical Rev., 1933, 44, 361 ; A,, 1101.Kronig, “ Band Spectra and Molecular Structure,” p. 9478 GENERAL AND PHYSICAL CHEMISTRY.of atoms, even when they are very This is the Stern-Gerlach experiment, which involves the splitting of levels in aninhomogeneous magnetic .field. 88In two papers of fundamental importance, R.Frisch and 0.Stern 89 and I. Estermann and Stern have described the measure-ment of the magnetic moment of the proton. That the mechanicalmoments of both electron and proton are $.7~./2n was statede 1 h above. The magnetic moment of the electron is 2 . _I 2mc'2'2x' aunit Bohr magneton. If similar relationships apply to the proton,its magnetic moment will, by virtue o€ the different mass, betimes smaller than a, Bohr magneton. This has been called aunit nuclear magneton. The hydrogen molecule is peculiarlysuitable for a Stern-Gerlach experiment : very low temperaturescan be used, giving large deflexions, and the technique of platedevelopment has been well worked out. Moreover, there arethe two forms, ortho and para, differing in nuclear spin arrange-ment.I n para-hydrogen the spins are opposed and the nucleiccntribute nothing to a magnetic moment. The latter can onlyarise from a molecular rotation with quantum numbers 0, 2,4, . . ,With ortho-hydrogen the spins are parallel, and in addition to themagnetic moment resulting from the molecular rotation (quantumnumbers 1, 3, 5, . . .) there will be a nuclear magnetic momentdouble that due to a single proton. Two determinations aretherefore made. At the temperature of liquid hydrogen, para-hydrogen correctly proves to have no magnetic moment. Atsomewhat higher temperatures, two quanta of rot'ation energy aretakeh up, and from the measured magnetic moment associatedwith these, that corresponding to one quantum is calculated. Atthe temperature of liquid hydrogen again, ortho-hydrogen possessesone quantum of rotational energy. If the magnetic momentassociated with this is subtracted from the whole measured, themagnetic moment due to the two nuclei follows. The remarkablefact is that the value for a single proton is approximately 2-5 nuclearmagnetons.I n addition the rotational magnetic moment is about one nuclearmagneton. This can only be explained if it is assumed that theelectron cloud " skids " during the r o t a t i ~ n . ~ lThis result is likely to be of profound significance.87 F. Knauer and 0. Stern, 2. Physik, 1926, 39, '7808; A,, 1927, 92; 1929,58, 766; A , , 1929, 490; J. B. Taylor, ibid., 1929, 5'2, 242.8 8 0. Stern, ibid., 1926, 39, 751.89 Ibid., 1933, 85, 4; A., 996.90 Tbid., p. 17; A., 996.91 G. C. Wick, ibid., p. 25; A., 997BOWEN : M0LECIlJLA.R STRUUTURE. 79JfisceZZaneo~s.-~4 few matters of a general nature may bementioned. Very many investigations during the year concernthe absorption spectra of polyatornic molecules, but the discussionof these must be postponed. The development of ultra-violet andvacuum spectroscopy technique has led to the study of the spectraof more highly ionised atoms. Typical examples are OIII, OIV, Ov,Brv, BrVI, BrVII, MnVI, FeVII, CoVIII, NIX, RbII, RbIII, RbIV, SbVI, etc.As the first of a series of papers compiling critically the data onatomic and molecular spectra of the halogens, W. E. Curtis andS. F. Evans have discussed iodine.92 Finkelnburg 93 has summarisedthe facts and interpretations suggested for continuous spectra ofgases and vapours. The application of the Franck-Condon principleto the potential-energy curves of molecules underlies the interpreta-tions. W. A. Noyes 94 has outlined the correlations between spectro-scopy and chemistry, and Norrish’s work 011 the photochemicaldecomposition of polyatomic molecules containing in particular thecarbonyl group has been continued.95 D. Chilton and E.Rabinowitsch 96 have measured the absorption spectrum of iodinein the adsorbed state. Mention of 0, has again been made bysome workers. H. W. T.Molecular Structure.The methods of determination of intramolecular distances, angles,and force constants from spectroscopic data have been consolidatedduring the year. The long-wave rotational absorption spectra ofwater, ammonia, and phosphine have been examined, leading, inthe case of ammonia, to the distances (in A.U.) : N-€€ = 1.01,M-H = 1.61.1 The most important advances have been madein the interpretation of vibrational data of molecules (obtained fromnear infra-red, Raman, and electronic spectra), from which anglesand force constants can be evaluated. Measured vibrationalfrequencies for a simple molecule must first be assigned to specificmodes of vibration according to the rules of Placzek and Dennison.2The assignment of vibrational frequencies to the molecules of sulphurdioxide and chlorine dioxide shows that their apex angles are both92 Proc. Roy. SOC., 1933, 141, [ A ] , 603; A., 1096.g3 Physikal. Z., 1933, 34, 529.94 Rev. Mod. Physics, 1933, 5, 280.9 5 Trans. Paraday SOC., in press ; J., 1933, 1533.96 Z. physikal. Chenz., 1932, [B], 19, 107; A., 2.1 J. Kiihne, Z. Physik, 1933, 84, 722; A., 998; N. Wright and H. Bf.Randall, Physical Rev., 1933, 44, 391; H. J. Unger, ibid., 43, 123; cf. An?%.Reports, 1932, 29, 62.2 G. Placzek, 2. Physik, 1931,70, 84; G. Placzek and E. Teller, ibid., 1933,81, 209; D. M. Dennison, Rev. Mod. Physics, 1931, 28080 GENERAL AND PHYSICAL CHEMISTRY.about 120" while that of the molecule chlorine monoxide is prob-ably about The assignment of the twelve fundamental0 0 frequencies of the molecule o>N-N<o is discussed by G. B. €3. M.Sut herland.A matter of great moment in work on the vibrations of moleculesis the question of the interaction of the vibrational frequencies.Generally, the separate frequencies of different links can be treatedas roughly independent and constant. The cyanogen halides, whichare linear and have the nitrile structure,6 form a good example ofthis, and it has been shown that the four fundamental frequencies, treated as three-CH, C1 CH, OHof the molecules \CH/ and \C*,/particle systems, do not interact to any great extent.' I n cases,however, where one fundamental frequency is nearly equal toor is nearly a multiple of another, resonance interaction must beconsidered, and this subject has received theoretical and experi-mental investigation.8 The two symmetrical vibration frequenciesof the water molecule are almost equal, and are represented by theabsorption band at about 3700 cm.-l (2.7 p) and the Raman line at3650 ~ m . - l , ~ while the unsymmetrical frequency appears as the6.25 p absorption band. Complete agreement, however, has notyet been reached as to the exact values of the two symmetricalfrequencies.10 Besides the question of resonance, the effects of thedeviations of the vibrations from harmonic motion l1 and of theinteractions of rotation and vibration are important. In one case,that of the carbon dioxide molecule, where the experimental dataare very accurate and extensive, a complete treatment, taking intoaccount all these factors, has been given by A. Adel and D. M.Dennison.12 They derive an equation containing twelve constants(viz., the three fundamental vibrational frequencies, their an-C. R. Bailey and A. B. D. Cassie, Proc. Roy. SOC., 1933, [A], 140, 605;Z. W. Ku, Physical Rev., 1933, 44, 376; L. Lotmar, Z. Physik, 1933,83, 765.C. R. Bailey and A. B. D. Cassie, Proc. Roy. SOC., 1933, [ A ] , 142, 129.Ibid., 141, 342.W. West and M. Parnsworth, J. Chern. Physics, 1933, 1, 402.P. C. Cross and J. H. van Vlock, ibid., p. 350; cf. G. Buss, Z . Physik, 1933,J. Horiuti, ibk?., 84,380 ; A., 998 ; TV. V. Norris and H. J. Unger, PhysicalD. A. Rank, J . Chem. Physics, 1933,1,604.82, 445; 0. Eichmann, ibid., p. 461.Rev., 1933, 44, 467; L. Tisza, 2. Physilc, 1933, 82, 48.lo J. H. van neck and P. C. Cross, ibid., p. 357; R. Mecke, W. Baumann,l1 Cf. the case of acetylene; G. B. B. M. Sutherland, Physical Rev., 1933,l2 Ibid., p. 716; 44, 99; E. F. Barker and A. Adel, ibid., 44, 185; A., 998.and I<. Freudenberg, 2. PhysQ, 1933, 81, 313, 345, 465.43, 883BOWEN : MOLECULAR STRUCTURE. 81harmonic constants, and rotational constants) which closely describesthe behaviour of the molecule up to a t least its fifth vibrational level.The anharmonic constants of vibration so determined open up anew line of inquiry, as they form a measure of the interaction ofthe non-linked atoms (on the older conceptions of valency) of themolecule. E. J. B.R. P. BELL.E. J. BOWEN.C. N. HINSHELWOOD.I€. W. THOMPSON.J. H. WOLFENDEN

ISSN:0365-6217    

DOI:10.1039/AR9333000013

出版商:RSC    

年代:1933

数据来源: RSC

摘要:

INORGANIC CHEMISTRY.1. ATOMIC WEIGHTS.THE striking contribution of mass-spectrum photometry to ourknowledge of the atomic weights of the elements forms the subjcctof F. W. Aston’s Liversidge lecture 1 in which the results obtainedup to December 1932 are summarised. Of the 35 elements examined,19 agree within 1 part in 4000 with the chemical values; a corre-spondence as close perhaps as can be expected in view of the presentuncertainty in the proportion of the oxygen isotopes. The remaim-ing elements examined exhibit a concordance of 1 part in 1000 orgreater, with the notable exceptions of B, P, Sc, Nb, Ta, Se, Ru, and0s. For all of these the International value exceeds the physicalvalue, varying from 1 part in 200 for ruthenium to 1 part in 780in the case of phosphorus.The physical value for boron is admittedly too low on accountof the enhanced photographic effect of its lighter isotope, but thediscordance of the elements scandium (1 in 320), niobium (1 in 232),and tantalum (1 in 360) calls for further investigation, especiallyin view of the recent redetermination of the atomic weight of thelast metaL2Aston found these elements to be simple, and no indication ofhigher isotopes was discovered, from which he concluded that thechemical atomic weights were too high.This has since proved tobe the case for another element, selenium, which in Aston’s tablewas shown to exceed the photometric estimate by 1 part in 330.0. Honigschmid and W. Kapfenberger ,3 after indicating sourcesof error in the older chemical values, have now linked selenium tosilver by the synthesis of silver selenide, a series of closely con-cordant ratios being obtained which give with Ag = 107.880, Se =78.962 instead of the older value 79.2 and in close agreement withAston’s figure 78-96.No recent values are available for rutheniumor osmium. The atomic weight of phosphorus calls for revision inview of the close agreement between the value found by M. Ritchie 4from the limiting density of phosphinc and that of Aston.K. R. Krishnaswami, J., 1930, 1277; A., 1930, 976.2. anorg. Chern., 1933, 212, 198; A., 659.Proc. Roy. SOC., 1930, [A], 128, 551; A., 1930, 1104.1 J . , 1932, 2888WHYTLAW-GRAY : ATOMIC WEIGHTS. 83A full account of the revision of the atomic weight of tellurium 5has now appeared.From the ratios TeBr4 : 4Ag : 4AgBr by the classicalmethods, Te = 127.61, was obtained whilst a synthesis of silvertelluride yielded an identical figure which is 0.1 unit higher thanthe previously accepted value, but 0.4 unit lower than that obtainedwith the mass spectrograph by Aston in 1931. That the chemicaland physical values now agree is due to K. T. Bainbridge,6 whosucceeded in obtaining a much better mass spectrum of this elementwhich disclosed three new isotopes of lower mass.During 1933 a number of important papers have been publishedby Baxter and his co-workers on the atomic weights of arsenic,lead, indium, thallium, potassium, and caxium. I n the light ofexperience gained with the preparation of various volatile inorganichalidcs, G.P. Baxter, W. E. Shaeffer, AT. J. Dorcas, and E. W.Scripture have prepared in a highly purified state arsenic trichlorideand tribromide, and analysed them by 6he standard method ofprecipitation with silver followed by nephelometric titration. Fromthe ratios AsC1, : Ag and AsBr, : Ag, a value for arsenic of 74.906was obtained, distinctly lower than the International value 74.93.In a later paper, Baxter and W. E. Shaeffer published furtherresults from the ratio AsC1, : 1,05, which gave mean values for arsenicranging from 74.906 to 74.916. They concluded that 74.91 was aclose approximation for the true value for this element and thatJ. H. Kfepe1ka”s result obtained in 1930 was too high. The newvalue is now 0.01 unit lower than the physical value 74-92.In the paper on thallium,1° Honigschmid, Birckenbach, andKothe’s value of 204.39 is codrmed by the analysis of thallouschloride, giving the value 204.40 in close conformity with Aston’sfigure of 204.41, but somewhat higher than the figure of 204.34found by H.V. A. Briscoe, S. Kikuchi, and J. B. Peel.10“Attention has often been drawn to the discrepancy between thechemical and the physical values for the atomic weight of czsium.As a rule when there is a divergence the chemical value is the higher.In the case of this element the reverse is true, and the physicalvalue is the greater by 0.1 unit or 1 part in 1330. Now czsium is asimple element and its packing fraction was measured in 1932 111933, 212, 242; A., 761.5 0. Honigschmid, R.Sachtleben, and K. Wintersberger, 2. anorg. Chem.,6 Physical Rev., 1932, [ii], 39, 1021 ; A,, 1099.7 J . Amer. Chem. SOC., 1933, 55, 1054; A., 442.8 Ibid., p. 1957; A., 659.9 Coll. Czech. Chem. Comm., 1930, 2, 255; A., 1930, 976.10 G;.P.Bsxter and J. S.Thomas, J . Arner. Chem. SOC., 1933,55,2384 ; A., 882.11 F. W . Aston, Proc. Roy. SOC., 1032, [A], 134, 67; A., 1932, 209.Pmc. Roy. SOC., 1931, [A], 133, 440; A., 1931, 134884 INORGANIC CHEMISTRY.and found to be - 5 5 2, corresponding to an atomic weight onthe chemical scale of 132.917 so that the physical evidence is verystrong. This discrepancy has now been explained and a chemicalvalue obtained of 132~91.1~ 1200 G. of the less soluble fractionsobtained from the recrystallisation of 4 kg.of cRsium nitrate fromt,he mineral pollucite were recrystallised three times as perchlorateand, after conversion into chloride and fusion, again recrystallisedthrice. The arc spectrum showed the absence of potassium andrubidium, and analysis after fusion in platinum in a quartz bottlingapparatus gave the value quoted.Results obtained with indium l3 by improved methods and withdue regard to the solubilities of the silver halides gave concordantvalues for the ratios InCI, : 3Ag and InBr, : 3Ag ; the average figurefound gives I n = 114.76.Interest attaches to the determination of the atomic weights ofradiogenic leads, especially in view of recent work with the massspectrograph on ordinary lead and lead extracted from purespecimens of pitchblende and thorite and used in the form of leadtetramethyl.Ordinary lead14 showed the presence of 5 newisotopes of which 204 was the most abundant. The mean massnumber found was 207-190, in close agreement with the acceptedchemical value and with new values found independently inAmerica and Germany. No lines corresponding to the new isotopeswere found in the uranium and thorium leads. The purest sampleof the uranium lead from Katanga consisted entirely of isotopes206 and 207 in the proportion of 93.3 to 6.7, giving a mean massnumber of 206.067. The atomic weights of three uranium leadsfrom the same source have been determined by 0. Honigschmid,R. Sachtleben, and 13. Baudrexler,15 who get values (agreeingwithin the limits of error) the mean of which is 206.03, in agreementwith Aston.Baxter and Alter,16 however, found from Katangapitchblende values slightly below 206, and for lead extracted fromRedford cyrtolite 1' (a hard zirconium silicate containing uraniumand lead) the lowest recorded figure for uranium lead, zlix., 205.92 -J=0.02. It is difficult t o believe that this lead can have a value aslow as this, for it would indicate a very small percentage of actiniumlead, the isotope 207, but since the Baxter and the Honigschmid valuefor ordinary lead agree closely, the cause of this divergence is obscure.12 G. P. Brtxter and J. 8. Thomas, J . Amer. Chem. SOC., 1933, 65, 858; A.,333.13 G. P. Baxter and C. M. Alter, ibid., p.1943; A., 659.14 F. W. Aston, Proc. Roy. SOC., 1934, [A], 140, 535.1 5 2. anorg. C h ~ m . , 1933, 214, 104; A., 1099.16 J. Arner. Chern. SOC., 1933, 55, 2785; A., 882.1 7 Jbid., p. 1445; A., 550WHYTLAW-GRAY : ATOMIC WEIGHTS. 85Another apparent discrepancy between the results of the Americanand the German specialists in the determination of atomic weightsby classical methods was furnished by the results recently publishedfor potassium by Baxter and W. M. MacNevin,18 who after a mostcareful investigation concluded that the true value for this elementfrom the mean of 31 experiments from the ratio KC1 : Ag was 39.094with C1= 35-457 and Ag = 107.880. This result confirms theolder value K = 39.096 obtained in 1907 by Richards and A.Staehler l9 and by Richards and E.Mueller,20 who analysed respect-ively the chloride and the bromide, but it is lower by nearly 0.01unit than the result of Honigschmid and J. Goubeau,21 K = 39.104,which was adopted by the German Atomic Weight Commissionin their 8th report 22 and agrees with the value obtained by Richardsand E. H. Archibald 23 as early as 1903. A little more than a fort-night after the appearance of Baxter’s paper, Honigschmid andSachtleben z4 reported a new determination of this constant fromthe ratio KC1 : Ag and MBr : Ag, in which, from the mean of 39 veryconcordant experiments, they deduced the value K = 39.096 withC1 = 35.457 and Ag = 107.880, a value, it is true, differing in-appreciably from Eaxter’s but diverging from their earlier figureby an amount far outside the errors of experiment.Though theauthors are convinced of the correctness of the new value, they areunable to explain why the earlier preparations gave the higherresult. There seems little doubt now in view of Baxter’s results thatRichards and his collaborators obtained a figure very near thetruth 27 years ago. No doubt this result will lead to a revision ofthe KNO, : KC1 ratio measured by E. Zintl and J. Goubeau 25 in1927, which, like the ratios for potassium chloride and bromide,gave the high value for potassium, but it is disquieting to find thatan error has crept into fundamental ratios which were regardedas affording confirmatory evidence of the correctness of the presentvalue for the atomic weight of silver.It was only in 1929 that the German Atomic Weight Commission 26concluded that sufficient evidence was available to decide betweenthe two values for this important standard, viz., Ag = 107.871,based on the ratios LiC10,:LiCl:Ag of Richards and H.H.18 J . Amer. Chem. SOC., 1933,55, 3185 ; A., 994.19 Ibid., 1907, 29, 623.20 Ibid., p. 639; A., 1907, ii, 615.21 2. anorg. Chem., 1927,163,93; A., 1927, 806.22 Ber., 1928, [B], 61, 1.23 Proc. Amer. Acad. Arts Sci., 1903, 34, 373 ; A., 1903, ii, 366.24 2. anorg. Chem., 1933, 213, 365; A., 994.2 5 Ibid., 1927, 163, 302; A., 1927, 806.26 Ber., 1929, 62, 186 INORGANIC CHEMISTRY.Willard,27 and Ag = 107.879 obtained by Richards and G . s.Forbes 28 by the synthesis of silver nitrate.This small difference of 1 part in 10,000 caused a much greateruncertainty in the values of a number of elements.Thanks chieflyto the work of the Munich school, it was found possible to decide infavour of the higher figure, and the Commission based their decisionmainly on two investigations : (1) the reduction of the purest silvernitrate to silver by hydrogen, and (2) the analysisz9 of bariumperchlorate and barium chloride which links silver to oxygen by theratios BaC10, : BaC1, : Ag. The first research, carried out with anexceptionally fine technique, yielded for the analysis of silvernitrate a mean value for the ratio AgMO, : Ag of 1.57479, identicalto the last place of decimals with the synthesis of the same saltcarried out 18 years before by Richards and Forbes,28 and gavewith N = 14.008, Ag = 107.879.The second method yieldedvalues for silver varying only from 107.877 to 107.882. Confirrn-atory evidence was afforded by a number of less direct ratios, amongwhich may be mentioned NaNO, : NaCl : Ag,30 the correspondingpotassium ratios, as well as the ratio 2Ag : SO2 : Ag,S04 : AgC1,31so that at the present time very strong evidence is available that theuncertainty in this important sub-standard does not exceed a t mosta few units in the third place of decimals, an accuracy much great,erthan anything achieved so far by mass-spectrographic methods.Many investigations of a similar character using every modernrefinement have in recent years been carried out at Munich ; mentionshould be made of a revision of the atomic weight of chlorine by twosyntheses 32, 33 of silver chloride which confirmed Richards’s value35.457, which, in turn, is only 0.001 unit lower than the mostprobable value for this element calculated from physicochemicalmeasurements by Moles.34 The synthesis of silver s ~ l p h i d e , ~ ~linking the two elements and giving S = 32.066 in conformity withthe older values of I?.P. Burt and F. L. Usher 36 and of Richardsand G. JonesF7 and the determination of the atomic weights of cal-27 2. anorg. Chem., 1910, 66, 229; A., 1910, ii, 292.28 Ibid., 1907, 55, 34; A., 1907, ii, 685.29 0. Honigschmid and R. Sachtleben, ibid., 1929, 178, 1 ; A., 1929, 370.30 E. Zintl and A. Meuwsen, ibid., 1924, 136, 223; A., 1924, ii, 608.31 0.Scheuer, Arch. Sci. phys. nat., 1913, 36, 381.32 0. Honigschmid and S. B. Chan, 2. anorg. Chern., 1927, 163, 315; A.,1927, 806; also J. Amer. Chem. SOC., 1931, 53, 3012.33 0. Honigschmid and L. Birkenbach, 2. anorg. Chern., 1927,163, 315; A.1927, 806.34 2. physikal. Chem., 1925, 115, 61 ; A., 1925, ii, 346.35 0. Honigschmid and R. Sachtleben, 2. anorg. Chem., 1931,195, 207; A.,36 Proc. Roy. SOC., 1911, 85, 82; A., 1911, ii, 389.37 J. Amer. Chern. SOC., 1907, 29, 825; A., 1907, ii, 685.1931, 279WHYTLAW-GRAY : ATOMIC WBIGIHTS. 87c i u n ~ , ~ ~ ytterbium,40 and iodine 41 are impoktant cantri-butions from this laboratory.The question of the exact value for nitrogen is of importame fncalculating the atomic weight of silver.This was discussed in 1927by Moles,42 who carried out an elaborate research on the densityof the gas and obtained the value 14.0082, and the very carefulwork of Baxter and H. W. Starkweather 43 leads t o an almost identicalfigure. Since then the attempt has been made to deduce the atomicweight of this element from the limiting densities of nitrous oxide 44and amrn0nia,~5 but it is doubtful whether it is possible to obtainresults of the highest accuracy yet from physicochemical measure-ments with condensable gases.A gravimetric synthesis of ammonium chloride and ammoniumbromide was carried out in 1931 by Baxter and C. H. Greene.46Gaseous ammonia was weighed condensed on cooled chabazite,allowed to evaporate into solutions of hydrogen chloride and bromideuntil these mere neutral, and the halide then compared with silver.The ratio Ag :NH, was then measured and from these resultsN = 144078 or 14.0080 according as Ag = 107.879 or Ag =107.880.In recent years the determination of the limiting densit'ies ofgases as a method of determining the exact atomic weights of anumber of elements has come into prominence.Suggested originallyby D. Berthelot and Lord Rayleigh, it was used by Guye and theGeneva school to check and control results obtained by purelychemical means, and with its help valuable independent data forthe atomic weights of nitrogen, chlorine, bromine, sulphur, andcarbon were obtained.The application of this principle is strongly advocated by Moles,who, in addition to the work on nitrogen, has recently with L.R.Pire 47 and later with M. T. Salnzar 48 revised the density of carbonmonoxide and obtained a value for carbon = 12.006. The admirable38 0. Honigschmid and K. Kempter, 2. anorg. Chem., 1931, 195, 1 ; A . ,1931, 279.39 0. Honigschmid and W. Kapfenberger, ibid., 1933, 214, 97 ; A , , 1099.40 0. Honigschmid and H. Striebel, ibid., 1933, 212, 385; A., 762.4 1 Idem, 8. physikal. Chem., Bodonstein Festband, 1931, 282; A., 1931,42 Ibid., 1927, 167, 49; A., 1927, 1120.43 Proc. Nat. Acad. Sci., 1926, 12, 703; A., 1927, 194.44 T. Batuecas, 8. physikal. Chem., Bodenstein Festband, 1931, 18; J .4 5 E. Moles and T. Batuecas, Anal. Pis. Qzciin., 1930, 28, 87; A . , 1930,46 J . Amer. Chem. Xoc., 1931, 53, 604; A., 1931, 407.4 7 Anal.Fis. Quim., 1929, 27, 267; A., 1929, 873.48 Ibid., 1932, 30, 182; A., 1932, 566.1208; 2. anorg. Chem., 1932,208, 53.Chirn. physique, 1931, 28, 572; A., 1931, 1222.135788 INORGANIC CHEMISTRY.work, too, of Baxter and Starkweather on the densities of helium,neon, argon, nitrogen, and oxygen a t a series of pressures hasextended the accuracy of the standard methods of weighing gasesto very fine limits.Unless the densities of gases can be compared at very lowpressures, which cannot be done at present with the requisite degreeof accuracy, a knowledge of the compressibility is necessary, andthe accurate determination of this factor from 1 atmosphere downto low pressures presents many difficulties, especially when, as isthe case with the liquefiable gases, the factor is large and extrapol-ation to zero pressure uncertain.Recent work on the compressi-bilities determined by the usual methods is contained in a paperby Batuecas, C. Schlatter, and G . where a study of thegases nitrogen, carbon monoxide, ammonia, hydrogen chloride,and hydrogen sulphide is reported. W. Cawood and H. S. Patter-son have recently made a study of the PF-P isothermals of aseries of gases over a pressure range of from 1 to 3 atmospheres,and they find, even with gases like carbon dioxide and sulphurdioxide, no indication of curvature. They contend that the low-pressure isat hermals used in calculating the limiting densities arestraight lines to within small limits (1 part in 5000) and there ismuch evidence from high-pressure data 51 and from modern gasequations such as the Beattie-Bridgman equation to support this.If this is SO, the problem of measuring the limiting densities ofliquefiable gases is simpued, though the effects of adsorption a tlow’pressures will need careful study.The same investigators 52 have redetermined the meniscus cor-rections to be applied to mercury surfaces in glass tubes of varyingdiameter, and have obtained data of great value for correctingmanometric readings when the highest accuracy is necessary.Another important error which may amount to 0.05 mm.is thatcaused by irregular refraction, and it is shown how this may bereduced to a minimum by constructing the manometer of verythin-walled tubing.An alternative pruccdure for comparing the densities of gases isby means of the quartz micro-displacement balance, a methodwhich has many advantages.It has been applied to find the atomicweights of xenon,53 of fluorine 55 from methyl fluoride,49 J . Chirn. physique, 1930, 27, 45; R . , 1930, 283.5O J., 1933, 619.s1 W. Wild, Phil. Mag., 1931, 12, 41; A., 1931, 899.53 Patterson and Cawood, Proc. Roy. Soc., 1931, [A], 134, 7 ; A., 1932, 106.54 H. E. Watson, Nature, 1931, 127, 631; A., 1931, 666.55 Cawood and Patterson, J., 1932, 2180.Trans. Paraday SOC., 1933,29, 514; A., 367WHYTLAW-GRAY : ATOMIC WEIGHTS. 89and of carbon.56 The case of carbon is of interest became thechemical atomic weight must be greater than the value obtainedby the mass-spectrograph, which does not take account of thehigher isotope CI3, but a t present it is actually lower, C = 12.00as against C = 12.001 or 12.0023 according to which factor is usedto convert the physical to the chemical scale.The results obtainedfrom a comparison of the densities of carbon monoxide and oxygenat a series of pressures gave 12.011 for the atomic weight of thiselement, in conformity with the latest measurement of the abundanceratio C12 : C13 made by F. A. Jenkins and L. S. Ornstein 57 by band-spectrum methods. It appears as if the higher isotope is presentto the extent of 1%, and that the atomic weight is close to 12-01,though if this is true it is difficult to understand why some lineinvolving C13 has not been detected with the mass spectrograph.Ever since the discovery that the majority of the chemical elementsare mixtures of isotopes, evidence has been accumulating that nodifference in isotopic composition is found in specimens of the sameelement from various sources.The chemical atomic weights arestill to be regarded as natural constants. The work on potassium,already quoted, again supports tbis view for, although both Baxterand Honigschmid extracted their material from various sourcesand each compared the potassium salts of vegetable origin with thoseextracted from minerals, yet no difference in the equivalent wasdetected. The less volatile potassium, however, obtained by idealdistillation by von Hevesy gave the value 39.109, 0.013 unit abovethe new figure. Baxter calculates the percentage of K41 to be 6.6in ordinary potassium and 7.3 in the heavier specimen.Products of radioactive disintegration are, of course, in a differentcategory, and it is of interest to note that again a difference in theatomic weight of calcium 5x, 593 6o in samples from different sourceshas been reported.This calcium of a higher atomic weight issupposed to originate from K41 by a P-ray change.R. W.-G.2. NEW COMPOUNDS.For many years it was supposed that no compound of oxygenand fluorine could exist. In 1927,l however, P. Lebeau andA. Damiens demonstrated that oxygen difluoride, OF,, was pro-duced when water was present in the electrolysis of molten acid56 M. Woodhead and R. Whytlaw-Gray, J., 1933, 846; A., 894.5 7 Proc.K. Alcnd. Wetensch. Amsterdam, 1932, 35, 1212; A., 333.6 8 J. Kendall, W. W. Smith, andT. Tait, Nature, 1933, 131, 688.59 A. V. Frost and 0. Frost, ibid., 1930,125, 48; A., 1930, 130.60 Honigschmid and Kempter, Eoc. cit., ref. (38).1 Compt. rend., 1927, 185, 652; A., 1927, 104490 INORGANIC CHEMISTRY.pota.ssium fluoride below 100". It was found later that the oxidewas best obtained by passing fluorine in a fine stream throughdilute aqueous sodium hydroxide.have isolated a new oxide from the two elements. I n their experi-ments, an equimolecular mixture of oxygen and fluorine a t 15-20 mm. pressure was introduced into a quartz container immersedin liquid air and equipped with two electrodes 12 em. apart. Anelectric discharge caused the formation of 0,F2 as a yellow solid,m.p. ca. - 160". Gaseous 02F2 is brown, and above - 100"decomposes into OF, a colourless gas of very low boiling and freezingpoint, which does not re-form 02F2 on cooling; OF reacts with HIas follows : OF + 3HI = 31 + HF + H,O. Various electronicstructures might be proposed for the molecule 0,F2. Ruff andMenzel suggest the constitution O=O=F, which would be rewrittenin modern nomenclature as O t O t I ? , . They do not, however,indicate what they consider the structure of OF to be, beyondremarking that it must be substantially different from that of 02F,since they have not succeeded in re-forming 02F2 from OF.Another interesting oxide which has recently been prepared issulphur monoxide.Its existence has frequently been postulatedin the Wackenroder reaction and in the genesis of polythionic acids.This monoxide is produced by an electric discharge in sulphurdioxide or, better, in a mixture of sulphur dioxide and sulphurvapour, at low pressures. Nearly pure SO has been isolated undersuitable experimental conditions. At room temperature the oxideis a gas which may be kept for some days in dry vessels. Organicsubstances (such as tap grease) and water vapour4 aid decom-position. The rate of decomposition is favoured by rise of tem-perature, and at 180" it is almost completely decomposed in aminute. Oxygen and the new oxide do not react a t room tempera-ture, but they unite when electrically sparked. Aqueous alkalisform a reducing liquid with the monoxide which decolorises indigoand may be described as a solution of sulphoxylate or hydro-sulphite. Sulphur monoxide reacts vigorously with metals, formingsulphides. When cooled in liquid air, it gives a red condensatewhich behaves in a remarkable way on evaporation, for it decom-poses completely with the separation of sulphur and the productionof sulphur dioxide.The ready spectroscopic detection of themonoxide has indicated that it is one of the products of the reactionbetween thionyl chloride vapour and certain metals such as silver,Z. anorg. Chena., 1933, 211, 204; A., 476.P. W. Schenk, 2. anorg. Chem., 1933, 211, 160; A., 475.E. Gruner, 2. anorg. Chem., 1933, 212, 393; A., 795; P. W. Schenk andNow 0.Ruff and W. Menzel* H. Cordes and P. W. Schsnk, Z. EZeEtrochem,, 1933, 39, 594; A., 1021.H. Platz, ibid., 215, 113WARDLAW : NEW COMPOUNDS. 91sodium, and magnesium, but this method is unsuitable for itspreparation. The monoxide was not detected in the reactionbetween hydrogen sulphide and sulphur dioxide or in the decorn-position of either sodium hydrosulphite or sodium thiosulphate.K. G. Denbigh and R. Whytlaw-Gray6 have made a valuableobservation in fractionally subliming solid sulphur hexafluoride a tlow temperatures. As fractionation proceeded, the vapour densityrose rapidly from ca. 73 (that of SF,) to ca. 125, at which it remainedsteady. They conclude that a new gas S2F1, ( M y 254) is probablypresent with a boiling point in the vicinity of 0".There areindications of other homologues of sulphur hexafluoride.A new hexafluoride has recently been prepared. 0. Ruff andW. Kwasnik (with E. Ascher) report that by heating rhenium ina current of fluorine at 125" the hexafluoride is formed as a colour-less gas which solidifies to light yellow crystals, m. p. 26.4". Fromthe position of rhenium in the periodic table, a fluoride betweenWF, and OsF, would be expected, but these investigators have sofar not obtained evidence of it. Previous experiments on the natureof the rhenium chlorides have shown that more than one compoundis formed when chlorine is passed over the metal. This is clearlydemonstrated by the results of 0. Honigschmid and R. Sachtlebenwhen they attempted to determine the atomic weight of rheniumby the analysis of the chloride. They found the best preparationsto be mixtures, since the proportions of rhenium to chlorine werenot constant and led in the more reproducible cases to the empiricalformula ReCl,.,.They also observed that this substance decom-posed on heating in an atmosphere of nitrogen with the formationof a less volatile product, whereas similar treatment in the presenceof chlorine resulted in complete volatilisation without apparentchange. These results are particularly interesting in view of thediscovery of a pentachloride of rhenium. So far the highest chlorideto be identified is ReCl,,S but it is now found that when chlorinereacts with rhenium, the pentachloride can be produced lo and ob-tained pure by sublimation in a high vacuum at 200".This chlorideis very sensitive to moisture, and with concentrated hydrochloricacid, chlorine is evolved and H,ReCl, is formed. Dilute hydro-chloric acid forms with the chloride a mixture of H2ReC1, andHReO,. Heated in dry nitrogen, the pentachloride yields thetrichloride and an equivalent amount of chlorine. The trichloride,6 Nature, 1933, 131, 763; A., 654.7 2. anorg. Chem., 1932, 209, 113; A., 40.8 Ibid., 1930, 191, 309; A., 1930, 1338.9 H. V. A. Briscoe, P. L. Robinson, and L. M. Stoddart, J., 1931, 2263; A.,10 W. Geilmann, F. W. Wrigge, and W. Biltz, Angew. Chem., 1933,46,223.1931, 125592 INORGANIC CHEMISTRY.which can be purified by sublimation in a vacuum a t 500-550",forms reddish-black lustrous crystals, apparently trigonal, whichare only slightly ionised in aqueous solution.lf This recent workwill possibly modify the conclusions which D.M. Yost and G. 0.Shull l2 drew from their vapour-density data on the rheniumchlorides .Hydrides of silver, copper, gold, beryllium, gallium, indium, andtantalum are formed by the action of atomic hydrogen on thesolid element,l3 and a number of most interesting products havebeen prepared from reactions involving atomic hydrogen at lowtemperatures. Atoms of hydrogen produced in a discharge at apressure of 0.5 mm. of mercury rarely react at room or highertemperatures to give addition products, but a t the temperature ofliquid air or liquid hydrogen such reactions are realisable.K. H.Geib and P. Harteckl* state that a t the temperature of liquidhydrogen it is possible to obtain 100% yields of hydrogen peroxideby the addition of atomic hydrogen to molecular oxygen. Theyconsider that this is not the normal form of the substance butthat it reverts to the usual form on warming to - 115", at whichtemperature there is some decomposition into water and oxygen,accompanied by considerable foaming. A solid substance con-taining 70% HgM is produced from the reaction of mercury vapourwith hydrogen atoms at low temperatures.P. Harteck 15 finds that when nitric oxide is passed, a t liquid-airtemperature, into a stream of hydrogen a t 0.5 mm. pressure, splitinto atomic hydrogen to the extent of 60%, a pale yellow trans-lucent deposit is obtained.This has the empirical formula HNOand proves to be mainly hyponitrous acid with some nitroamine.The same product is obtained from ammonia and oxygen atoms.Although under special conditions hyponitrous acid can be syn-thesised from hydrogen and nitric oxide, apparently sodium hypo-nitrite is not formed when nitric oxide reacts with sodium dissolvedin liquid ammonia at - 50". This reaction is well known butdoubt has frequently been expressed as to whether the product isreally sodium hyponitrite. An investigation by E. Zintl andA. Harder l6 seems to settle this point, for not only does the sodiumnitrosyl, (NaNO),, prove to have different chemical properties fromthose of sodium hyponitrite but Debye-Scherrer diagrams showl1 Idem, Nack.Ges. Wiss. Gbttingert, Math.-phys. Kl., 1932, 579; A., 1130.l2 J . Amer. Chem. Soc., 1932, 54, 4657; A., 218.l3 E. Pietsch (with F. Souferling, W. Roman, and H. Lehl), 2. Elektrochem.,l4 Ber., 1932, 65, [B], 1551; A., 1032, 1098.l5 Hey., 1933, 66, [B], 423; A., 361.l6 Ibid., p. 760; A., 578.1933, 39, 577; A., 1020WARDLAW : STRUCTURES, ETC. 93that they are quite different. Evidently the nitric oxide moleculesare co-ordinated with the sodium in a similar way to carbon mon-oxide molecules when this gas is passed into a solution of the alkalimetals in liquid ammonia at - 50" to give the carbonyls NaCO,LiCO, etc.1' w. w.3. STRUCTURES, ETC.J. R. Partington and C. C. Shah 18 report that molecular-weightdeterminations on benzene solutions of ethyl, propyl, butyl, andbenzyl esters of hyponitrous acid indicate the formula R2N,0, inaccordance with accepted practice.They have also studied thedecomposition of the alkaline-earth hyponitrites and noted that onheating they give chiefly the monoxide of the metal, nitrogen, andnitrous oxide, together with some nitrate, nitrite, and nitric oxide.T. G. Pearson and P. L. Robinson 19 have measured the densityand surface tension of pure sulphur hexafluoride, provided byWhytlaw-Gray, and infer from the derived parachor that the octetvalency ruleismaintainedin this compound and that two of the fluorineatoms are attached to the sulphur atom by covalent linkages andthe remaining four by semipolar singlet bonds, a possibility suggestedby S.Sugden.20 They maintain that the parachors of selenium andtellurium hexafluorides reveal a similar structure for these com-pounds. It is very difficult to accept such a constitution forsulphur hexafluoride, for it does not accord with the known chemicalproperties of the substance, and moreover, as N. V. Sidgwick 21 hasemphasised, the meaning of the results from parachor measure-ments is at present often obscure. I n particular, the effects ofchanges of valency and covalency do not seem to be understood.This opinion seems to be justsed, for L. 0. Brockway and L. Paul-ing 23 arrive at a fundamentally different conclusion from theirexamination of the structures of these hexafluorides by the electron-diffraction method. The diffraction patterns obtained give thefollowing separations (in A.u.) :S-F, 1.58 & 0-03 ; Se-F, 1-70 & 0.03 : Te-F, 1.84 & 0.03.These values for the interatomic distances agree very well withthose expected for an ionic structure.The sums of the ionic radiiare 1.65, 1.78, and 1-92 A.U. respectively, and these sums areexpected to be somewhat larger than the equilibrium distances1' T. G. Pearson, Nature, 1933,131, 166; A., 238.lS J . , 1932, 2559.20 " Parachor and Valency," London, 1930, p. 136.21 " The Covalent Link in Chemistry," 1933, p. 94.z2 Proc. Nat. Acad. Sci., 1933, 19, 6s; A., 341.Is J . , 1933, 142794 INORGANIC CHEMISTRY.between small highly charged cations and large anions. Thedifferences 0.07, 0.08, and 0.08 A.U.are very close to that betweenthe observed Si-F distance 1-68 & 0.02 A.U. in the crystal ofammonium silicofluoride and the ionic radius sum 1-77 B.U. Brock-way and Pauling conclude that the structures are ionic, and thefluoride ions are situated at the corners of an octahedron aboutthe appropriate central atom.Doabt has always existed as to the correct representation ofhypophosphoric acid and its salts. This acid is readily formulatedas H4P206, but as H2P03 it is an anomalous compound of phos-phorus, with the possible structure OtP(OH),. There is con-siderable evidence in favour of the double formula, but the molecularweights in solution for the esters, determined by A. Rosenheimand M. Pritze in 190~3,~~ indicated the constitution R,P03.Recently,however, A. E. Arbusov and B. A. Arbusov 24 prepared some pureethyl hypophosphate by the reaction of bromine on sodium diethylphosphite, (EtO),P*ONa, and found that it gave a molecular weightin accordance with the formula Et4B206. Moreover, they provedthat the pure ester could not be obtained from the alkyl halide andsilver hypophosphate, the method used by Rosenheim and Pritze,whose molecular-weight determinations are thereby invalidated.Further support for the doublc formula has been obtained frommagnetic susceptibility measurements.I n general, atoms or ions which contain completed sub-groupspossess zero or very small magnetic moments and are diamagnetic,whilst those in which the sub-group is incompletely filled areparamagnetic.On this basis, since H2PO, is an “ odd ” mole-cule and H4P20, is not, it follows that if the hypophosphates havethe simpler structure they should be paramagnetic. 3’. Bell andS. Sugden 25 have therefore determined the magnetic susceptibilitiesof Na2H2P20, (and with 6H20), Ag,P20,, and (CN3H5)4H4P206,2H20,and find them all diamagnetic in accordance with this constitution.They also showed that the products from alkyl halides and silverhypophosphate did not give the reactions of true hypophosphates,confirming the conclusions of Arbusov and Arbusov. Recently,P. Nylen and 0. Stelling 26 concluded that the cryoscopic measure-ments on aqueous solutions of the acid and its sodium salt supportthe formula H,P206. Moreover, X-ray examination of the sodiumand the barium salt shows only one K-absorption edge correspondingwith that expected for the structure R4(03P*P0,).In spite of numerous investigations, the problem of the con-stitution of the perchromic acids remains unsolved.It will be23 Ber., 1908,41,2708. 24 J.pr. Chem., 1931,130,121; A., 1931, 1268.26 2. anorg. Chem., 1933, 212, 160; A., 664. J., 1933, 48; A., 212WARDLAW : STRUCTURES, ETC. 95recalled that earlier 27 workers considered the red perchromates tobe M3Cr08, where chromium is septavalent, whilst the blue saltswere thought to be MCrO,, with sexavalent chromium. The lattersometimes contained a molecule of hydrogen peroxide of crystal-hation. Later, different views were expressed on this matter,and in recent years K. Gleu28 has maintained that the red per-chromates should be formulated as the co-ordination compoundsM3[Cr(02)4] containing quinquevalent chromium, whilst A.Rosen-heim and his collaborators 29 suggest the inset constitution wherethe chromium is sexavalent. R. Schwarz and H.r-(02K)3 Giese 30 have now degraded potassium perchromatein a tensimeter at 0" and shown that no stoicheio-0=Cr=(02K)3 metric relationship exists between the loss of waterand oxygen, thus excluding the formula KCr05,H,0,. Moreover,analyses of the very unstable thallous perchromate indicate theformula TlCrO,, whence the potassium salt should be KCrO,,H,O.Addition of the blue perchromate to excess of acidified potassiumpermanganate containing a trace of ammonium molybdate andtitration of the excess shows the presence of five [O*O] groups inthe molecule thus leading to the constitutionO=F 10 2 1 -[( x>),C.( OK)*O* .I 2Similar treatment of the red perchromate demonstrates that 3+[O*O] groups are present to each chromium atom, which is fully inaccordance with the formula proposed by Rosenheim and hiscollaborators.Nevertheless, B. T. Tjabbes 31 contends that themagnetic susceptibility of K,CrO, indicates a valency of 5 or 7for the chromium, and that the value 5 is supported by electronicconsiderations, isomorphism with the perniobates and pertantalatescontaining quinquevalent niobium and tantalum, and analogy withthe magnetic properties of other compounds of quinquevalentchromium. At present, therefore, there is not general agreementas to the constitution of the perchromates.It has long been known that alkaline oxidation of potassiumhydroxylaminedisulphonate, HO-N(S03K)2, gives a solution similarin colour to permanganate from which, on evaporation, yellowcrystals are deposited with the empirical formula ON(SO,I<),.When these crystals are dissolved in water they yield the deeppurple solution again.The relationship between the two colouredforms of this compound has always been obscure, but recently27 Ephraim, " Inorganic Chemistry," 1926, p. 419.28 2. anorg. Chem., 1931, 204, 67.29 Ibid., 1932, 209, 175.81 2. anorg. Chem., 1933, 210, 385; A., 449.30 Bw., 1933, 66, [B], 310; A., 24196 INORGANIC CHEMISTRY.R. W. Asmussen 33 has shown that the yellow solid has only a feebleparamagnetism whilst the aqueous solution gives nearly the theor-etical paramagnetism for an " odd " molecule.These results aresuitably interpreted as indicating the equilibrium, solid [ON(SO,K),],2(KSO,),NO, where (KSO,),NO is an " odd " molecule accord-ing to the electronic theory of valency. It is interesting to learnthat the X-ray diagram of the supposed mercurous oxide Hg,0,33prepared by different methods, shows only interferences character-istic of mercuric oxide with some diffuse blackening due to mercury.The investigators remark that no satisfactory evidence of theexistence of Hg,O has yet been produced. In the previous Reportattention was directed to the announcement of the isolation ofcompounds of krypton with chlorine and bromine.34 It is nowproved that the red compound stated to be krypton chloride con-tains no krypton,35 and other investigators 36 record negativeresults in attempts to combine the rare gases with the halogens.Nevertheless, L.Pauling 37 predicts the isolation, in due course,of KrF6 and XeF',, with XeF, as an unstable compound. Finally,it may be mentioned that L. Pauling and L. 0. Brockway38 con-clude from electron diffraction experiments with carbon suboxide,that the molecule is linear. This is in accordance with the usualformulation. The interatomic distances are G O = 1-20 0.02,C-C = 1-30 & 0.02 B.U. These values are intermediate betweenthose expected for triple and double bonds, and the normal stateof the molecule may be represented conveniently by the formula0"C"C"C" ........ .... .... 0. w. w.4. CO-ORDINATION COMPOUNDS.Modern physics has already provided a secure foundation for theprinciples which Werner fist outlined in his co-ordination theory,and the outstanding feature of present-day researches on thistheory is the increasing utilisation of physical methods.L. Paulingl has recenfly made some striking observations on32 2. anorg. Chem., 1933, 212, 317.33 R. Fricke and P. Ackermann, ibid., 211, 233 ; A., 579.34 A. von Antropoff, H. Franz, and H. Westerhoff, Naturwiss., 1932, 20,688; A , , 1932, 566.35 A. von Antropoff, H. Frauenhof, and K. H. Kruger, ibid., 1933, 21, 315;A., 579.36 H. Kliding and N. Riehl, ibid., p. 479; A., 776; D.M. Yost and A. L.Kaye, J . Amer. Chem. Xoc., 1933,55,3390; A., 1128; 0. Ruff and W. Menzel,2. anorg. Chem., 1933, 213, 206; A., 914.37 J . Arner. Chem. SOG., 1933, 55, 1895; A., 664.38 Proc. Nut. Acad. Sci., 1933, 19, 800.1 J . Amer. C h m . Xoc., 1033,55, 1895; A., 664WARDLAW : CO-ORDINATION COMPOUNDS. 97co-ordination compounds in which the central atom is exhibitingits maximum valency and is associated with the units oxygen,hydroxyl, or water. These units are attracted by the strongcentral field of a small cation, and the configuration of maximumstability is reached when they are as close to it as possible. Ifthe ions are considered as rigid spheres, then four anions at thecorners of a tetrahedron are in contact with a cation at the centrewhen the ratio p of cation radius to anion radius is 0.225.If theratio be smaller, three anions in contact with the cation preventa fourth from approaching to within the same distance of thecation, and the arrangement with co-ordination number 4 becomesless stable than with co-ordination number 3 (triangular arrange-ment). Pauling deduces that the tetrahedron is stable only forp = 0.225 to 0.414; the octahedron (co-ordination number 6) forp = 0.414 to 0.713, and the cube (co-ordination number 8) wherep is above this. The ionic radii used in these considerations arethe " univalent radii " which represent the relative extensions inspace of the outer electron shells of the ions.Ratios of Univalent Radii : Cation/Oxygen.N5-t R03 Region."+0.14 0.25 I 0.20 0.17Be++ B3+si4+ p5+ S6 + c17+I~ g + + ~ 1 3 +0.48 0.41 0.37 0.34 0.31 0.28I R04 Region.Zn++ Ga3+ Ge4 -I- 1 As5+ S06+ Br7+ Krs+0.50 0.46 0.43 0.40 0.37 0.35 0.330.42Cd++ In3+ Sn4+ Sb5+ TOG+ 17+0.65 0.59 0.55 0.5 1 0.47 0.44RO, Region.Reference to the above table indicates that the values for boron,carbon, and nitrogen are less than 0-225 in accordance with the- existence of H,BO,, H2C0,, and HNO,.The value0 for boron is not far below that required for a co-'B' ordination number 4, so that possibly some meta-which are usually written as derivatives of HBO,are in reality salts of the polymerised acid (HBO,),,in which each boron atom is surrounded by threeoxygen atoms as in H,BO, (see inset).The X-rayIn this compound each boron is0- (I, borates may be salts of HB(OH),. Metaborates\B/\0study3 of Ca(BO,), shows this.2 L. Pauling. J . Amer. Chem. SOC., 1927, 49, 765; A., 1927, 399.3 W. H. Zachariasen, Proc. Nut. Acad. Sci., 1931, 17, 617; A., 1932, 114;W. H. Zaohariasen and. G . E. Ziegler, 2. Krist., 1932, 83, 354; A., 1933, 13.REP.-VOL. XXX. I98 INORGANIC CHEMISTRY.surrounded by three oxygen atoms, in a plane. The existenceof the tetra,hedral hydrate [Be(H,O),]++ in Be(H,O),SO, confirmsthe co-ordination number demanded by the value for beryllium.Salts of acids with formuls usually written as €P2Si0, and HP03are actually polymerised in such a way as to retain the tetra-hedral arrangement of four oxygen atoms round the silicon andphosphorus.The prediction that aluminium should occur withco-ordination number 4 as well as 6 has been verified by theexamination of many crystals including mica, sodalite, etc. Theradius ratio for I7+ is not far above the tetrahedron-octahedrontransition value. Salts of periodic acid, such as KIO,, containinga tetrahedral 10,’ ion, are known, but confirmation is still lackingof an octahedral configuration for the acid and salts representedby the formula? HJIO,], Ag,[IO,], K,H,[IO,], eke. Telluric acid isgenerally recognised as H,TeO, or Te( OH), rather than H,TeO,,ZH,O,and X-ray data show that the six oxygen atoms in the moleculeare equivalently related to the tellurium atom. A representativesalt is Ag,TeO,. Tin shows its expected co-ordination number of 6in CaSnO,.This crystallises with the perovskite structure, eachtin being surrounded by six oxygen atoms at the corners of aregular octahedron, and each oxygen atom being common to twooctahedra in such a way as to make the entire crystal one giantmolecule of the formula (CaSnO,), . Lastly, Pauling examinesthe case of the antimonates, and holds that they are satisfactorilyformulated by assigning to antimonic acid the structure H[Sb( OH),].That sodium antimonate can be written NaSb(OH), had previouslybeen suggested by L. P. Bammett.6Important results have been obtained during recent years on theco-ordinated compounds of the alkali metals. N. V. Sidgwick andS. G. P. Plant in 1925 published an account o€ the first recogaisedco-ordinated compounds of these metals when they described thelithium, sodium, and potassium derivatives of +-indoxylspirocycZo-pentane of the general formula C,,H1,ON*~~~C1,H1,ON.All thesecompounds were decomposed immediately by water, regenerating theindoxyl compound. The sodium and potassium derivatives, however,could be crystallised from toluene although the lithium compounddecornpased. Later, N. V. Sidgwick and F. &f. Brewer 8 preparedfurther co-ordinated compounds of the alkali metals of which twotypical examples are sodium o-nitrophenol salicylaldehyde (I) andlithium salicylaldehyde dihydrate (11). The latter, although a* C. A. Bcevors and H. Lipson, %. Krist., 1932, 82,297 ; A., 1932, 681.L. Pauling, J. Amer.Chem. SOC., 1929, 51, 1010; A., 1929, 748.See ibid., 1933, 55, 3063; A., 919.J., 1925,127, 209; A., 1925, i, 298. * Ibid., p. 2379; A., 1926, 71WARDLAW : CO-ORDINATION COMPOUNDS. 99dihydrate, gives a colourless solution in t>oluene, thereby affordinga clear proof that it is not an ionised but a covalent compound.0 (1.1 (11.1I n 1931, H. King and G. V. R~tterford,~ employing p-alanine,NH,*CH,-CH,*CO,H, as the organic component, isolated derivativesof the type LiX*C,H,0,M,l~5H2O (X = C1, Br, or I), whilst in 19320. L. Brady and W. H. Bodger lo recorded chemical evidence ofchelation in the sodium salts of o-hydroxybenzaldehyde. Morerecently, 0. L. Brady and M. D. Porter l1 have described somestable quadricovalent orange-coloured compounds of the alkalimetals, prepared from react ions involving 4- isonit ro so- 1 - phenyl- 3 -methyl-5-pyrazolone. Particular value, however, attaches to thework of N.V. Sidgwick and F. M. Brewer l2 on the derivatives withsalicylaldehyde, for all the alkali metals will co-ordinate with thisunit to produce the simple structure (111).HH HI n certain cases a further molecule of the aldehyde can be co-ordinated. The experimental results have established the followingcovalency numbers for the alkali metals : lithium and sodium, 4 ;potassium, rubidium, and cssium, 4 and 6. I n this series, the4-covalent compounds of sodium and potassium are much morestable than the others, whilst the potassium derivative with acovalency of six is formed only with difficulty.Sodium may,however, under appropriate conditions, also exhibit a covalency ofsix. The isolation of the compound NaQ,2HS (where HQ = 8quinizarin, and HS = salicylaldehyde) proves this.13Considerable interest attaches to recent work on the co-ordinationcompounds of the currency metals. By the classical methods ofstereochemistry, copper, like arsenic, has been shown to be tetra-hedral when its co-ordination number is 4 and octahedral when itis 6. The resolution of the copper derivative of benzoylpyruvicJ . , 1931, 3131; A., 1932, 150. lo J., 1932, 952; A., 1932, 513.l1 J., 1933, 840.12 J., 1925, 127, 2379; A., 1926, 71; J . , 1931, 361; A., 1931, 443.1s N. V. Sidgwick and F. M. Brewer, J . , 1925, 127, 2379; A., 1926, 71100 INORGANIC CHEMISTRY.acid by W. H.Mills and R. A. Gottsl* in 1926 established thatcopper may have a tetrahedral structure in its 4-covalent com-pounds, whilst the resolution of the co-ordination compound[Cu(en),2H,0]X2 by W. Wahl15 in 1927 demonstrated the octa-hedral codiguration of this 6-covalent compound. This latterresolution was achieved through the tartrate, and the active frac-tions could be converted into active iodides. N. V. Sidgwick16has recently emphasised the surprising nature of these experimentalresults. He points out that copper is generally unwilling t o form6-co-ordinated compounds and readily stops a t 4. However, thiscomplex ion must be remarkably stable, for if the water were evenpartly split off in solution the compound would racemise.Heconcludes that, although this stability is unexpected, there is nodoubt that the attachment of chelate groups to an atom may eitherincrease or diminish its power of further co-ordination.The value of co-ordination in stabilising unstable simple salts iswell known. This is particularly well illustrated by some recentresults obtained by co-ordinating ethylenethiourea (etu ; IV) withcuprous, argentous, and aurous salts. G. T. Morgan and F. H.Burstall 17 have shown that cuprous nitrate, which appears incapableof existence as a simple salt, forms the stable [Cu 4etu]N03, whilstunstable cuprous sulphate becomes quite stable as [Cu 3etu],S04.Moreover, the co-ordination compounds of silver chloride and silverbromide, [Ag 3etuIC1 and [Ag 2etu]Br, are not affected by light.Further, aurous chloride, which is readily decomposed by hotwater t o auric chloride and gold, gives the extremely stable[Au 2etu]C1,H20.The co-ordination compound [Au 2etu]N03,where the central atom has an effective atomic number of only 82,is a very stable complex salt which crystallises unchanged fromboiling solutions and is not reduced to metallic gold by form-aldehyde.It is generally stated that when copper i s exposed to the atmo-sphere it becomes coated with a green basic carbonate. W. H. J.Vernon and L. Whitby l8 find, however, from analyses of typicalpatinas, that the older specimens are identical in composition withthe mineral brochantite which may be formulated as (V). Nearthe coast, the sulphate ion may be replaced, in varying degree, bythe chlorine ion, giving the analogue of the mineral atacamite.The preparation by 0.L. Brady and E. D. Hughes l9 of co-ordin-l4 J . , 1926, 3121; A., 1927, 149.15 Acta Sci. Pennicae, Comm. Phys. Math., 1927, 4, 1; A., 1928, 395.l6 “ The Covalent Link in Chemistry,” 1933, p. 202.17 J . , 1928, 143; A., 1928, 278.1* J . Inst. Metals, 1930, 44, 389; B., 1930, 992. J., 1933, 1227WARDLAW : CO-ORDINATION COMPOUNDS. 101ation compounds such as (VI) involving 2 : 2’-diphenol as a chelategroup adds interesting examples to the limited number of caseswhere a chelate ring may contain more than six atoms.T. S. Moore and (Miss) M. W. Young 2o have directed their atten-tion to the problem of the relative stability of copper derivativesof p-diketo-compounds. They find that the co-ordinating affinitydecreases in the order (1) dibenzoylmethane, (2) acetylacetone,(3) benzoylacetone, (4) acetonedicarboxylic ester, (5) acetoaceticester, and that there is a large gap between (3) and (4).This isalmost the order representing the percentage of enol form in thepure substance 21 and affords strong confirmation of the view thatthe enol forms of p-diketo-compounds are chelated.22Co-ordination with suitable units enables many metals to displayvalencies not shown in their stable simple salts. Silver is mostinteresting in this respect. It is stated that silver and fluorinemay form an unstable black material which contains ca. 75%AgP,.Z3 Many stable co-ordination compounds of bivalent silverhave, however, been isolated in recent years.As far back as1912, G. A. Barbieri prepared the compound [Ag4py]S2O, bymixing aqueous solutions of potassium persulphate and silvernitrate containing excess of pyridine. In 1927 he obtained thenitrate [Ag 4py](NO,), by anodic oxidation of silver nitrate in thepresence of pyridine, and in the following year W. Hieber andF. Miihlbauer, substituting o-phenanthroline (phenan) for pyridine,obtained a series of argentic salts of the general formula[Ag 2phenan]S,O, or X,, where X = HSO,, NO,, ClO,, or C10,.Later G. T. Morgan and P. H. Burstall stabilised the bivalent silverion in a number of compounds by employing 2 : 2’-dipyridyl as achelate group,24 and recently G.A. Barbieri 25 has applied theelectrolytic method to isolate similar compounds containing di-pyridyl. He has also prepared a new type of complex compound2O J . , 1932, 2694; A., 1933, 36.21 K. H. Meyer, Annalen, 1911, 380,242.22 N. V. Sidgwick, “ Electronic Theory of Valency,” p. 147.23 35. S. Ebert, E. L. Rodowskas, and J. C. W. Frazer, J . Amer. Chem. SOC.,24 Ann. Reports, 1931, 28, 132.25 Atti R. Accad. Lincei, 1932, [vi], 16, 44; A., 1933, 34.1033, 55, 3056; A., 916102 INORGANIC CHEMISTRY.of bivalent silver by using picolinic acid as the associating unit.26Silver picoliiiate, on oxidation with either potassium persulphateor anodic oxygen, gives an orange-red crystalline compound of thecomposition Ag( C,H,NCO,),, which with hydrochloric acid givessilver chloride and chlorine.Presumably this is the internallycomplex compound (VII).(VlI.)The possibility that in some of these compounds the bivalencyof the complex is not due to a bivalent silver ion but to incom-plete co-ordination of the nitrogen atoms, leaving one of them freeto function as a base, has been examined from the point of viewof magnetic susceptibilities. In the univalent condition the ionsof copper and silver possess a complete sub-group of 10 electronsand hence should be diamagnetic. On the other hand, the bivalentions have only 9 electrons in this group and therefore should beparamagnetic. S. Sugden2' has described the results he hasobtained from measurements of the magnetic susceptibilities of 10compounds of copper and 9 compounds of silver.From these,and other data in the literature, it is found that both copper andsilver in the univalent condition have zero magnetic moments,whilst the bivalent atoms of these elements exhibit moments of1.72-2-16 Bohr units. Amongst the compounds examined werethe salts [Ag 4py]S,O,, [Ag Zdipy]S,O,, [Ag 3dipy](C104),. Inevery case these salts were paramagnetic, with moments of thesame order of magnitude as those of bivalent copper. Theseresults provide st'rong evidence for the view that the bivalency ofthe complex is due to the presence of a bivalent silver ion.W. Klemm 2* has also published data for the magnetic susceptibilityof [Ag $py]S,O, and [Ag Z(phenan)]S,O, which support this view.In 1907 J.Meyer and H. EggelingZ9 reported that colourless andyellow sodium silver thiosulphates existed. They were supposedto be isomerides, silver being attached to oxygen in the one caseand to sulphur in the other. H. Bassett and J. T. Lemon30 con-clude from their experiments that this idea is erroneous, the colourdifference being merely due to decomposition. They mention,however, that colourless and yellow sodium cuprous thiosulphatesundoubtedly exist but that they are not isomeric.Much fundamental work on the chemistry of tervalent gold hasZ 6 Atti R. Accad. Lincei, 1933, [vi], 17, 1078.3 7 J . , 1932,161 ; A., 1932,324. 28 2. anorg. Cliem., 1931,201,32 ; A , , 1932,lO.29 Ber., 1907, 40, 1361 ; A., 1907, ii, 347. 30 J., 1933, 1434WARDLAW : CO-ORDINATION COMPOUNDS.103been done by C. S. Gibson and his collaborators. Tervalent goldis commonly 4-covalent, and this is substantiated by the observationof Gibson and J. L. Simonsen31 that diethylgold bromide is bi-molecular in benzene (VIII) and that with ethylenediamine orthallous acetylacetone it yields the complex salts (IX) and (X).Et\&Brk Au/Et [ I >AuEtq Br \Au/OoCMe >CHEt ’ ‘Br’ ‘Et CH,*NH, Et’ ‘O-CMe(VIII.) (IX.) (X. 1In 1931 Gibson and W. M. Colles 32 isolated the following interestingexamples of 4-covalent gold, [Au(en),]Br,, [Au 2py Br,]Br, and[Au py Br,], and quite recently Gibson 33 has announced the prepar-ation of two auric derivatives of unusual type, (XI) and (XII).CH,*NH, EtvH2 ”;JH, (XII.)CH2-CH,The second salt contains G-covalent gold and is the first gold com-pound to be described in which the possibility of optical activityarises.Another group of co-ordination compounds which has beeninvestigated in recent years is that of thallium.Here, R. C. Menzies,N. V. Sidgwick, and their collaborators have made noteworthycontributions. It is remarkable that thallium forms no monoalkylderivative. The trialkyl derivative, however, has been preparedby H. P. A. Groll in good yield from diethylthallium chloride andeth~l-lithium,3~ but one of the alkyl groups is very reactive, and thesubstance tends to form the stable dialkyl compounds which aresalts of the strong base Alk,Tl*OH. It has been found that Et,TlXcan be conveniently prepared by the action of ethylmagnesiumbromide on thallous chloride or thallous e t h o ~ i d e .~ ~ Amongst thedifferent investigations, one of particular interest is directed todetermine whether the derivatives of the dimethyl and the diethylbase Alk,Tl*OH with P-diketones, acetoacetic ester, and salicyl-aldehyde are chelate compounds or not.36 The presence of the two3 1 J., 1930, 2531; A., 1931, 75. 32 J., 1931, 2407; A., 1931, 1316.33 Nature, 1933, 131, 130; A., 267.34 J . Amer. Chern. SOC., 1930, 52, 2998; A., 1930, 1302.35 R. C. Menzies and I. S. Cope, J . , 1932, 2862; A., 1933, 152.36 R. C. Menzies, N. V. Sidgwick, E. F. Cutcliffe, and J. M. C. Fox, J., 1928,1288; A., 1925, 745104 INORGANIC CHEMISTRY.alkyl groups would give the thallium in a monochelate derivativea stable covalency of four.The products, which are of the typeAlk,TlA (where A is a radical of the diketone or similar substance),prove to have unusual properties. They are crystalline solidswhich can he sublimed with ease under reduced pressure and arereadily soluble in benzene, indicating that they are chelate com-pounds of the types (XIII) and (XIV). On the other hand, withthe exception of the benzoylacetone compounds, they are extremelysoluble in water, yielding solutions with a strong alkaline reaction.Evidently, in water, these derivatives, like hydrogen chloride, passfrom the covalent to the ionised state. The alkyloxides andaryloxides of thallium, such as TlOC,H, and TlOC,H,, behavesimilarly.37 I n water they are salts, but they dissolve in benzeneand are presumably covalent in that solvent and in the solid state.These substances, however, do not display thallium with a co-valency of one, for molecular-weight determinations in benzeneshow that they exist as four-fold polymerides, The structure ofthis complex is not yet settled, although some ingenious speculationsare made as to its configuration.The formation of the stable four-molecular complex shows that unicovalent thallium is very unstable,and this may be correlated with the fact that as yet no monoalkylderivative of thallium has been isolated. On the other hand, thegreat stability of the dialkylthallium radical makes it unlikelythat here the valency group of the thallium is completed by co-ordination.It is interesting to find, therefore, that an examinationof the crystal structure of dimethylthallium iodide, TMe,I, hasdisclosed that the C-T1-C atoms lie in a straight line.3s Thisevidence suggests, according to N. V. Sidg~ick,3~ that for 2-covalentcompounds of atoms with “ incomplete octets,” the structure isrectilinear.Dithionic acid is generally considered to have the constitutionHO,S*SO,H, the two sulphur atoms being directly linked. Con-firmation of this view is afforded from the X-ray examination ofpotassium d i t h i ~ n a t e . ~ ~ Each sulphur in the ion [S,O,]” is sur-rounded, approximately tetrahedrally, by three oxygen atoms andone sulphur atom. This accords with the formulation (XV). An3 i N. V. Sidp-ick and L.E. Sutton, J., 1930, 1461; A., 1930, 1052,58 H.M.Powelland(Miss)D.M. Crowfoot,Nature,1932,130,131; A., 1932,904.3* Op. cit., ref. (16).40 G. V. Hclwig, Z. Krist., 1932, 83, 486; A., 1933, 13WAIXDLAW CO-ORDINATION COMPOUNDS. 105X-ray examination of sodium metabisulphite, Na2S20s,41 has dis-closed that the structure of the S20, group is the same as that ofthe S20, unit, with one oxygen atom removed. Presumably itshould be written as (XVI),T. M. Lowry and F. L. Gilbert 42 assumed the existence of single-electron links to the halogens in the co-ordination compoundsa-l\le,TeX,, where X = C1, Br, or I, whereby an inert-gas structureis attained by the tellurium. These substances, however, provedto be diamagnetic, so they concluded that all the electrons weremagnetically paired just as they are in compounds in which thevalency electrons are present in pairs of shared electrons or as“lone pairs ” of unshared electrons.If these compounds docontain single-electron bonds to the halogens there should be ELparamagnetic effect due to the unpaired electrons. Actually, thiswill be masked by the diamagnetic effect of the rest of the molecule.However, if there is any paramagnetic effect it will be shown bythe variation of magnetic susceptibility with temperature (sincediamagnetism is independent of temperature). S. S. Bhatnagarand T. K. Lahiri43 have examined the compounds a-Me,TeR,(R = Br, I, NO,) and find the susceptibility t o be constant up t oand beyond the melting point. Further, assuming all the bondsto be electron pairs, they have calculated the susceptibilities andfind that they all agree with the observed values.They conclude&herefore that there is no evidence for the single-electron bond inthese substances.The structure of the heteropoly-acids which are typical co-ordination compounds has long been a subject of speculation, anduntil recently chemists were content to adopt the Miolati-Rosenheimconceptions based on Werner’s theory and to assign to the important12-phosphomolybdic and -phosphotungstic acids the constitutionsH7[P(R,07),],aq., where R = Mo or W. I n 1929, L. Pauling 44on theoretical grounds gave the formula of the phosphotungsticacid as H,[P04W1201,( OH),,]. X-Ray investigations have nowshown that the formula is best written as H,P(W,010)4,nH20 andthat the other 12-heteropoly-acids have this structure.45 The4 1 W.H. Zachariasen, Physical Rev., 1932, 40, 923; A., 1932, 903.42 Nature, 1929,123, 85; A., 1929, 127.43 2. Physik, 1933, $4, 671; A., 1002.44 J . Amer. Chem. Xoc., 1929,51,2888; A., 1929, 1367.45 J. L. Hoard, 2. Krist., 1933, 84, 217; A., 215; J. F. Keggin, Nature,1933,131,908; A., 768. D 106 INORGANIC CHEMISTRI?.arrangement differs from that of Pauling. The complex acidicanion is a co-ordinated structure, more or less spherical. At thecentre is one phosphorus, silicon, or boron atom, surrounding whichare 4 oxygen atoms in the form of a tetrahedron. Each of these iscommon to three oxygen oct'ahedra, making 12 octahedra in all.At the centre of each octahedron is a molybdenum or tungsten atom,i e ., 12 altogether. I n each group of three octahedra there is onecommon corner (one of the four central oxygen atoms) and threecommon edges, so that the whole complex has tetrahedral symmetry.The water of crystallisation packs in the interstices between theanions [PW,,0,,]3.Finally, some observations on the transition elements may berecorded. Here, covalencies of 8, 6, and 4 are possible. Atpresent, little is known of the valency distribut,ion with &covalentatoms, but without exception 6-covalent atoms have an octahedralarrangement. Amongst the transition elements, optically active6-covalent compounds of iron, cobalt, rhodium, platinum, iridium,and ruthenium have been known for some time, and recentlyG.T. Morgan and F. H. Burstall,46 by resolving trisdipyridyl-nickelous chloride, [Ni 3dipy]C1,,6H20, have added nickel to the list.Molecular dissymmetry amongst purely inorganic compounds isvery rare. Unusual interest attaches therefore t o the preparationby F. G. Mann 47 of the optically active 6-covalent rhodium deriv-ative Na[ (2H,O)Rh( S0,N2H,),].I n the 4-covalent compounds, the unique behaviour of certainof the transition elements is important. Whilst, in general, atomswith a covalency of 4 show a tetrahedral distribution of valencies,there are three exceptions to the rule, vix., nickel, palladium, andplatinum. As far back as 1893 Werner proposed a planar con-figuration in the case of platinum, and palladium was added later.Not until 1931, however, did L.P a ~ l i n g , ~ ~ from considerations basedon wave mechanics, deduce that nickel, like palladium and platinum,should have a planar distribution of valencies in its 4-covalentcompounds. Moreover, he asserted that the electronic formationof this plane configuration would involve a definite decrease of the(para) magnetic moment, so that in nickel it would become zero.The isolation by S. Sugden 49 of two isomeric forms of nickel benzyl-methylglyoxime, coupled with the fact that, like the nickel deriv-atives of dimethyl- and diphenyl-glyo~irne,~~ they are diamagnetic,46 J . , 1931, 2213; A., 1931, 1168.48 J. Amer. Chem. SOC., 1931, 53, 1367; A,, 1931, 670.49 J., 1932, 246; A., 1932, 272.50 W.Klemm, H. Jacobi, and W. Tilk, 2. anorg. Chem., 1931, 201, 1 ; A.,*' J . , 1933, 412; A., 580.2932, 10; L. Cambi and L. Szego, Ber., 1931, 64, 2591 ; A., 1932, 10WARDLAW CO-ORDINATION COMPOUNDS. 107is strong confirmation of Pauling’s views. On the other hand,the chemical evidence for the plane structure of 4-covalent palladiumcompounds has proved to be valueless. Pink and yellow forms ofthe 4-covalent compounds (NH,),PdCl,, (~y)~PdCl,, (EtNH,),PdCl,have long been known, and in recent years, by analogy with theplatinum derivatives, they have been considered as cis- and truns-isomerides of planar configuration. This view was supported in1927 by the results of F. Krauss and F. B r ~ d k o r b , ~ ~ who foundthat the compounds gave simple molecular weights in freezingphenol.In 1931, however, F. Krauss and K. Mahlmann 52 repudi-ated these results, explaining that the nature of the solvent madethem unreliable. They suggested instead that all the compoundswere bimolecular. A re-examination of the substances by otherinvestigators showed that, whilst the yellow forms had the simpleformula, the pink were bimolecular and actually salts of the type[Pd(NH3)4][PdC14].53 This was proved by a variety of reactionsand has been confirmed since by further work on the pallado-~ulphines.53~ Whilst, therefore, there is only one compound(py),PdC12, there is no doubt that there are two monomeric isomerides(~y)~PtC1,,5~ and it is not unreasonable to suppose that theserepresent the cis- and trans-forms of a planar structure. Thus,in the case of platinum chemical evidence favours the planar con-figuration. Additional chemical evidence has been advanced byTschernjaev in the isolation of three compoundsand by A.A. Griinberg and B. W. P t i ~ y i i , ~ ~ who consider that theyhave prepared two compounds of the formulz (XVII) and (XVIII).[Pt (NH3) (NH2OH) (PY )~O,I,[PtC1,1~ 55Ho(XVII . )OC--0, ,N‘CH, 1 pt I (XVIII.)H ~ C - N ~ ‘o-coH2Lastly, H. I>. K. Drew and F. S. H. Head 57 have found that byacting with isobutylenediamine upon P-[(NH,)(NH,Et)PtCI,] amixture of two isomeric tetrammines results, which they have51 2. anorg. Chem., 1927, 165, 73; A., 1927, 951.52 Siebert Festschr., 1931, 215; A., 1932, 240.53 H.D. K. Drew, F. W. Pinkard, G. H. Preston, andW. Wardlaw, J., 1932,53a H. D. K. Drew, G. H. Preston, W. Wardlaw, and G. H. Wyatt, J., 1933,54 A. Hantzsch, Ber., 1926, 59, 2761 ; A., 1927, 93.2 5 Ann. Inst. Platine, 1928, 6, 55; A., 1928, 974.56 J . pr. Chem., 1933, [ii], 136, 143; A., 381.5 7 Nature, 1933, 132, 210; A., 1040.1895; A., 1932, 824.1294108 INORGANIC CHEMISTRY.separated and characterised by their different plato-salts. Pre-sumably these are the dichlorides of the planar structures (XIX)and (XX). A similar pair of isomeric p-platotetrammines[Pt (NH,*CH,*CMe,*NH,),]Cl, has also been obtained in admixtureand separated as two different plato-salts. All these examplesseem t o afford definite evidence of a planar structure for $-covalentplatinum.Valuable information on this matter has been obtained duringrecent years from the X-ray examination of co-ordination com-pounds of palladium and platinum.It will be recalled that R. G.Dickinson 58 in 1922 examined potassium chloro-palladite and-platinite by X-ray methods, and concluded that the four chlorinesin the complex anion lay in a plane with the palladium or platinumatom. In 1932 59 E. G. Cox showed by an X-ray investigation of[Pt(NH3),]C12,H20 that the four nitrogen atoms lay in a planewith the platinum atom, whilst an examination of Magnus’s salt[Pt(NH3)4] [PtCl,] disclosed a similar distribution.60 Quite recently,E. G. Cox and G. H. Preston have carried out X-ray examinationsof [Pt(en),]Cl,, [Pd(en),]CI,, and [Pd(NH,)&Cl,,H,O ; all these com-pounds gave further evidence of the planar distribution of palladousand platinous valencies. In addition, the results have led to theconclusion that the valencies of the metal atom are not all crystallo-graphically equivalent but are differentiated into pairs.More pre-cisely, only those valencies which are trams to each other areequivalent. This applies to all the above compounds and also to[Pt(NH,),]C12. Whether this pairing is confined t o the amminesis a matter for further investigation. They suggest that theirwork affords a physical basis for Werner’s hypothesis of trans-elimination of groups from the tetrammines and triammines whichhas been proved to explain satisfactorily the reactions of a largenumber of mixed tetrammines and triarnmines, recently inves-tigat8ed.62 Whilst it is certain from the X-ray results that the fourcovslencies are not crystallographically equivalent, it is possible5 8 J.Amer. Chem. Soc., 1922, 44, 774, 2404.6e J., 1932, 1912; A., 1932, 797.6o E. G. Cox, F. W. Pinkard, W. Wardlaw, and G. H. Preston, J., 1932,2527; A , , 1933, 41.61 J., 1933, 1089; A., 1040.62 F. W. Pinkard, H. Saenger, and W. Wardlaw, J., 1933, 1056; A,, 1001WARDLAW : CO-ORDINATION COMPOUNDS. 109that this may be explained as being due to polarisation forces inthe crystal. However, if this were the explanation, it would beexpected that the effect would be found in many other substancessuch as potassium chloroplatinite, particularly as the negativeion [PtCl,]” is more polarisable than the positive ion Pt(NH,),”.Experimental evidence shows this is not the case.61 It is inkerestingto find that the a-tetrachloride derived from the chlorination ofCGP~(NH,),CI, proves, on X-ray examination, to be suitably form-ulated with the ammonia groups in trans-positions in accordancewith Werner’s view of its configuration (XXI).Whilst the resultsfor the p-tetrachloride were not so definite, they could be mostsimply explained by an octahedral distribution of valencies with theammonia groups in cis-positions.c1From time to time resolutions of 4-covalent nickel, palladium,and platinum compounds have been described,63 but in no casehas the investigator replaced the active radical used by an inactiveone whilst still retaining the optical activity.It still remains,therefore, an interesting problem whether these three elementscan give configurations other than planar when they are 4-covalent.I n the case of the pp’p”-triaminotriethylamino (tren) platinoussalt F4 a planar configuration is unsatisfactory for the molecule.Here again the position is uncertain since it is not known whetherthe compound is unimolecular or not. An X-ray examination ofNi(i~en)SO,,~* which is known to be monomeric, would be veryuseful. F. Rosenblatt and A. Schleede 65 have suggested that theresolutions published by H. Reihlen and W. Huhn 63 do not meanthat the valencies of the platinum and palladium are non-planar.They consider that such results can be explained by asymmetryof the complex as a whole, the valencies remaining planar with thepalladium and platinum atom.This may be true in the case ofunsymmetrical chelate groups, but it seems most unlikely thatthe various unsymmetrical models proposed by them for com-pounds of the type Pt(en),Cl, have any real existence. The ion63 H. Reihlen and W. Hiihn, Annalen, 1931,489,42; A., 1931, 1167; 1932,499, 144; A., 1933, 74; A. Rosenheim and L. Gerb, 2;. anorg. Chem., 1933,210,289 ; A., 497.84 F. G. Maim and (Sir) W. J. Pope, J . , 1926, 482; A., 1926, 1233.6s Annalen, 1933, 505, 51110 INORGANIC CHEMISTRY.Pt(en)," has in fact been shown to possess a symmetrical struc-ture 61 different from any of those proposed by F. Rosenblatt andA.Schleede. These investigators 66 have recently questioned thecxistence of the third isomeride of Pt(NH,),CI, basing their con-clusions entirely on the interpretation of their Debye diagrams.In view of the known difficulty of obtaining reproducible results byX-ray powder photographs of highly absorbing compounds of lowsymmetry, it is unfortunate that these authors give no details oftheir experimental technique. Their deduction from the pub-lished diagrams that this third isomeride is a- or trans-Pt(NH,),CI,whose lattice has been modified by the presence of a small quantityof (3- or cis-Pt(NH,),CI, seems unjustified and opposed t o all thechemical evidence. An interesting example of a, 4-covalent atomis osmium in the osmiamates such as K[OsO,N].F. M. Jaegerand J. E. Zanstra6' have recently shown that the ion [OsO,N]'is very nearly tetrahedral. The nitrogen is equivalent to theoxygen atoms as far as X-ray considerations are concerned. Thissuggests that the structure is (XXII) and that the nitrogen carriesthe negative charge. The per-rhenates MeReO, have the samestructure as the osmiamates.In conclusion, it may be mentioned that much valuable inform-ation on co-ordination compounds will be found in " A Survey ofModern Inorganic Chemistry " by G. T. Morgan 69 and in hisrecent publication on " Experimental Researches on Co-ordin-ation." 70 Also the new book by N. V. Sidgwick 71 on "SomePhysical Properties of the Covalent Link " deals with many im-portant aspects of co-ordination compounds.The undersigned gratefully acknowledges the valuable help ofMr.E. G. Cox. w. w.5. THE COVALENCY RULE.The covalency rule1 is the generalisation that the maximumcovalency of an element is limited in accordance with its period inthe Periodic Table, being 2 for hydrogen, 4 for elements of thefirst short period (Li-F), 6 for those of the second short period6 G Ber., 1933, 66, [B], 472; A., 476.6 7 PTOC. K . Akad. Wetensch. Amsterdam, 1932, 35, 610, 787; A., 1933, 13.6 8 F. M. Jaeger and J. Beintema, &id., 1933, 36, 528; A., 892.69 Institute of Chemistry, 1933.'O Roprinted from the Acharyya Sir P. C. RBy Commemoration Volume,71 The George Fisher Baker Non-resident Lectureship in Chemistry at1 N. V. Sidgwick, " Electronic Theory of Valency," 1927, p.152.1932.Cornell University, Cornell University Press, 1933SIDGWICK : THE COVALENCY RULE. 111(Na-Cl) and the first long period (K-Br), and 8 for elements ofhigher atomic number. That in the first short period the limit is4 was suggested by Werner, who pointed out that this might accountfor the chemical inactivity of carbon tetrachloride. The extensionto the rest of the Periodic Table was made in practically its presentform in 1923.2 It is of interest to see how far the validity of therule has been affected by the developments and discoveries of thelast ten years.I n the nature of the case, the evidence in its favour is largelynegative: that no compounds have been found which conflictwith it. This test it has stood satisfactorily.Of the numerouscompounds discovered in the last ten years, whose molecularweights have been determined, there is none in which an atomshows a higher covalency than the rule permits. A few compoundswhich appeared to contradict it have been shown not to exist, orto be compatible with it. For instance, a supposed 8-covalentcobalt compound was found by W. R. Bucknall and W. Wardlawto be really 6-covalent. Again, the highly complex borotungstates,which were written R,[B(W,O,),], were claimed as proving thatboron could have a covalency of six, whereas the rule requiredthat it should be limited to four. It has, however, recently beenfound4 that in these compounds, as in the phosph~tungstates,~the boron atom is surrounded by four oxygen atoms and is there-fore 4-covalent, in the same way as the central beryllium atomin the " basic " beryllium acetate Be,0(CO*CH,),.6On certain points, however, direct evidence has appeared infavour of the rule.The first of these is from the side of theoreticalphysics. The theory of the covalent link elaborated by Heitlerand London seems to suggest that the maximum covalency of anatom is equal to one-half of the number of electrons in its outer-most quantum group, so that the covalency maxima should be4, 9, 16, whereas the chemical evidence indicates that they are4, 6, 8. The theory of molecular orbitals developed by J. E.Lennard-Jones 8 and R. S. Mulliken 13 seems to be leading to aresult more in accordance with the chemical evidence.Their3 N. V. Sidgwick, Trans. Faraday Soc., 1923, 19, 469; Sidgwick and R. K.3 J . , 1928, 2648; A., 1928, 1345.4 J. F. Keggin, unpublished work.5 See this vol., p. 105.6 (Sir) W. H. Bragg and G. T. Morgan, Proc. Roy. SOC., 1923, [ A ] , 104, 437;7 See C. N. Hinshelwooii, Ann. Reports, 1930,27,14-17.8 Trans. Faraday Soc., 1929, 25, 668; Faraday SOC. discussion, Sept. 1933.9 Ph.jeiCal Rev., 1932,40,55 ; 41,49, 751 ; 1933,43,279.Callow, J., 1924, 125, 532; A., 1924, i, 506.A . , 1924, i, 7112 INORGANIC CHEMISTRY.methods have a more geometrical basis than those of Heitler andLondon. They consider the quantum groups and sub-groups ofthe molecule as a whole, and find that these are usually differentfrom those of an atom isolated in space.The method suggeststhat in an octahedral group, for instance, there is an outer quantumsub-group which requires twelve electrons to fill it, and in a moleculeof the type XU, six of these may be contributed by X and one eachby the U’s. Detailed consideration of the electron structure ofdifferent molecular arrangements indicates that the tetrahedralarrangement will be succeeded by an octahedral, and that perhapsby a cubic, which would give the observed covalency values of4, 6, 8.Another point is the covalency maximum o€ 2 for hydrogen.On the chemical side there is no reason to doubt this. It is requiredto account for the association of hydroxylic compounds, the existenceof the ion [F’,H]’, the peculiar non-associated character of o-nitro-and aldehydo-phenols and of the enolic forms of p-diketones, theaction of water on the binary halides, etc.But on the physicalside the position is not so simple. It was originally assumed thata covalency of 2 implied a group of 4 shared electrons. The physicalobjections to this are very strong. We know that an a%om cannothave more than 2 electrons in the first quantum group, and if thehydrogen in these molecules has four, two of these must be in thesecond quantum group. But the energy with which 2-quantumelectrons would be held is too small to account for the stabilityof the co-ordinated hydrogen. It seems more probable, therefore,that we have to deal with some kind of resonance between twostructures, in the first of which the hydrogen is attached to one,and in the second to the other, of the two atoms which it holdstogether :It : H : R z+ R : H : RIf such resonance is possible, it will iiivolve a lower energy statethan either of the two separate structures; that is, the resonancecondition will be stable.The time of change from one structure tothe other (if, indeed, we are allowed to speak of such a change) Millbe excessively short, and can involve an oscillation only of theelectrons and not of the hydrogen nucleus, which is too heavyto oscillate with the necessary frequency,Whatever may be the exact mechanism of the co-ordinatedhydrogen, its occurrence is beyond doubt ; we have good evidencethat a hydrogen atom can hold two other atoms together, as inthe examples quoted above. It has also recently been foundpossible t o determine the dimensions of a hydrogen atom in thiSIDGWICH : THE COVALENCY RULE.113state, by means of X-ray measurements of crystal structure. Thesedo not of course show the position of the hydrogen, but they givethe distance between the nuclei of the two atoms which the hydrogenlinks. Thus W. H. Zachariasen10 has shown that in sodium bi-carbonate, NaHCO,, there is a unique distance between one pair ofoxygen atoms in two neighbouring CO, groups, and this can onlybe the pair which are united through the co-ordinated hydrogenO-HtO. This distance is 2.55 A.U., the other distances betweenoxygens of neighbouring CO, groups being all close to 3.15 B.U.Precisely the same distance (2.55) has been found l1 in potassiumdihydrogen phosphate, KH,PO,.On the reasonable assumption(see below) that the hydrogen valencies lie in the same straight line,the distance between the oxygen nuclei in O-HtO is twice theradius of the oxygen plus the diameter of the co-ordinated hydrogen.The oxygen radius being 0.70 A.U., we get 1.15 A.U. as the diameter,or 0.58 A.U. as the radius, of the co-ordinated hydrogen atom.Nearly the same result is obtained from the structure of the F-HtFion in the “ acid ” fluorides; lo the distance F F is found to be2.36 A.U., and subtracting twice the radius (0-68) of the fluorineatom we get 0-50 as that of the hydrogen. These values aremarkedly larger than the radius of the ordinary l-covalent hydrogenatom, which is 0.37 A.U.; but the influence of the extra electronswill enlarge the radius; that of the negative hydrogen ion H’ is1-3, but this includes an “ envelope ” of perhaps 0.5, which wouldmake the true radius of the ion 0-8 A.U.This conception of the co-ordinated hydrogen is in accordancewith many of its properties.I n its chelate compounds, such asthose of the P-diketones (as well as in the metallic derivatives), ithas long been evident that the unsymmetrical formula? which areusually employed, z‘ix., (I) and (II), do not really represent theR\R” R”facts : that i t is not to be supposed that one of the oxygen atomshas permanently a different relation to the hydrogen (or the metal)from the other, and that if there is any difference between thetwo it must be supposed t o oscillate from one form to the otherlike the double links in the Kekul6 formula for benzene.This hasbeen confirmed by the observation l2 that the electrical dipolelo J . Chem. Physics, 1933, I, 634.11 J. West, Z . KTist., 1930, 74, 306; A., 1931, 289.12 L. E. Sutton, PTOC. Roy. SOC., 1931, 133, 668114 INORGANIC CHEMISTRY.moment of beryllium acetylacetone (1.18 D.) is not more than mightbe expected from the flexibility of the molecule, and far less thanshould result from the presence of two fixed co-ordinate links.Such an alternation is in agreement with the theory that in thehydrogen compounds there is an oscillation between the twostructures, and this would also explain the absence of isomericenols in diketones (or keto-esters) in which R and R' (above) aredifferent.It has been objected that the enols are unlikely to have chelatestructures analogous to those of the metallic derivatives, becausethe great difference in size between a hydrogen atom and that of ametal will introduce too much strain into the ring.This objectioncan be shown to be of small weight. The atoms of the heavierelements are not so much larger than that of the hydrogen as onemight expect, as Table I shows :TABLE I.Radii of neutral atoms (A.U.).-H+ Li Be Mg Z n Pi; Th c0.58 1.50 1.12 1.60 1.32 1.38 1.79 1.49The strain due to the small size of the hydrogen is also relieved inanother way. It has been shown l3 that when a 2-covalent atomhas less than an octet of valency electrons, like thallium in[Tl(CH,),]+ or mercury in Hg(C,H&, the natural angle is not 110"hut 180°, and it is probable that this angle is more easily distortedfrom its natural value. Hence in a chelate 6-ring the effect onthe strain of the small size of the hydrogen atom will be to someextent compensated by the larger angle between its valencies.The recognition of the fact that the valency angle in co-ordinatedhydrogen is 180" removes another difficulty, that of formulatingthe dimeric molecules of the carboxylic acids.Such acids, ofcourse, readily polymerise in non-ionising solvents (some even inthe vapour) up to double molecules, but no further. It shouldbe possible to explain both why the polymerisation occurs, andwhy it stops st this point instead of proceeding indefinitely, asit does with the alcohols and phenols.It is clear that in (111) thehydrogen has a strong tendency to act as an acceptor, and thedoubly linked oxygen as a donor. Co-ordination within the moleculeto (IV) is impossible, because the strain in a 4-ring would be toogreat; if it occurred, acetic acid would behave as a non-associatedsubstance, like salicylic aldehyde, where intramolecular co-ordinationgives a stable 6-ring. I n the acid, co-ordination must thereforetake place between molecules, and if we suppose it to occur twice,l3 See N. V. Sidgwick, '' The Covalent Link," 1933, p. 230SIDGWICK : THE COVALENCY RULE. 115we get the symmetrical dimeric structure (V). This would haveno power of further co-ordination, which would explain why the0 0 O+H-0,-C/ -c/ 4H R-CH /C-R‘O-H ‘0’ \O-HtO(111.) (IV.) (V.)association stops a t this point; it should also have a low dipolemoment, which is found to be the ~ a s e .1 ~ The formula is also inagreement with the crystal structures.15 The one objection thatcould be urged against it was that chelate rings of as manyas 8 atoms are almost unknown.16 Now that we realise that thecovalencies of the hydrogen are in one straight line, it is evidentthat the symmetrical introduction of two hydrogen atoms into a6-ring as in the formula above will not increase the strain.Further evidence in favour of the covalency rule has been affordedby the behaviour of the recently discovered fluorides of nitrogenand oxygen NF, and OF,.It has already been pointed out l7that the behaviour of different binary halides with water can beexplained if we assume that reaction takes place through a compoundformed (not necessarily to more than a minute extent) by co-ordination between the central atom of the halide and the oxygenor the hydrogen of the water; and that this theory gives supportto the covalency rule, and indicates where the covalency maximumchanges. For example, in carbon tetrachloride the carbon, havingalready its maximum valency group fully shared, cannot co-ordinateand does not react, whilst in silicon tetrachloride, where the maxi-mum group is not 8 but 12, co-ordination to Cl,Si+-OH, occurs,and is followed by hydrolysis; in the same way the hexafluorideof selenium (maximum 12) does not react with water, while that oftellurium (maximum 16) does so.On this hypothesis the hydro-lysis of nitrogen trichloride to hypochlorous acid and ammonia isdue to co-ordination between the nitrogen and a hydrogen atomof the water giving ‘l\N/ ; this is followed by the lossof HOCI, which involves the replacement of the chlorine on thenitrogen by hydrogen. Nitrogen fluoride, NF,, was discovered byRuff in 1928.17a If we apply this theory of hydrolysis to it, we can14 C. P. Smyth and H. E. Rogers, J . Amer. Ch,em. SOC., 1930, 52, 1824; A.,1930, 841.15 A. Muller, J., 1923, 123, 2043; A. Miiller and G. Shearer, ibid., p. 3156.16 See “ Electronic Theory of Valency,” p. 252.1 7 Sidgwick, J., 1924,125, 2672 ; “ Electronic Theory of Valency,” p.166.0. Ruff, J. Fischer, and F. Luft, 2. anorg. Chem., 1928, 172, 417; 2.ClCI/ ‘H-O-Hangew. Chem., 1928, 41, 1289; A., 1928, 854116 INORGANIC CHEMISTRY.see that the first stage, the formation of the co-ordination compound,should occur as easily as with nitrogen chloride, but that the secondstage, which here would be the elimination of hypofluorous acid,is impossible, because that acid does not exist. Hence if this theorygives the actual (and only) method by which such halides can behydrolysed, nitrogen fluoride should not be acted on by water.This is found to be the fact; the gas is unaffected by water oralkalis even on heating; it is only when it is sparked with steamor hydrogen that decomposition occurs, with the formation ofhydrogen fluoride and the oxides of nitrogen. Oxygen fluorideis also remarkably inert, though less so than the nitrogencompound ; it reacts with strong alkalis and acids, but is indifferentto water.lg It is also remarkable that if the gas, at a low pressureor mixed with an inert gas, is treated with yellow phosphorus, thelatter removes the oxygen from it, but has no action on the oxygenfluoride.20 Dennis and Rochow have found that with concentratedalkali it gives an oxy-acid of fluorine, but have obtained evidencethat this is not hypofluorous acid but fluoric acid, HFO,.Thus the behaviour of these fluorides supports the view that thereaction of halides with water takes place through an intermediateco-ordination compound ; and if this view is accepted, the nature ofthe reaction with different halides is direct evidence of the truthof the covalency rule.The Covalency Rule and Atomic Dimensions.Some writers have argued that the increase of the covalencymaxima with atomic number is merely a result of the sizes of theatoms, and that the reason why, for example, [SiP,]" can existbut not [CF,]" is only that there is not room for six fluorine atomsround the smaller carbon.It is interesting to consider this argumentin detail, and see how it has been affected by the increase of ourknowledge of atomic dimensions. It was first used by those whoassumed that the general type of linkage was ionic, and that carbontetrachloride, for example, should be written C4+(Cl-),. Now, inthis question of the space available for attached atoms, it is offundamental importance whether we regard the links as ionised1* P.Lebeau and A. Damiens, Compt. rend., 1927,185,652; A., 1927,1044;1929,183, 1253; A., 1929, 779; Ruff and W. Mend, 2. anorg. Chem., 1930,190, 257; A., 1930, 877; Ruff and K. Clusius, ibid., p. 267; A., 1930, 986;H. von Wartenberg and G. Klinkott, ibid., 193, 409; A., 1931, 42.19 The small solubility of these gases in water is no objection t o the theory,since we need not supposo that more than a small concentration of the co-ordination compound is attained.20 L. Ra. Dennis and E. G. Rochow, J . Amer. Chem. SOC., 1932,54,832; A.,1932, 485; 1933, 55, 2431 ; A., 797SIDGWICK : THE COVALENCY RULE.117or covalent, because the presence of a positive charge diminishesand that of a negative charge increases the atomic radius. Thiscan be seen from Table 11, which includes certain ions, such asC4+ and S6+, whose radii can be calculated but cannot be observed,since they do not really exist, the molecules which were supposedto contain them being actually covalent.TABLE 11.Radii of atoms and ions (H.U.).C ......... 0.77 C4+ ...... 0.2 C1 ...... 0.97 C1- ......... 1.81Si ......... 1.17 Si4f ...... 0-4 Br ...... 1.13 Br- ......... 1-96S ......... 1.06 S6+ ...... 0.32 I ......... 1.35 I- ......... 2.19F ......... 0.68 F- ...... 1.33The geometry of the systems is simple. For AB, we havenormally a tetrahedron, with a valency angle of 189.5"; for AB,the arrangement is octahedral, and the angle 90"; the same angleof 90" occurs in the rare cases where we have four covalencies in aplane, as with nickel ; for a molecule AB, the structure is presumablycubic.We assume that the atoms are spherical, and that thecondition of stability is that every B should be able to touch A.If a, by are the radii of A and By then the maximum values of bwhich permit of this are :For AB, (tetrahedral) ... b = 4-45 a... h -1 2.42 a... ...For AB, (tetrahedral) or AB, (plane)For AB, (cubic) ... ... ... ... b = 1.37 aIt is a t once evident from these results and the radii in Table I1that on the hypothesis of ionisation there is not room for the atomsof even so stable a molecule as carbon tetrafluoride ; C4+ has a radiusof 0.2 A.U., and so the largest atom of which it could accommodatefour would have a radius of 0.89, whereas even the fluorine ion is50% larger than this (1.33 A.U.).The results are shown graphicallyfor carbon fluoride and iodide and for silicon fluoride, in the figure.I n the left-hand series of diagrams the radii used are those of theions (C*+, Si4+, I?, I-), and in the right-hand series those of theneutral atoms. Each of the six figures gives the positions of theatoms of the molecules AB, and AB,; the halogen atom at thetop makes a valency angle of 90" with that on the left of the centralline, and one of 109.5" with that on the right. The B atoms aredrawn to touch A, so that the condition of stability is that theyshould not overlap.When we pass from the ionised to the covalentfigures the position is entirely changed, since the absence of chargemakes A much larger and B smaller. It now appears that thereis ample room, not only for the number of attached atoms in actualstable molecules, but for far more than are ever found; even i118 INORGANIC CHEMISTRY.the imaginary CI, there would be space left between the iodineatoms; so that it would seem that the steric influence can havenothing to do with the covalency limits.Ionised.Valency angle.Covalent.Valency angle.90" 109-5" 90"F6 6 c'6 6 SiF6dcF6a , dh dA d .AB4 AB6109.5"This is shown more fully in Table 111, in which the ratio b/a(calculated for covalent links) is given for a series of atoms; aswe have seen, the number of B's which can surround one A is 8, 6,or 4 if this ratio does not exceed 1.37, 2.42, and 4.45 respectivelySIDGWICK THE COVALENCY RULE.TABLE 111.Values of bla for covalent links.A== CB = H 0.48C 1.000 0.91F 0.88C1 1.26Br 1.47I 1.7500.531.101.000.971-391.611.93Si0.320.660.590.550.830.971-15119S0.350.730.660.640.921.071.27On this basis there is room round an atom of carbon (whose ob-served covalency maximum is 4) for 8 atoms of fluorine or oxygen,and for 6 atoms of almost any kind, and round silicon or sulphur(where the maximum is 6) for eight atoms even of iodine.It is certain that forthese compounds the covalent picture is the true one, and this isnow generally recognised.But it has become evident that thistype of model, in which the atoms are represented by spheres, withradii calculated from the lengths of the links, is imperfect. Suchmodels express satisfactorily the distances between the centres oflinked atoms; but the distances between those of unlinked atoms,even when they belong to the same molecule, are much greater.We ought to regard the spheres as covered on their outsides witha sort of " envelope " which is of the order of 0.5 B.U. in thickness;or t o speak more scientifically, we have t o take into account themutual repulsion of the electrons of unlinked atoms, which isconsiderable even when the spheres of the simple model are 1 A.U.apart ?l The simplest example is a'ff orded by carbon tetrachloride,in which the spheres representing the chlorine atoms are 1.05 A.U.apart, or the " envelope " is 0.53 A.U.thick, and yet the electronicrepulsion is still str0n.g enough at this distance to stretch the linkby 0.09 X.U., and to diminish the heat of its formation by some3,500 calories. The magnitude of this repulsion, and its variationwith the distance and with the nature of the atoms, cannot yetbe calculated, but the example of carbon tetrachloride is enoughto show that it seriously interferes with the usefulness of the simplespherical models, and that steric influences may play an importantpart iii limiting the stability of molecules even where these modelsgive no indication of their presence. At the same time it is doubtfulwhether size is the determining factor in fixing the covalency maxima.This can be seen if we calculate the thickness of the " envelopes ''in a series of actual molecules, and compare them with those whichwould occur in molecules in which the covalency maxima wereexceeded. The results for molecules of the type of CB, and SB,,2 1 See Ann.Reports, 1932, 29, 67.The story does not, however, end here120 INORGANIC CHEMISTRY.together with thoae for the imaginary type CB,, are given in TableIV, being calculated on the values obtained for the atomic radiiin unstrained links ; the behaviour of carbon tetrachloride showsthat in many of these molecules the links would be more or lessstretched, and the “ envelopes ” correspondingly increased.Itwill be seen that the “ envelope ” would be practically the same inCF6 as it is in CI,.TABLE IV.Thickness of “ envelopes ” (B.U.).CB, (CB,) 8%B = F 0.5 1 0.35 0.55c1 0-45 0.26 0.46Rr 0.42 0.22 0.42I 0.38 0-15 0-350 0.50 0.34 0.54Whatever may be the true connexion between atomic dimensionsand the covalency maxima, the evidence available up to date isentirely in favour of the covalency rule as expressing the relationof the values of these maxima to the position of the elements inthe Periodic Table. N. V. s.6. THE “INERT PAIR” OF VALENCY ELECTRONS.This name has been given to the property observed in certainelements, especially in the heavier elements of the B sub-groups ofthe Periodic Table, of behaving in some compounds as if two of thevalency electrons were absent, or had become part of the core.This results in the element resembling in these compounds the onewhich is two places in front of it : thus we have T1’ like Au., Pb”like Hg”, ICl, and IF, like SbC1, and SbC1,.The theoretical reason for this has been pointed out by H.G.Grimm and A. Sommerfe1d.l The maximum sizes of the quantumgroups are given by 2n2, where n is the principal quantum number ;they are 2, 8, 18, and 32. Now a group of 8, which is the maximumfor the quantum number 2, can also behave as practically complete,showing no tendency to take in more electrons, even when itsquantum number is 3, 4, 5, or 6, as we see in the outermost groupin argon, krypton, xenon, and emanation; and in the same way18 can act as a complete group in xenon and emanation, where itsquantum number is 4 and 5 respectively.It is not clear why thesame is not true of the group of 2, but in general it is not, and aclosed group of 2 electrons is only found where its principal quantumnumber is 1. These instances of the inertness of the first two valency2. Physik, 1926, 36, 36; A., 1926, 560SIDGWICK : THE " INERT PAIR " OF VALENCY ELECTRONS. 121electrons of a group obviously occur in atoms in which this group of2 is able to show something of the completeness which it has inhelium or lithium.But while we can thus see how the phenomenon arises, we haveno theoretical guidance as to where it will occur ; we cannot give areason even for its most obvious characteristic, that it is foundchiefly among the heavier elements of the B sub-groups. Hence inorder to discover the range of its occurrence we have to rely entirelyon the chemical evidence, and examine the properties of thesuccessive periodic groups in turn.The inert pair manifests itself mainly in two ways: (1) in themonatomic cations and (2) in the covalent compounds.The outerelectronic group of a cation is normally either a complete group of8 or 18 electrons, or one of an intermediate size between these two,as in the transitional elements (in Fe" 14, in Fe"' 13); a group ofless than 8 electrons is too unstable to exist in an ion. Where,however, two of the electrons can become inert, we can have a stableion with this pair as its outer group, as in Sn" (2, 8, 18, 18, 2) orT1' (2, 8, 18, 32, 18, 2).In the covalent compounds the presence ofthe inert pair is shown in a different way. I n general, a " mixed "valency group-one containing both shared and unshared electrons-is never larger than 8, though pure (wholly shared) groups of up to16 electrons are, of course, well known. The only exceptions arewhen the atom contains an inert pair, and then it is always foundthat the valency group can be reduced to a normal form-eitheran octet or a wholly shared group-by the removal of two unsharedelectrons from the valency group to the core. For example inSbC1, and SbF, we have normal valency groups, in the former2, 5, a mixed octet, and in the latter lo, a fully shared decet (theshared electrons are underlined). I n ICl, the iodine has the group4, G, and in IF, 2, ; these are abnormal, but reduce to the normalforms of the antimony halides if two unshared electrons are con-sidered as part of the core.Thus the effect of the inertness of the pair is always to change thevalency by two, but while it diminishes it in the ion, it increases itin the covalent compounds.It need scarcely be said that no element exhibits the inert pairin all its compounds; there is always a number in which the pair isactive; the strength of the tendency for it to assume the inertstate varies greatly in different elements.There is nosign of the inertness in the A sub-groups, so that we are concernedonly with the typical elements and those of the B sub-groups.Thephenomenon obviously cannot occur in Group I. In Group I1 itWe may now consider the periodic groups in turn122 INORGANIC CHEMlSTRY.would involve the disappearance of all the valency electrons, andthe acquisition of a character like that of an inert gas. There aredistinct signs of this in mercury; it is the only element other thanthe inert gases which gives a monatomic vapour in measurableconcentration anywhere near the ordinary temperature. Cadmiumand zinc show no such behaviour (b. p.'s Zn 906", Cd 778", Hg 357").We may therefore recognise the inert pair as occurring to someextent in mercury, but in no other element of the second group.The inert pair appears first in indium, which formsa unimolecular monochloride, InC1, which in the fused state is agood conductor of electricity, and must be supposed to contain In'ions; this chloride is very unstable and is decomposed by water.Gallium does not form a monochloride. The bivalent state of bothgallium and indium is without parallel elsewhere in the PeriodicTable, and is unexplained.I n the covalent state, as we have seen,these lower valencies imply, not more than an octet, but less.In thallium, as we shou!d expect, the inertness is much moremarked. The thallous ion Tl', with an outer group of two (inert)electrons, is quite stable, and resembles the ion of an alkali metalor silver. I n the covalent thallous compounds, such as the allcox-ides Tl*O*Alk, the imperfect valency group of two shared electronstends to make itself up to 8 by polymerisation; the alkoxides formquadruple polymerides in which the octet is completed by co-ordinat ion.The evidence in this group needs to be examined withcare.The bivalent state in general is very stable in lead, a i d thestability falls off steadily as we ascend the series until it is nearlyzero in silicon; it then reappears in carbon. This, however, is noevidence of the inert pair where the compounds in question arecovalent, as they are with carbon, silicon, and germanium; theirsimple formulation as C Z O , Ge<C1 would give the atom a valencysextet. This will have a strong tendency to grow to an octet, whichit can do by co-ordination. With carbon, as has been shown,3 theco-ordination occurs within the molecule, as in C'O.With siliconand germanium this is for some reason impossible, but the fact thatthe bivalent compounds of these elements are all solids shows thatco-ordination takes place between molecules, as inGroup I I I .Group I V .c12 Siclgwick arid Sutton, J . , 1030, 146 1.I, Langmuir, J. Amer. Chem. XOC., 1919, 41, 1543; D. L1. Hammick,1%. G. A. New, N. V. Sidgwick, and L. E. Sutton, J . , 1930, 1876; A., 1930,1239; Sidgwick, Chenz. Reviews, 1931, 9, 77SIDGWICK: THE “INERT PAIR” OF VALENCY ELECTRONS. 123with the completion of the octet and the production of a solidpolymeric form. Why there should be this difference we do notknow; it is perhaps due to a greater strain in the multiple links ofsilicon and germanium ; the contrast between the volatile carbonmonoxide and the non-volatile SiO and GeO obviously recalls thatbetween carbon dioxide and silica.It is at any rate evident thatwe cannot infer the presence of the inert pair from the bivalentcompounds of these elements when they are covalent, but onlywhen they are ionised. The f i s t element of the group to givebivalent ions is tin (there is no satisfactory evidence of a germanousion). I n lead, as usual, the inertness is much more marked; whilethe stannous ion is unstable, and readily passes into the stannio,with the reappearance of the activity of the pair, the plumbous ionPb” is the stable form, and has no reducing properties.Hitherto our evidence of thc inertness has been derivedexclusively from the ions.In this group we have also a second lineof argument, from the covalent compounds. The two may be takenseparately. A tervalent cation of a fifth-group element has twoof the electrons of its valency group left, obviously in an inert state.It has the structure : (core), 2, as have Sn” and T1’. Such an ionis not formed by nitrogen or phosphorus, and there is no goodevidence of its occurrence with arsenic; but there are clear in-dications of the existence of an antimony ion Sb***, and with bis-muth the phenomenon is very marked. There is a whole series ofsalts in which bismuth acts like a metal of the third group such aslanthanum, as in the isomorphous series of complex nitratesM,”[M”’(N03)6]2,24H20, where M” may be Mg, Zn, Co”, Ni; andM”’ may be Bi, La, Ce, Pr, and other rare-earth metals.Thus theionic evidence indicates that the inert pair begins with antimony.When the tervalent ion passes into the covalent state, we merelyhave a normal octet produced, and the inactivity disappears;thus in [BiICl, the bismuth has the structure (60)(18)2, and inG~oup V .Bi-C1 P1 it has (60)(18) 2, 5 ; there is no doubt that the trihalides of\Clantimony and bismuth are largely in the covalent state. But ifa covalent derivative of this type forms a complex anion, this mustcontain an inert pair of electrons, because its valency group hasincreased beyond 8. For example, in K[SbCl,] the antimony, withfive electrons to start with, has gained four more from the chlorinesand one from the potassium : it has a decet with 8 shared electrons,2, 8.Ashas been said, a mixed valency group with as many as 8 sharedelectrons is never found to have wore than two unshared electrons,I n KJBiI,] the bismuth has the valency group 2, 14124 INORGANIC CHEMISTRY.and this only when these two are inert.* Now there are a numberof complex salts of this type derived from the trihalides of antimonyand bismuth. In addition to those mentioned above we haveK[SbF,], KCBiCl,], and K3[Bi(CNS)6]. No such complex deriv-atives of tervalent nitrogen or phosphorus have been described, noruntil recently of tervalent arsenic.has described a series of salts of the types M[AsCl,], M[AsBr,],M,[AsCl,], in addition to more complicated forms containing severalatoms of arsenic in the molecule.The cations are alkylamrnoniumsand pyridinium. The salts (the chlorides are colourless and thebromides pale brown or yellow) are soluble in concentrated halogenacids, but are at once decomposed by water. This is the onlyevidence we have of the occurrence of the inert pair in arsenic, andfrom the properties of the compounds i t is clear that this state ofthe arsenic atom is very unstable.It should perhaps be pointed out that the formation of complexanions from the pentahalides is no proof of the inert pair ; K[SbCl,],for example, has a vnlency group of 12 shared electrons.This group raises questions of considerable interest.There is no doubt about the occurrence of the inertness in tellurium,which gives a definite quadrivalent cation Te"", with the structure2, 8, 18, 18, 2.The evidence for a corresponding selenium ionSe"" is very weak.Apart from the ions, we should look for evidence of the inertnessin the tetrahalides such as XC14, with a valency group of (2) 8.Tellurium tetrachloride, though its melting point (214") and boilingpoint (414") are high, and the liquid from its high conductivity(0.1145 a t 236"),5 must be largely ionised, can evidently exist inthe covalent state, since its vapour density at 536" is 90% of thetheoretical value.Selenium tetrachloride is a stable solid, but is relatively non-volatile, and its vapour density at 200" is only about half of thatcalculated for SeCl,, indicating that it is largely dissociated intoSe,Clz and chlorine.Selenium tetrafluoride, however,6 melts at-13" and boils a t 93" and is no doubt a genuine covalent compound.The presence of the inert pair in tellurium and selenium is con-firmed by the existence of complex salts derived from the tetra-This year, however, W. PetzoldGroup V I .2. anorg. Chem., 1933,214, 355; A., 1258.A. Voigt and W. Biltz, ibid., 1924, 133, 297; A., 1924, ii, 552.E. B. R. Prideaux, J., 1928, 1603.* The only exception to this rule is the series of salts of the type of K[IC14],in which the iodine appears to have the valency group 4, S, which on theassumption of the inert pair reduces to (2), 2, S. As this stands as an isolatedexception we may suppose that the true formula, is a multiple of that givenabove; the molecular weight has not been determinedSIDGWICK : THE " INERT PAIR " OF VALENCY ELECTRONS.125halides, such as KCTeCl,], K,[SeBr,], and K2[TeC1,]. Seleniumtetrachloride is remarkable for forming a 1 : 1 addition compoundwith antimony pentachloride. This must have the structureCl,Se+SbCl,, the inertness having disappeared, and the seleniumhaving acquired a valency group of 10 and the antimony of 12shared electrons.We may therefore conclude that the pair can be inert in telluriumand selenium.When we come to sulphur we find a different state of affairs.Sulphur tetrachloride, SCl,, can only exist in the solid state at verylow temperatures ; it already has a dissociation pressure (givingSC1, and chlorine) of more than an atmosphere at -30".has shown that there is no reason to think that it exists a t all in theliquid state in a solution of sulphur in chlorine. He has also pointedout that the dielectric constant of the solid tetrachloride is muchhigher than we should expect for anything but a salt, and he suggestsnot improbably that this substance, which can only exist as a solidand is apparently insoluble in a non-ionising solvent like liquidchlorine, is actually a salt, trichlorosulphonium chloride [SCl,]Cl,in which, of course, the sulphur would have the normal octetstructure 2, 5.LowryI n the same way the active sulphonium compound of W.J.S/CH2*Co2H has all the character- CH3 Pope and S. J. Peachey 8 C 2 H 2 [XIistics of a salt. It is nonrvolatile, it has a high melting point (ifany), and it is precipitated from alcoholic solution by ether in anon-solvated form.The salt character of the sulphonium com-pounds is most clearly shown by comparing trimethylsulphoniumiodide with tetramethylammonium iodide on the one hand andtrimethyltin iodide-a genuine covalent compound-on the other :(CH3 )m. (CH3 1381. (CH,),Snl.Decomposes without Dissociates without B.p. 170"melting a t 214"Sol. in water Sol. in water Sol. in waterInsol. in ether Insol. in ether Sol. in ether, benzene,Another strong indication that the sulphur atom cannot assumethe true 4-covalent form, i.e., that the pair cannot become inert, isderived from the optically active 3-covalent sulphur compounds,such as the sulphoxides.melting a t 230'etc.These might have two formulae :Valency group, 2,s. Valency group, 2,g.7 T.M. Lowry, L. P. McHatton, and G. G. Jones, J., 1927,746 ; T. M. Lowryand G. Jessop, J., 1929, 1421; 1930, 782. * J., 1900, 77, 1072126 INORGANIC CHEMISTRY.The first of these involves an inert pair, but the second does not;the activity of the compounds definitely rules out the first formulaand establishes the second. Now, if the first structure could exista t all, it should be present in solution a t least in small quantityin tautomeric equilibrium with the second; but every time anactive molecule went over into it, it would lose its activity, and sothe compound would racemise rapidly. The fact that the activesulphoxide does not racemise shows that the doubly linked form isnot even temporarily assumed, and this implies that there is somefundamental objection to it, and the only one that can be suggestedis that the inert pair which it involves cannot occur in sulphur.One compound has been described which if it is real proves thepossibility of the inert pair in sulphur.This is sulphur tetra-fluoride SF,. According to J. Fischer and W. Jaen~kner,~ this isformed by heating cobaltic fluoride, CoF,, with sulphur, and is acolourless gas, m. p. -1124", b. p. - 40°, mol. wt. 107 (theory 108),which when dry does not attack glass, sulphur, or rubber, but doesattack mercury. If this exists it is obviously a covalent fluoride,and the sulphur atom has the inert pair. The evidence against thephenomenon in sulphur is so strong that an isolated exception needscareful scrutiny. The paper of Fischer and Jaenckner is called a" preliminary communication," and they have published no sub-sequent paper on the subject during the last four years, nor doesRuff, in his numerous papers on the fluorides, seem to have madeany reference to it.Also the differences between the properties ofthe substance and those of the undoubted selenium tetrafluorideare far greater than we should expect the replacement of seleniumby sulphur to produce, as may be seen by comparing the meltingand boiling points of these compounds with those of the hexa-fluorides :SF,: m. p. - 124", b. p. - 40". SF,: m. p. - 56", b. p. - 62".SeF,: ,, - 13", ,, + 93". SeF6: ,, - 35", ,, - 39".We may therefore assume provisionally that there is some mistakeabout sulphur tetrafluoride, and that the inert pair, which is verymarked in tellurium and selenium, cannot appear in sulphur.I n this group the evidence depends wholly on thecovalent compounds; a simple cation with two inert electronswould have 5 positive charges, and this is more than any atom cancarry.The covalent compounds of polyvalent halogens are of twotypes, neutral molecules, as IC13, and anions, as [IC12]'. In theanions of this type the linking atom has a valency group of 10,i e . , 6, 4, which involves the inert pair, giving (2), 4, 4. It is wellGroup VII.2. anyew. Chein., 1929, 42, 810; A . , 1929, 1156SIDGWICK : THE '' INERT PAIR " OF VALENCY ELECTRONS.127known that t,he trihalides are more stable the heavier their com-ponent atoms, and that the stability depends on the presence ofone heavy atom in the complex; for example, while [Cl,] is almostif not quite unknown, [ICl,] is nearly as stable as [I,]. This is whatBB we should expect; in a trihalogen ion A< it is only the A atomwhich exhibits the inertness, and this is always more marked in theheavier atoms. Definite evidence that the heavier atom is in themiddle of the complex is afforded by the crystal structure ofFor some reason, while iodine, bromine, and chlorine are almostequally ready to act as the terminal members of the complex ion,fluorine as a rule does not do so. This, of course, has no relation tothe question of inertness; it is presumably due to the strongtendency of fluorine to pass from the covalent to the ionised state.Until lately it was not known that fluorine could ever form part of apolyhalide ion, although it forms neutral compounds with otherhalogens a t least as easily as any other element of the group.Re-cently Booth and his co-workers l1 have prepared the salts Cs[ICl,F]and Rb[ICI,F] by the action of iodine trichloride on the alkalifluorides. These belong to the type M[I,], and presumably haveCS[ICI,] .lothe structure I<" C1 where the iodine and the linking chlorineCl<Fhave the inert pair. They are of interest as showing that fluorinecan form part of a polyhalide ion, but perhaps only when it is a longway from the negative charge.From the point of view of the inert pair, the question is, whatis the lightest halogen which can form the linking atom in a poly-halide ion ? There is no doubt that bromine can do this ; whetherthe same can be said of chlorine is less certain.The evidence forthe existence of [Cl,]' is not satisfactory; but we may perhapsaccept the above formulation of Booth's compounds as making itprobable.When we coine to the neutral inter-halogen compounds (see below,p. 128) there is no doubt that the pair can be inert in chlorine;chlorine trifluoride, CIF,, is a perfectly definite though very re-active substance (m. p. - 83", b. p. + 13").We can sum up our knowledge of the occurrence of the inertpair of electrons as follows :(1) It is never found except in the sub-typical (second shortperiod) and B elements.10 1%.\V. G. Wyckoff, J. Anaer. C'hem. Soc., 1920, 42, 1100; A., 1920, ii, 489.11 H. S. Booth, C. F. Swinehart, and W. C. Morris, J . Amer. Chem. Xoc.,1932, 54, 2561; J . Physical Chem., 1932, 36, 2779; A., 1932, 823, 1219128 INORGANIC CHEMISTRY.(2) In every group it becomes more stable as the atomic weight(3) As we pass from the earlier groups to the later, we find thatThe limits of its occurrence are marked by the line in the table.and number increase.it extends to lighter elements.Occurrence of the inert pair of electrons.Be B C N 0 FMg A1 Si P s 1 c1Cd 1 In Sn Sb To IZn Ga Ge 1 As Se BrHg T1 Pb BiN. v. s.7. THE INTER-HALOGEN COMPOUNDS.Eleven compounds of the halogens with one another have nowbeen described. Some of these have been known almost as long asthe halogens themselves, but several have only recently been dis-covered, largely through the work of Ruff and his collaborators.None is known which contains more than two different halogens,and all those which have more than two atoms in the molecule areof the type AB,, in which B is either fluorine or chlorine, and n is3, 5, or 7 ; it is evident that all the B atoms are directly joined toA.The valency group of A in the various types is therefore :Valency group of A ............ 6, 2 (2) 2, 5 (2) Ilo_ --It will be seen that in AB, and AB, the atom A has an inert pair ofelectrons (see p. 127), and that a molecule AB, is only possible whenA is iodine, since chlorine and bromine are limited by the covalencyrule to a covalency of 6.Type AB.Type.AB. AB,. AB,. AB,.14These Compounds are obviously of the same type as the freehalogens themselves. Of the six possible compounds, all are knownexcept iodine monofluoride, which is probably too unstable to exist,since bromine monofluoride is highly unstable. Their boiling andmelting points (the latter in parentheses) are given in Table I,along with those of the elementary halogens.Chlorine rnonojluoride, C1F.l Made by the action of slightly moist1 0. Ruff et al., 8. angew. Chem., 1925, 41, 1289; A , , 1929, 160; 2. anorg.Chem., 1928,176,258; A., 1929, 40; Ruff and F. Laass, ibid., 1929,183, 214;A., 1929, 1226; Ruff and W. Menzel, ibid., 1931,198, 375; A., 1931, 912; 0.Ruff, F.Ebert, and W. Menzel, ibid., 1932, 207, 46; A., 1932, 902; Ruff andA. Braida, ibid., 1933, 214, 81; A . , 1130; K. Fredenhagen and 0. T . Krefft,2. physikal. Chm., 1929, 141, 221; A,, 1929, 664SIDQWICK : THE INTER-HALOGEN COMPOUNDS 129chlorine on fluorine, or by heating dry chlorine with fluorine at250" ; the reaction is exothermic and reversible. The compound isalmost colourless; it reacts like fluorine, but even more readily.TABLEBoiling and melting pointsF. c1.F - 188O( - 219O)CI - 100" - 33.7"(- 156") ( - 102.3")Br + 20" + 5 O( - 33") ( - 66") + 97.4"I I (u + 27.2')(jl + 13.9")I.of the AB compound&.Br. I.+ 58.8"( - 7.2")116" 184"(36") (114')Bromine monoJluoride, BrF.Owing to its instability this com-pound was not obtained until this year.2 Bromine reacts very slowlywith fluorine in the gaseous state a t O", while at 50" the only productof the reaction is bromine trifiuoride. By saturating bromine withfluorine at + 10" and fractionating under reduced pressure, themonofluoride was obtained in a nearly pure condition, but itchanges continuously into the trifluoride and free bromine. It is apale red gas, which gives a dark red liquid, and orange crystals likepotassium diehromate. It is remarkable that the monofluoride isonly slightly soluble in bromine, though BrF, BrF,, and BrF, areall miscible with one another. It is clear that the monofluoride isthe first product of the action of fluorine on bromine, but that itchanges spontaneously into the other fluorides and bromine.Itsmarked instability as compared with chlorine monofluoride makes itprobable that iodine monofluoride, which has never been prepared,cannot exist.Bromine monochloride, BrC1. This has a curious history. Balardin 1826 noticed that when bromine is mixed with chlorine thecolour diminishes, but this observation was disregarded for a century.Then in 1928 G. M. B. Dobson noticed the same thing when usingthe mixture as a light filter for observing the ozone bands in theatmosphere, and his conclusion was confirmed by S. Barratt and c. p. Stein 3 with the spectrophotometer. Meanwhile B. J. Karsten4in 1907 had worked out the complete phase-rule diagram for bothvapour pressures and freezing points of the system bromine-chlorine, and concluded from his results that no compound was2 0.Ruff and A. Bmida, 2. anorg. Chem., 1933, 214,81; A., 1130.4 2. anorg. Chem., 1907, 53, 365; A*, 1907, ii, 447.3 P ~ o c . ROY. SOC., 1929, 122, 682 ; A*, 1929, 411.REP.-VOL. XXX. 130 INORGANIC CHEMISTRY.formed. All the curves-vapour, liquidus, and solidus, both forvapour-liquid and for liquid-solid equilibrium-proceed con-tinuously from Br, to C1, and give no sign of combination. This is,however, compatible with the existence of a compound, if it ispartly dissociated and in the solid forms a continuous series of solidsolutions with both of its components, which is not improbable inview of the identity of molecular type of the three substancesBrz, BrC1, and Cl,.The spectroscopic results of Barratt and Stein were c o n b e d byL.T. M. Gray and D. W. G. Style; 5 from the fall of intensity of theabsorption band of bromine in the visible, it appears that at 20"the equilibrium constant [BrC1]2/[Br,]*[C12] is about 5, so that atthe ordinary temperature the mixture contains about 80% of BrC1.Confirmation of the existence of BrCl from another side wasobtained by N. H. W. Hanson and T. C. James,6 who found that amixture of bromine and chlorine adds on to unsaturated acids andesters much more rapidly than either element separately, and givesa mixture of a-chloro- p-bromo- and a-broino- p-chloro-compounds.T. W. J. Taylor and L. A. Forscey found the same for the reactionwith diazoacetic ester, which gave mainly the chlorobromoacetate ;this reaction has the advantage that it is practically instantaneous,and may be supposed to give a true measure of the proportion ofBrCl present; on this hypothesis they find the proportion to beSOYO, in agreement with the spectroscopic results.W.Jost 8 has measured spectroscopically the equilibrium inmixtures of bromine and chlorine at different temperatures, andalso the rate (of the order of 15" for the half change) of the reaction.He finds the heat of reaction t o be very small, and the heat ofactivation about 14,000 calories.According t o S. An~ar-Ullah,~ if chlorine is passed into bromineunder water below 18", a crystalline hydrate BrC1,4H20 is formed,which is more stable than the hydrate of bromine or chlorine.Discovered by Davy and by Gay-Lussac in 1814. It is formed by the combination of the elements,or by the oxidation of an iodide in hydrochloric acid. The phase-rule diagram lo of the system iodine-chlorine is familiar. Iodinemonochloride occurs in two solid forms, a, red, m. p. 27.2", and p,brown, m. p. 13.9". They give the same liquid, b. p. 97.4"; thereIodine monochloride, IC1.5 Proc. Roy. SOC., 1930, 126, 603; A., 1930, 510.6 J., 1928, 1955, 2979.* 2. physikd. Chenz., 1931, 153, 143; ibid., [Bl, 14, 413; A., 1931, 431.9 J . , 1932, 1176.J., 1930, 2272.10 W. Stortenbeker, Rec. trav. chim., 1888,7,152; A,, 1889,102 ; Z. physikal.Chem., 1889, 3, 11SIDGWICK : THE INTER-HALOGEN COMPOUNDS. 131is little dissociation in the vapour even at Chemically thecompound behaves like a mixture of its components ; e.g., with waterit forms hydrochloric and hypoiodous or iodic acids.Iodine monobromide, IBr. Discovered by Balard ; formed bydirect combination. The solid looks like iodine and the vapouris dark red. Iodine and bromine form a continuous series of solidsolutions, like chlorine and bromine. The heat of reaction and theequilibrium constants have been determined by D. M. Yost, T. F.Anderson, and F. Skoog.12 They find the reaction to be slightlyendothermic. The degree of dissociation at 25" is about 8% in thevapour and about 9% in carbon tetrachloride solution.Type AB,.Three compounds of this type are known, ClF,, BrF,, and IC1,.Chlorine trijiuoride, ClF,.13 This is produced by heating chlorineor chlorine monofluoride in excess of fluorine. Colourless gas andsolid, pale green liquid; m. p. - 83", b. p. + 13". It is highlyactive ; it reacts with most substances explosively.Bromine trijluoride BrF,. l4 Made from its elements. Colourlessgas, liquid, and solid; m. p. 8%", b. p. 127". The heat of evapora-tion is 10,000 calories, which gives a remarkably high Troutonconstant. The values of this constant are : Br, 23.2; BrF, 25.3;BrF, 23.7. These values are all high for non-associated liquids,and that for the trifluoride is almost as high as for an alcohol.Bromine trifluoride is very reactive and fumes in air; it is, ofcourse, the acid fluoride of bromous acid. Its formation by thespontaneous decomposition of the monofluoride has already beenmentioned.Iodine trifluoride, like the monofluoride, is unknown, and in viewof Ruff's work on the heptafluoride may be presumed not to exist.Iodine trichloride, ICl,. Discovered by Gay-Lussac in 1814.Formed by direct combination of the elements. Lemon-yellowneedles ; melts at 101" under its own (dissociation) vapour pressureof 16 atmospheres. The density shows that the vapour is almostwholly IC1 + Cl,. It is miscible with iodine in all proportions.11 J. McMorris and D. M. Yost, J . Amer. Chem. SOC., 1932, 54, 2247; A.,12 J . Aqner. Chem. Soc., 1933, 55 552: A.. 351.1s 0. Ruff and H. IZrug, 2. anorg. Chem., 1930, 190, 270; A., 1930, 878;Ruff, F. Ebert, and W. Menzel, ibid., 1932, 207, 46; A., 1932, 902.14 P. Lebeau, Compt. rend., 1905, 141, 1018; A., 1906, ii, 80; Ann. Chim.Phys., 1906, [viii], 9, 248; E. B. R. Prideaux, J., 1906, 89, 316; 0. Ruffet al., 2. anorg. Chem., 1932,206, 59; 0. Ruff and A. Braida, ibid., 1933, 214,91; A., 1130.1932, 906132 INORGANIC CHEMISTRY.Type AB,.Bromine pentajluoride, BrF5.15 Obtained by heating brominetrifluorida with fluorine at 200". Colourless liquid; m. p. - 61*3",b. p. + 40.5". Very reactive.Obtained fist by G. Gore16 by theaction of iodine on silver fluoride; prepared by €5. Moissan by thedirect combination of the elemen%s.17 It is a colourless heavy liquid,m. p. - 8", b. p. 97". Dissolves iodine and bromine, and fumes inthe air, being converted by water into hydrogen fluoride and iodicacid.l'ype AB,.Iodine pentujluoride, IF,.Iodine heptu$uoride, IF,. This remarkable substance was pre-pared by Ruff in 1930.1s It is made by acting with fluorine on thepentafluoride at 270--5OO", and fractionating the product. Itforms white crystals, which melt at + 5-6" and boil a t 4.5". Theheat of evaporation is 7,300 cals., which gives a Trouton constantof 26.4, the same value as for an alcohol, which suggests polymeris-ation in the liquid. The vapour density of the gas is that requiredfor IF,. It is not improbable that it should poljmerise to someextent in the liquid; it is the only known neutral binary compoundwith a covalency of 7, and presumably is very ready to increasethis to the more stable 8, which it could do by co-ordination toF,I-F+IF,; it is even possible that it may form a chelate ring,like aluminium or ferric chloride.c1 FC1,Al' 'AlC12; F61( >IF,kl' PIt is absorbed (rather slowly) by water to give periodic acid;most substances are as readily attacked by it as by chlorine trifluorideand even more vigorously. N. v. s.N. V. SIDGWICK.W. WAXDLAW.R. WHYTLAW-GRAY.15 0. Ruff and W. Menzel, 2. anorg. Chem., 1931, 202,49; A., 1932, 133.16 Phil. Mag., 1871, [iv], 41, 309.17 Compt. rend., 1902, 135, 563; see also Prideaux, J., 1906, 89, 316.18 0. Ruff and R., Keim, 2. anorg. Chem., 1930,193, 176; A., 1930, 1390

ISSN:0365-6217    

DOI:10.1039/AR9333000082

出版商:RSC    

年代:1933

数据来源: RSC

摘要:

ORGANIC CHEMISTRY.PART I.--AT,IPHATIC DIVISION.Acetylenes, OleJin-acetylenes and Allenes.THE discovery of a convenient means of preparing vinyl- anddivinyl-acetylene by the polymerisation of acetylene has givenan impetus to the investigation of acetylenic behaviour as it appearsboth in simple acetylenes and in olefin-acetylenes. A speciallyinteresting feature of work on the latter group is the light that isthrown on the formation and reactivity of allenic systems (*CZC=C*)and on the reactivity of such systems when associated with additionalunsaturated units.Jletallic and Orglano-metallic Derivatives.-The mercury deriv-atives of monoalkyl- and monoaryl-acetylenes are formed by theaction of potassium mercuri-iodide (K,HgI,) or mercuric cyanideand potassium hydroxide on the hydrocarbons.These compoundsof the general formula Hg(CiCR), are crystalline substances ofsharp melting point which are suitable for the characterisation ofthe Sodium vinylacetylide, C131,:CH*CiCNa,formed by the action of sodium or sodamide on vinylacetylene,is a reactive powder which can be employed for introducing thevinylacetyleaiie group into organic compounds ; when the sodiumderivative is formed in the presence of a ketone (or less satisfactorilyan aldehyde), vinylethinylcarbinols, CH,:CH*CiC*C( OH)R,, areproduced.* The formation of crystalline bis-organomercuricacetylides by passing acetylene into an alkaline solution of alkyl-or aryl-magnesium halides (C2M, + RHgX -+ RHgCiCNgR)affords a valuable means for the identification of the latter groupof substances ; but although the bis-derivatives are the only isolableproducts of reaction, the mono-acetylides, RHgGCM, appear tobe formed in solution when an excess of acetylene is empIoyed.5r\;ionosubstituted acetylenes appear to react in a similar way toacetylene, since vinylacetylene yields with et hylmagnesium bromide,1 Ann.Reports, 1932, 29, 112.2 T. H. Vaughn, J. Amer. Chem. SOC., 1933, 55, 3453; A., 1033.3 W. H. Carothers, R. A. Jacobson, and G. J. Berchet, ibid., p. 4206.4 W. H. Carothers and R. A. Jacobson, ibid., p. 1097; A., 455.5 R. J. Spahr, R. R. Vogt, and J. A. Nieuwland, ibid., pp. 2465, 3728;A., 815, 1177134 ORGANIC CHEMISTRY.-PART I.vinylethinylmagnesium bromide, CH,:CH*CICMgBr-a compoundwhich reacts with acetone and with carbon dioxide in the mannernormal to organomagnesium halides, but falls far short of thehigh degree of reactivity towards alkyl and aryl halides whichis displayed by sodium vinylacetylide (CH,:CH*CiCNa +CH,:CH*CCR) .6HaZogenoacetyZenes.-Acetylene and sodium acetylide react withiodine dissolved in anhydrous liquid ammonia to give almosttheoretical yields of di-iodoacetylene.Certain alkyl- and aryl-acetylenes, R*CiCH, also react satisfactorily at -34" with the samereagent to give iodo-derivatives, but others react only at a highertemperature and with reduced yield ; the sodium derivatives ofthe monosubstituted acetylenes, however, are iodinated instantlyin liquid ammonia.' cx-Chloro- and a-bromo-vinylacetylenes,CH,:CH*CiCX, are obtainable as distillable liquids by the directaction of chlorine and bromine on bisvinylethinylmercury.3 A veryvaluable general method, however, for the preparation of halogeno-derivatives from acetylenic compounds which retain one or morereactive hydrogen atoms consists in submitting the acetyleniccompound to the action of alkali hypoiodite, hypobromite, orhypochlorite. By this means dichloro- and dibromo-acetylene,chloro- and bromo-propiolic acid, the a-halogeno-derivatives ofnumerous Au-acetylenk alcohols and of vinylacetylene: and thedihalogeno-derivatives of diacetylene lo have been successfullyobtained .Additive Beactions.-The hydration of primary acetylenic alcohols,R*CiC*CH,*OH, or the corresponding acetates by means of mercuricacetate solution results in the addition of the elements of waterat the acetylenic bond in one direction only.Thus only the ketonesR*CO*CH,=CH,*OH (or their dehydration products, R*CO.CH:CH,)are formed.ll When substituted divinylacetylenes are hydratedby means of sulphuric acid-acetic acid, not only is the acetylenicbond affected, but further hydration, leading to cyclisation, occurs.A., 485 ; W. H. Carothers and R. A. Jacobson, ibid., p. 1622 ; A., 590.741 ; A., 930.1157.4667.1930, 1158.2, 645; A,, 1933, 161.RX6 W. H. Carothers and G. J. Berchet, J . Amer. Chem. SOC., 1933,55,1094;7 T. H. Vaughn and J. A. Nieuwland, ibid., p. 2150; A,, 694; J., 1933,8 F. Straus, L. Kollek, and W. Hsyn, Ber., 1930, 63, [B], 1868; A., 1930,a R. A.Jacobson and W. H. Carothers, J . Amer. Chem. SOC., 1933, 55,10 F. Straus, L. Kollek, and H. Hauptmann, Ber., 1930, 63, [B], 1886; A.,11 E. D. Venus-Danilova and S. N. Danilov, J . #en. Ohm. ( U.S.S.B. ), 1932FARMER. 135In this way tetra-alkyl-divinylacetylenes, CHR:CR’*C:C*CR‘:CHRpass into tetra-alkylcyclohexenes, the stages being presumably : l2CHR CHR CHR*OH CHR// // / A R’Q RHR ~!?$RIV GHR F$R’QH GHR I?$? R’VH RRC CR’ OC CR’ OC CR’ OC CR’\/CH2Hydration of vinylacetylene occurs with mercuric acetate solutionat 60-70” : here, however, it is suggested that the additive re-action involves the breaking of only one of the bonds of the triplelinkage (instead of two as normally occurs with acetylenes), therebyyielding an olefinic mixed salt, CH,:CH*C( O*H~*OAC):C(H~*OAC),.~~Addition of liquid hydrogen bromide to vinylacetylene l3 yields,as does concentrated hydrochloric acid,14 an allenic derivative (I)by terminal addition.This derivative readily suffers ay-change,to yield (11). When, however, concentrated hydrobromic acid(aq.) is employed as the additive reagent, the re-arranged bromide(11) is obtained dire~tly.1~ Addition of hydrochloric acid toa-alkylacetylenes (111) yields in each of the cases examinedCH,:CH*CiCH CH,:CH*CiCR (111.1innr JEta(1.1 CH,Br*CH:C:CH, CH,:CH*CCI:CHR (IV. 1an ap-derivative (IV).16 The course followed in the reactionbetween hydrochloric acid (aq.) and divinylacetylene is difficultto establish, since the monohydrochloride reacts with the reagentmore rapidly than does the hydrocarbon.There is some groundfor supposing that the first reaction product has the allenicconstitution (V).17 This mode of reaction, however, is unlikeany other addition of hydrogen halide to conjugated compoundsyet described in that (a) the hydrogen atom does not unite at a12 A. T. Blomquist and C. S. Marvel, J . Amr. Chem. SOC., 1933, 55, 1655;A . , 591 ; D. T. Mitchell and C. S. Marvel, ibid., p. 4276; A., 1270.13 W. H. Carothers and G. J. Berchet, ibid., p. 2807; A., 930.14 Ann. Reports, 1932, 29, 109.15 W. H. Cardhers, A. M. Collins, and J. E. Kirby, J . Amer. Chem. Soc.,16 W. H. Carothers and R. A. Jacobson, ibid., p. 1624; A., 690.17 D. D. Coffman, J. A. Nieuwland, and W.H. Carothers, ibid., p. 2048;1933, 55, 786; A., 371.A., 694136 ORGANIC CHEMISTRY .-PART I.terminal carbon atom of the olefin-acetylene system, and (b) as-addition occurs in an extended conjugated chain despite the factthat such addition has not yet been observed to occur in a conjugatedhexatriene or in any other conjugated polyolefin, and that migrationof the halogen atom from the position assigned (ay-change) wouldyield a stable conjugated chloro-triene.CH,:CH*CXWH:CH, 3 CH,Cl*CH:C:CCl*CH:C€I, PI.)I l O l (as-addition) -1CH,Cl-CH:C:CH*CH:CH, C 7H *S*CH,* CH,* CiC*CH,*CH,*S C ?H(V. 1 WI).An allenic product (VI) results from the interaction of chlorinewith divinylacetylene,l* but it is impossible to state with certaintywhether it arises by direct &addition or by ap- or a<-addition,followed by ccp-rearrangement.It is not improbable that as-addition may be a common mode of reaction in conjugated olefin-acetylenes such as divinylacetylene, since the S-halogen atom ofthe addition product then occurs in an allenic system-a fact whichpresumably restricts its mobility and therefore precludes furtherrearrangement.lg It is to be observed, however, that migrationof the a-chlorine atom, which would yield a fully conjugated system,does not occur, although the analogous system, CH2:C:CH*CH,C1,re-arranges with great ease.Remarkably enough, divinylacetylene reacts towards thiocresolas an unconjugated diolefin, the reagent adding at the etliyleniclinkages of the vinyl groups as shown in (VII).20The mode of chlorine addition to a-chloromethylallene,CH,:C:CH*CH,Cl, is the one which would be expected, i.e., additioncan occur at both double bonds, to give a mixture ofCH,:CCl*CHCl*CM,Cland CH,C1*CC1:CH-CH2C1; it is of interest, however, that the ratioof the products varies considerably with the temperature of reaction.21In vinylallem systems such as those represented in (VIII) and (IX)there is no uniform tendency for an addendum to be added at theends of the conjugated system present in each, since, although inone of four such examples examined, ' u z ' ~ ., (VIII), chlorine appearsto be added at the ends of the conjugated system to give (X), yet in18 D. D. Coffman and W. H. Carothers, J . Amer. Chern. Soc., 1933, 55,2040; A., 694.19 The tendency to form a completely conjugated system does not appearto be sufficient to bring about ay-change in the dihydrochloride of &vinyl-acetylene, C&Cl*CH:CH*CCI,*CH:CH,, also.20 W.H. Carothers, J . Amer. Chern. Soc., 1933, 55, 2008; A., 695,21 W. H. Carothers and G. J. Berchet, ibid., p. 1628 ; A., 590FARMER. 137the same example hydrogen chloride is added (almost certainlyas the primary reaction) at one of the allenic double bonds to give(XI).18 In the case of (IX), there is reason to believe that hydrogenchloride is added at the ends of the conjugated system to give(xII).~~(VIII.) CH,Cl*CH:C:CCl*CH:CH, 1 CII,Cl.CH:C:cH*CH:C€I, (IX. )CH2C1*CH:CC1*CC1:CH*CH2Cl CH,Cl*CH:CCl*CH:CH*CH,(X.1 (XII.)CH,Cl*CHXH*CCl,*CH:CH, (XI.)I n connexion with the well-known ability of diazomethane t oadd to Aa-olefinic and acetylenic esters, K.von Auwers and 0.Ungemach 22 draw the conclusion that, although in the pyrazolinederivatives formed from olefinic esters the diazo-group of the reagentis customarily attached to the a-carbon atom of the ester, yet inthe formation of pyrazole derivatives from acetylenic esters of thetype RCiC*CO,R, the diazo-group becomes attached t o the a-carbonatom if R = alkyl, but to either the a- or the p-carbon atom ifR = aryl.Reactions and Syntheses.-Application of selenium dioxide tothe oxidation of simple alkylacetylenes has shown that oxidativeattack occurs at the y-carbon atoms of the latter, n- Aa-heptinenethus giving n- Aa-heptinen-7-01 and n-Aa-octinene giving n- Am-octinen-7-01.~~ An interesting synthesis of acetylenic bases of thetype RCiC*CH,*NR, has been carried out by combining variousarylacetylenes with formaldehyde and a secondary amine, accordingto the equation :By the successive action of bromine and alcoholic potash on cyclo-penhdecene and cycloheptadecene L.Ruzicka and his collaboratorshave obtained the 15- and 17-membered cyclic acatylenes, (XIII)and (XIV).25RCiCH + CH2O + NHRZ --3 RCiC*CH,*NR, + H2O 24Catalytic Hydrogenation.Much new information has been gained concerning the applicationof catalysts to the promotion of different varieties of reductivechange. The new observations relate to ( a ) the specificity, (b) the21 Ber., 1933, 66, [B], 1205; A., 1171.23 R.Truchet, Conapt. rend., 1933, 19$, 706; A., 486.2.4 C. Mannich and F. T. Chang, Ber., 1933, 66, [B], 418; A., 387.86 L. Ruzicka, M. Hiirbin, and H. A. Boekenoogen, Helv. Chim. Act@, 1933,16, 498; A., 599.E 138 ORGANIC CHEMISTRY.-PART I.selectivity, and (c) the conditions affecting both the specificit'y andthe selectivity of catalyst action.The cleavage of oxygenated compounds with simultaneousaddition of hydrogen has a t various times been observed to occurduring catalytic hydrogenation. Such cleavage is not confinedto the t C - 0 linkage, but occurs also a t >C-C< linkages. Theemployment recently of high pressures of hydrogen , and especiallyof a copper-chromium oxide catalyst in place of the nickel andplatinum catalysts previously used, has extended the field of oper-ation and utility of this type of change, which is in its nature hydro-genolysis-strictly analogous to hydrolysis (2 C-0- 4 X X I + HO- ;@-CC + 3CH + HC<).26 It is found that the linkagesindicated by the dotted lines in formulze (XV) to (XXI) are allsusceptible of hydrogenolysis in the presence of copper-chromiumoxide a t elevated temperatures and pressures (120" to 250" ; 140-210 atm.) and to this list must be added a large number of open-chain and cyclic glycols, p-hydroxy-esters, malonic and acetoaceticesters, and other compounds which have been observed to sufferH"/ H,R.C-4- Po OEt HO ---CHR*CH,*CO,Et Ph*CHR- OH Ph*CHR*CH,---OH(XV.1 (XVI.) (XVII .) (XVIII.)CHx=:C-CH2--- OH CHMe(0H) ---CMe,*CO,Et Ph-CH,--- CMPh,CHLCH..' (XIX.);:a0(XX.) (XXI.) Ifission a t C-0 or C-C linkages in the presence of the same catalyst.Certain structural features recognisably affect the ease (ie., thetemperature and rate) of fission a t these linkages and determinethe course of reaction when the two types of cleavage are in com-petition ; the results, however, apply to certain specified conditionsof reaction and cannot be assumed to afford direct evidence con-cerning the relative strengths of the linkages in various molecules.The hydrogenolysis of saturated alcohols occurs smoothly in thepresence of nickel at 250"/100--200 atm. n-Dodecyl, n-tetradecyl,n-octadecyl, and y-cyclohexylpropyl alcohols suffer almost quantit-ative scission a t the terminal carbon atoms of the molecules, accord-ing to the equation : R*CH2*OH + 2H2 -+ RH + CH, + ~ ~ 0 .2 7The reaction applies moreover to secondary alcohols such as cycko-hexanol and octan- p-01, which yield cyclohexane and %-octanerespectively, and also to diols such as decan-aK-diol and octadecan-28 R. Connor and H. Adkins, J . Arner. Chem. rsIoc., 1932, 54, 4678; A.,1933, 143.1 7 H, Adkins, C. E. Kommes, E. F. Struss, and W. Dasler, ibid., 1933, 55,2992; A., 936FARMER. 139apdiol, which yield octane and m-heptadecane respectively. When,however, the vapours of saturated alcohols are submitted alone tothe action of a copper-chromium oxide catalyst, reaction mayfollow one or other of two courses according to the existing con-ditions of temperature and pressure Dehydrogenation (reactionA) and a little dehydration (reaction B) occur a t 300°/1 atm.;whilst at 300400'"/100-300 atm.the aldehydes formed by de-hydrogenation suffer either condensation (Le., aldolisation, followedby dehydration) (reaction C) or ester formation (reaction D);even a t 1 atm. pressure ester formation may become important.CHPra:CEt*CHO + H,O(A*) (C.)PPCHO + H,B ~ ~ ~ Q H ~ 2Pra.CHOfLC,H, + H,Q ' Pra*C02Bua(B.1 (D.1The formation of aldols from dehydrogenated alcohols is doubtlessrelated to the appearance of aldol derivatives during G. T. Morgan'smethanol synthesis. I n the production of methanol from carbonmonoxide and hydrogen under high pressure by a chromium-manganese oxide catalyst the formation of higher alcohols becomesappreciable if the catalyst contains traces of the alkaline precipitant.The extent of such formation becomes enhanced if catalysts con-taining rubidium and czesium hydroxides are employed, and with achromium-manganese oxide catalyst containing 15% of rubidium,42% of the carbon is converted into methanol, 38% into otheralcohols, and 15 yo into carbonyl compounds and derived acetals.28Probably the higher alcohols are formed by the aldolisation of simplealdehydes, followed by dehydration and hydrogenation ; but quiteapart from the correctness or otherwise of this view it is clear thataldols and their dehydration product,s offer, under available con-ditions of temperature and pressure control, attractive startingmaterials for the production of a wide variety of alcohols, glycols,aldehydes, and ketones which are not readily obtainable by othermeans.29A remarkable result recently reported concerns the reduction ofketones by the aid of a platinum-charcoal catalyst which is furtheractivated with palladium.30 Acetophenone and its homologues inits presence are reduced to ethylbenzene and its homologues(respectively). The carbonyl group, however, of cyclopentanone,28 G. T.Morgan, D. V. N. Hardy, and R. A. Procter, Chem. and Ind., 1932,29 G. T. Morgan and D. V. N. Hardy, ibid., 1933, 618; A., 809.30 N. D. Zelinski, K. Packendorff, and L. Leder-Packendorff, Ber., 1933,51, 1; B., 1932, 250.66, [B], 872; A., 715140 ORGANIC CHEMISTRY.--PART I.cyclohexanone, hexahydroacetophenone and its homologues, and ofketones in which it is separated from the benzene nucleus by atleast one methylene group, remains unaffected.It thus becomespossible to separate aromatic ketones from cyclic ketones or thosewhich do not contain the carbonyl group in the cx-position to thebenzene ring. The reduction of the carbonyl group precedes thatof the hydroxyl group in benzoin (Ph*CO*CHPh*QH --+OH*CHPh*CHPh*OH --+ CH,Ph*CH,Ph),30 and the rate ofreduction of the carbonyl group to >CH*OE in o-acetyl esters ofthe type CH,*CO*[CH,],*CO,Et (n = 1-5) decreases with increasein the value of n (platinum ~ a t a l y s t ) , ~ ~ whilst that of l m d i c esters,CH3*CO*[CH,],C0,R, decreases with increase in the size of R.32By employment of a palladium-charcoal catalyst in the reduction ofdibasic acid chlorides, various yields of aldehydo-monocarboxylicacids are obtained.33Reductive fission of a C-N linkage in certain quaternaryammonium salts has been observed t o occur in the presence of apalladium-charcoal catalyst suspended in acetic acid : thus phenyl-benzyldimethylammonium chloride yields toluene and dimethyl-aniline.34 Another type of reductive fission of great practicalinterest is that involved in the decomposition of ozonides by hydro-gen in the presence of a palladium-calcium carbonate catalyst.For many ozonides this method is free from the disadvantagesattending decomposition by water or zinc and acetic acid, and itaffords an elegant means of obtaining good yields of aldehydesand dialdehydes from the ozonides of mono- and di-olefinic com-pounds respectively, provided that suitable conditions of ozonis-ation are employed and the temperature during hydrogenation iskept reasonablyCarbocyclic esters may be reduced either to the alcohol or to thehydrocarbon stage (RCO-i-0Me --+ R*CH,I,-OH --+ R-CM,).Reduction to the latter stage may be effected directly by using amixture of two catalysts (vix., copper-chrornium oxide and nickel) ;but it is best to conduct the operation in two stages, since the waterformed in the second stage prevents completion of the &st stage.3631 E.J. Lease and S. M. McElvain, J . Amr. Chenz. Soc., 1933, 55, 806;89 R.W. Thomas, H. A. Schuette, and M. A. Cowley, ibid., 1931,53, 3861;33 N. Froschl, A. Maier, and A. Heuberger, Monatsh., 1933, 59, 25.6; A.,34 H. Emde, Helv. Chim. Acta, 1932, 15, 1330; A , , 1933, 55.35 F. G . Fischer, H. Dull, and L. Ertel, Ber., 1932, 65, [B], 1467; A., 1932,36 B. Wojcik and H. Adkins, J. Amer. Chent. SOC., 1933, 55, 1293; A., 484.A,, 376.A,, 1931, 1397.1932, 499.11 13FARMER. 141Reduction of ethyl phenylacetate in the presence of a copper-chromium oxide catalyst at 250" yields both the alcoholPh*CH,*CM,*BH and the hydrocarbon Ph*CH,*CH, :37 as the estergradually disappears from the reaction mixture, the proportion ofalcohol to hydrocarbon present in the latter decreases--a result tobe expected if the ratio of the products depends only on the relativeconcentrations of the ester and the alcohol in the mixture. Nomodification of the catalyst has resulted in the attainment ofpreferential hydrogenation of the carbethoxyl group as comparedwith the carbinol group derived therefrom; on the other hand,various substances are found to exercise a deactivating influenceon the catalyst, thereby lowering the proportion of alcohol produced.The best yield of alcohol is obtained by interrupting hydrogenationat the point a t which 5--lOY0 of unchanged ester remains, but thehighest degree of conversion of ester into alcohol (80%) is securedin the earlier stages of the reaction.If desired, reduction of thecarbethoxyl group in ethyl phenylacetate and analogous estersmay be preceded by hydrogenation of the nucleus : in this case anickel catalyst is first employed and subsequently a copper-chromium oxide catalyst.I n aromatic esters the position of thephenyl group has a marked effect on the ease of scission of the C-0group, but specific effects of this nature disappear wi$h hydrogen-ation of the nucleus.Amines are obtainable by the catalytic reduction of oximes(> C:NQIi -+ >CONK,). W. H. Hartung and his collaboratorsreported the preparation of bemylamine and various other aminesby hydrogenation of the appropriate oximes in hydrochloric acidsolution in the presence of palladium,38 but no such reduction hashitherto been satisfactorily effected with a nickel catalyst. Goodyields of primary arnines are, however, reported as procurable fromoximes by employing a nickel-kieselguhr catalyst F9 and in somecases by the use of platinum black.40 Mandelonitriles are smoothlyreduced to amines in the presence of platinum and a slight excessof hydrochloric acid.41Extensive hydrogenolysis of glucose, sorbitol, mannitol, sucrose,37 E.Adkins, B. Wojcik, and L. TV. Covert, J . Arner. Chern. SOC., 1933, 55,1669 ; A., 604.3 8 Ibid., 1928, 50, 3370; A., 1929, 184; ibkZ., 1929, 51, 2262; d., 1929,1066; ibid., 1930, 52, 3317; A., 1930, 1286; ibkZ., 1931, 53, 2248; A., 1931,1057.39 C. F. Winans and €1. Adkios, ibid., 1933, 55, 2051 ; A., 700.40 M. IGajEinovib and D. Vranjicnn, Bull. SOC. chim., 1933, [iv], 53, 145;A.., 595.4 1 J. S. Buck, J . Amer. Chem.SOC., 1933, 55, 2593, 3388; A., 821, 1049.Compare W. H. Hartung, ibid., 1928, 50, 3370; A., 1929, 184142 ORGANIC CHEMISTRY.-PART I.lactose, maltose, and related compounds occurs when these sub-stances are hydrogenated over copper-chromium oxide at 250"/300atm., the products being methyl and ethyl alcohols, propane-@-diol, and various hydroxy- h e ~ a n e s . ~ ~ Hydrogenolytic fission ofvarious ring systems can also be readily effected in the presence ofplatinum-charcoal at 300°.43 I n contrast to this, all the resourcesof technique and equipment which have been applied to thcreduction of diaminodimesityl and 6 : 6'-dimethoxydiphenic acidhave failed to bring about reduction.44 It is possible that thesame factors which prevent the rotation of the rings with respectto each other in these diphenyl derivatives also prevent adsorptionof the benzenoid compound by the catalyst and thus inhibit thefirst step of the reduction process.With regard to selectivity in hydrogenation, recent experimentsshow beyond any doubt that, in those cases in which two or morecourses of reaction are offered, a large degree of selective discriminationmay be exercised by the catalyst. Two main categories of selectivephenomena are to be distinguished : (1) relating to examples inwhich mixtures of different mono-olefins, or mixtures of two or moresubstances displaying different types of unsaturation (e.g. , mono-olefins, conjugated di- or poly-olefins, acetylenes, allenes), undergohydrogenation; and (2) relating to examples in which differenttypes or kinds of unsaturation or reducibility occur together withinthe molecule of substance to be reduced.With regard to the f i s tof these groups it was suggested by S. V. Lebedev and his collabor-as early as 1925 that the main factor which determinespreference of reduction (as also rate of reduction of the individualcompounds) in mixtures of mono-olefins is the degree of substitution :thus mono-substituted olefinic compounds, whether hydrocarbonsor not, are reduced in preference to di-, di- to tri-, and tri- t o tetra-substituted compounds ; and further, mono-, &-, tri-, and tetra-substituted compounds are reduced at rates which diminish for therespective typcs in the order cited. Further observations byLebedev and A.0. Yakubchik indicated that butadienes are reducedin preference to mono-olefinic s~bstances.~~ It is pointed out by3. H. Farmer and R. A. E. Galley 47 that the principle of consecutive(selective) reduction here embodied is rigidly observed in (a) various42 W. H. Zartman and H. Adkins, J . Arner. Chem. Soc., 1933, 55, 4562.43 N. D. Zelinski, B. A. Kazanski, and A. F. Plate, Ber., 1933, 66, [B], 1415;A . , 1150.44 C. R. Waldeland, W. H. Zartman, and H. Adkins, J . Amer. Chem. Soc.,1933, 55, 4234; A., 1294.4 5 J., 1925, 127, 418; A., 1925, i, 350.46 Ibid., 1928, 823, 2190; A., 1928, 613, 1111.4 7 Ibi&., 1933, 687; A,, 936FARMER. 143binary mixtures of An-, AD-, and A"-olefinic acids which theyexamined and ( b ) mixtures of butadienoid and mono-olefinic acids.There is, nevertheless, good reason to challenge the generalisationthat the order of preferential reduetion in mixtures of olefinic sub-stances necessarily runs parallel to the relative speeds of reductionof the pure components, since in mixtures of ally1 alcohol andAa-hexenoic acid 47 and of pinene and cinnamic acid48 the com-ponent which is reduced the more slowly when each is taken separ-ately is preferentially reduced in the binary mixture.I n conjugated butadiene-cc-carboxylic esters of the sorbic acidseries, reduction may occur selectively a t one or other of the doublebonds or a t the terminals of the butadienoid system.With platinumat room temperature and atmospheric pressure, the dominant modeof reduction for sorbic acid (XXII) is unselective as regards theap- and 78-double bonds of the butadiene system; but such selec-tivity as exists reveals the 78-double bond as much more prone toattack than is the orp-double bor1d.47,49 With an active nickelcatalyst, however, recent observations of the reporter 50 show thatpractically quantitative reduction of sorbic acid to the dihydro-stage can be achieved at room temperature and atmosphericpressure; and in the course of reduction the 78-double bond is byfar the most extensively attacked, but some cc8- and @-reductionCHMe:CH*CH:CH*CQ,H CIP:CH*CO,Na CH:CH*CO,NaCH,*CH,*CO,Na CH,Me*CH,CH:CH*CO,H CH:CH*CO,Naalso occurs.I n the case of sodium muconate (XXIII), where botholefinic centres are of orp-type, 50% hydrogenation gives rise to only4 8 H.Adkins,F. F. Diwoky, and A. E. Broderick, J. Amer. Chem. SOC.,1929, 51, 3418; A., 1930, 40.49 The scheme of reaction is considered to be the following,Conjugated acid /".- h Y ?-3 1 (Nil(XXIII. ) (main product)I ( ~ ~ 1 1 . ) E2pi)(main product),, (a,!?-reduction) ---+ Saturated mid i Aa-Dihydro-acid (yS-reduction) --++AD- ,, ,, (&reduction) -+(a,!?yS-reduction)\ - -3in which complete disappearance of conjugated material is achieved before theethylenic products begin to suffer reduction. Here there are two types ofreaction in competition at the outset, w ~ z . , A and B in the scheme : Dihydro-acid f- Conjugated acid 4 Saturated acid (aPy8-addition); reduction ofthe dihydro-acids does not compete with these.The cam differs from thatof acetylenic compounds (see below).A B50 Unpublished144 ORGANIC CHEMISTRY .-PART I.a 50% yield (m.) of dihydro-derivative, which includes some pro-portion of the Ap-form. Thus the dcgree of selectivity exercisedby a platinum catalyst (which becomes enhanced by '' ageing " ofthe catalyst) 47 is greatly inferior, under the conditions of experiment-ation employed, to that exercised by an active nickel catalyst, butto the extent that selective discrimination does occur, the mostreadily (although not solely) attacked centre in a conjugatedbutadiene-a-carboxylic acid is for both platinum and nickel they8-double bond. The appearance of marked selective action inpoly-olefinic systems has obvious synthetic and diagnosticapplications.The difference in catalytic action between platinum and nickelis shown in the hardening of arachis and other oils containing poly-olefinic components.Platinum tends at the ordinary temperatureand atmospheric pressure to promote the complete saturation of theunsaturated molecules of these substances ; nickel, on the otherhand, tends a t about 180" and atmospheric pressure to promoteselective hydrogenation, resulting in the production of glycerides ofoleic acid or of its i~ornerides.~~ A quite similar difference to theforegoing, however, appears between the products obtained byhardening in the presence of nickel at low pressures and hightemperatures and those obtained with nickel a t very high pressures(100-300 atm.) and the ordinary temperature : the former tend tobe composed of mono-olefink glycerides agd the latter of fullysaturated glycerides. 52In the case of certain acetylenes S.V. Lebedev and V. J. Schtern 53report that hydrogenation in the presence of platinum yields ethyl-enes (CR{CH 3 CHIt:c'H,), but these suffer further hydrogenationimmediately subsequent to their formation at rates which dependon the nature of the substituent group (R) and on the relativeconcentrations of the aceeylenic and ethylenie components presentin the reduction mixture at the time. According to J. S. Salkindand his collaborators 54 the process of hydrogenation of acetylenesproceeds in two stages (c% -+ CH-CH -+ CH2*6H,) andthe relative speeds of the two consecutive reactions vary veryconsiderably with the nature of the substitution in the originalacetylenic system.Thus in presence of palladium the first stage of51 H. I. Waterman, J. A. van Dijk, and C. van Vlodrop, Rec. truv. chim.,1932, 51, 653; A., 1932, 1018.52 H. I. Waterman and (Frl.)M. Zaayer, ibid., p. 401: A., 1932, 601. Coin-pare H. P. Kaufmann and E. H. Schmidt, Ber., 1927, 60, [B], 50; B., 1927,226.53 J . Gen. Chenz. (U.X.S.R.), 1932, 2, 249; A., 1932, 1231.54 Ibid., 1933, 3, 91; A., 805FARMER. 145reduction may proceed more or less rapidly than, or a t the samerate as, the second, and the result is independent of tho mode ofpreparation and the state of dispersion of the catalyst and of thekind of support-material employed. The relative yield of theethylenic product of reaction may apparently depend on the velocityof hydrogenation, which in turn is stated to depend on the con-centration of the catalyst and on the nature of the solvent employed.Analogous results are reported for platinum, but it is claimed thatpalladium has, and platinum has not, a special affinity for triplelinkages.Natural Polyene Pigments (Lipochromes) and Vitamin A .During the past year, owing mainly to the researches of P.Karrerand R. Kuhn and their numerous collaboraDors, striking progresshas been made in the elucidation of the constitutions of a number ofnaturally occurring polyene pigments belonging to the group ofcaroten~ids.~~ Pigments of polyene type are widely distributed inthe vegetable and animal kingdoms and the group comprises hydro-carbons, alcohols (xanthophylls, phytoxanthins), ketones, carboxylicacids, and carboxylic esters.In most of these substances there occurs a conjugated C,,-chaincomposed of two double-isoprene skeletons linked together by adouble bond (*) to form the system (XXIV).:CH-CMe: CH*CH:CH*CMe:CH*CH~CH*CH:eMe*CltT:CH*CH:CMe*CPI:-v ~Fragment A.(XXIV. )This system, being symmetrical about the doable band, does notconsist merely of isoprene units joined end to end, but may beregarded as the middle portion of a still larger symmetrical systemcontaining the fused carbon skeletons 01 two phytol molecules.This larger system is the one which OCCUFS in open-chain form inlycopene (XXV), and in semi-cyclic form in p-carotene (XXVI):it incorporates at each end the C,, carbon framework belonging toa- or @-ionone, whilst its middle portion is composed of the conjug-ated straight-chain fragment designated by A in formula (XXIV).For convenience in describing the natural polyencs the systemderived from A by interchanging the positions of the double andsingle linliings (Le., the system which occurs in the terminallyreduced derivatives of substances containing A, and in at least55 For the history of the carotenoids and the early work on their isolationand characterisation, the reader is referred t o P.Karrer, 2. angew. Chcm.,1929, 42, 918 ; R. Willstatter and Stoll, “ Untersuchungen uber Chlorophyll,”Berlin, 1913; L.S. Palmer, “ Carotinoids and Related Pigments,” New York,1922146 ORGANIC CHEMISTRY .-PART I.one of the naturally occurring polyenes) is designated B, the satur-ated system derived by complete hydrogenation of A or B is,/CH,-CHMe,\CH,-C€lMe,CH2-[CH,]2*CHMe*[CH2]3*CMe:C€I-CH2*OHPhytolCH2H2 Me,/-\-CH:CH*COMeH”\_//H, Me 8-Ionone/CH=CMe, CMe,=CH\Lycopene (XXV.) \CMe-CH,/ \CI€,-CMc’CE.1-CH: CH*CMe=AyCMe *CH: CH-CH CH2 CH2H2 Me, Me2 H2\-/ H,/-\-CH:CB*CMe=A=CMe*CH:CH- O H 2Me H, H, Me p-Carotene (XXVI.)designated C, and the trimethylcyclohexene rings of p- and a-ionone /J<--L a n d c =/---- respectively . are symbolically represented byThus :B = *CH:CH*CH:C~e*CIP:C~*CH:C~~CMe:CH*CH:~H*and C = *[CH,],*CHMe~[CH,],*CHMe~[CH,],*Other natural polyenes contain modifications of the carbonsystem present in p-carotene, such modifications including hydr-oxylation or ketonisation of the ionone rings, curtailment of thechain, and replacement of the A-system of conjugation by theB -system.Bixin and Crocetin.-Last year it was reported 56 that Y.Kuhnand A. Winterstein had suggested a symmetrical formula for bixin,the yellow pigment of the seeds of annatto (Rim orellam), in placeof an earlier formula in which the four methyl substituents knownto be present in the chain had been assigned unsymmetricalpositions in order to allow for the reported non-equivalence of thecarboxyl groups. At that time, however, the positions occupiedby the methyl groups had not been demonstrated, but the newformulation was of interest in revealing the bixin chain as con-stituting the middle fragment of the lycopene molecule accordingto P.Karrer’s formulation (XXV) 57 of the latter.HO,C*CH:CH*CMe:A:CMe*CH:CH*CO,Me(XXVII.) Bixin (K. and W.)H02C*CMe:A:CMe-C021?I (XXVIII.) Crocetin.\/ \ \56 Ann. Reports, 1932, 29, 120.67 Hclv. Chim. Acta, 1930,13,1084; 1931,112,435; A., 1930,1422; 1931,697FBRMER. 147The correctness of the position assigned to the methyl groups, andthe symmetry of (chain) constitut’ion, have now been demonstratedby the synthesis of perhydronorbixin (XXIX) 58 by the stages :-CMeNa(C 0,E t) Reduction Br[CH2I3Br --- -$ H02C*CHMefC€12]3*CHMe*C02H ofeeter +HO2C~[CH212~CH~e~[CH,I3~CHMe*[CH212*CO,H gEtsaltOl.) 3(Hydrol.)PBr,; CHNa(CO,Et), CH,( OH)*CHMe-[CH2],*CHMe*CH2*OH --(zol.)+Electrolysis ofH02C*[CH2],*CHMe*C*CHMe*[CH2],*C02HNot only is the synthetic product identical with that derived bythe hydrolysis of catalytically reduced natural bixin, but it is foundthat when the terminal carboxyl groups of perhydronorbixin aredegradatively removed by the ingenious procedure represented inthe stagesPerhydronorbixin (XXIX.)P b( 0 Ac) R*CH( OH)*CMe,*OH ~ -$- RCHO + Me2C0the product is a dialdehyde, showing clearly that no methyl groupsare attached to the a- and or’-carbon atoms of bixin and perhydro-norbixin. In this respect perhydronorbixin differs from per-hydrocrocetin, since by the same procedure the latter gives adiketone, C18H3402.To crocetin, the pigment derived from Crocus sativus (saffron),the formula C19H2204 was originally assigned,59 but later this wasreplaced by the formula CzoH2404 in which four of the carbonatoms were almost certainly present as methyl groups.60 The factthat at least one of the four methyl groups is attached to an or-carbonatom was established by comparative oxidations of dihydro-(terminally reduced)-crocetin and dihydronorbixin, but all un-certainty as to orientation in the carbon chain was dispelled bythe production of the above-mentioned diketone, C18H3402 (actuallyMe*CO*[CH2]3*CHMe*[CH2]4*CHMe*[CH2]3*CO*Me), and by the syn-thesis of perhydrocrocetin, HO,C*CHMe* C*CHMe*CO,H, from5 * P.Karrer, P. Benz, R. Morf, H. Raudnitz, M. Stoll, and T.Takahashi,Helv. Chim. Acta, 1932,15, 1218; A., 1932, 1234; ibid., p. 1399; A., 1933, 52.It is remarkable that the synthetic acid is obtained in the same stereo-form asthe perhydrogenated acid derived from natural bixin, although the latterpossesses no fewer than four asymmetric carbon atoms and can exist in sixinactive forms.59 P. Karrer and H. Salomon, ibid., 1928, 11, 711; A., 1928, 869; R. Kuhn,A. Winterstein, and W. Wiegand, ibid., p. 716; A., 1928, 869.60 R. Kuhn and F. L’Orsat, Ber., 1931, 64, [B], 1732148 ORGANIC CHEMISTRY.-PART I.pC-dimethylheptane-ctq-diol61 in a manner somewhat similar tothat of perhydronorbixin from trimethylene bromide. Any furtherevidence necessary t o the complete establishment of the relation-ship existing between crocetin and bixin has been amply supplied(1) by the building up of perhydronorbixin synthetically from theester of perhydrocrocetin 62 and (2) by t)he degradation of perhydro-norbixin to perhydrocrocetin : 63 the former process involves thelengthening of the perhydrocrocetin chain by four CH,-groups, andthe latter the reverse transformation.The cis-trans nature of the isomerism existing between bixin(E-) and 13-bixin was referred to in last year's Report; anothervariety of bixin, &., isobixin, reported by S.F. B. van Nasselt 64to result from exchange of the esterified and the unesterified carb-oxyl group, has not been obtainable by subsequentAn isomeride of crocetin has now been isolated from saffron, andthis on account of the extreme ease with which it changes intocrocetin and the fact that its dihydro-derivative is identical withdihydrocrocetin, is regarded as a geometrical isomeride of crocetincon%aining one or more cis double bonds.66 Since ordinary bixin(IX-) and crocetin must now be considered to possess cis- and trans-configurations respectively and p-bixin a trans-configuration, therewould be no rational significance in designating the new cis-isomeride of crocetin p-crocetin : indeed the prefixes ct, p-, and y-have been employed in the past to distinguish between free crocetinand its monomethyl (m.p. 218') and dirnethyl (m. p. 221") esters.67Accordingly the use of the terms stable and labile (or the numbers1 and 11) in describing the isomeric bixins and crocetins is suggested.B1.p. Geomot.form. Variety.,%pixin ........................ 2 SO" trans stable (I)Bixin (ordinary) ............ 196 C i S labile (11)a-Crocetin ..................... 285 truns stable (I)Crocetin (new) ............... 141 CiS labile (11)Although crocetin and bixin represent the middle sections of themolecules of carotene and lycopene, the suggestion that theyare derived in the plant by the oxidative degradation of carotenoidsubstances 68 has not remained un~riticised.~~ Interesting evidenceP. Karrer, F. Bern, and M. Stoll, Helv. Chim. Acta, 1933,16, 297; A., 594.6% P. Karrer and F. Benz, ibid., p. 337; A., 594.63 H. Raudnitz and J. Peschel, Ber., 1933, 66, [B], 901; A . , 807.64 Rec. truv. chim., 1911, 30, 1 ; A., 1911, i, 550.6 5 P.Karrer and T. Takahashi, Helv. Chim. Acta, 1933, 16, 287; A., 594.G6 R. Kulin and A. Winterstein, Ber., 1933, 66, [B], 209; A., 258.6 7 P. Karrer and H. Salomon, Helv. Chim. Acta, 1927, 10, 397; A., 1927,G8 R. Kuhn and C. Grundmann, Ber., 1932,65, [B], 1880; A., 1033, 142.5 7 1 ; ibid., 1928,11, 513; A,, 1928, 644FARMER.. 149on this point is afforded by the recent discovery that picrocrocin(C16H260,), the bitter principle of saffron, yields on hydrolysisequimolecular quantities of saffronal (a cyclic aldehyde of con-stitution XXX) and glucose, so that picrocrocin i s cz (P-)glucosideof hydrated saffronal, probably having the constitution (XXXI).69I3 M,Me,H, Me, €,I ?H? 31 &<‘WHO / -//H/-\CHQ OH.CH,.c-~-(i--(i-- C\14: W,Rfe \A,(XXX.) L O - A (XXXI. )I HO H OH( I41 MeSaffronal PicrocrocinNow in view of the facts (1) that in saffron large proportions ofglycosidically-linked trans- and cis-crocetin occur in the form ofcrocin side by side with smaller proportions of lycopone, p- andy-carotene and zeaxanthin, and (2) that the framework of 2 mols.of saffronal and 1 mol.of crocetin together make up the frameworkof a dicyclic C,,-carotenoid pigment, it appears that such &cycliccarotenoids suffer oxidative scission of the terminal ring systemsand that the two varieties of scission product SO formed becomeglucosidically combined t o yield picrocrocin and crocin respec-tively. This hypothesis is fortified by the observation that theproportion of these two substances in fresh saffron is 1.4 : 1.Lycopene and the Carotenes.-Lycopene, which occurs as a redcolouring matter in the berries of bitter-sweet and of other plants,and in rose-hips and tomatoes, has the empirical formula C,,HC,,.70Since it contains thirteen double bonds, as indicated by hydrogen-ation and halogenation tests, I?.Karrer and his collaboratorsproposed the formula (XXXII) on the assumption that i t s doubleCMe,:@H.[CH,],~CMe=@H.CH~C€I*CMo:A:C~~e.CI-P:CH.~~=C~e.[CH ] .CHI I (XXXII.) CMe,1 1 L ycopenal CMe,.1 CMe,:CH. [CH,] ,CO.Me + CKO.CI-P:CN.C~~e:A:CMe.CII: :CH.CH=CMe. [ CH,],. CHMethylheptenoneMethylheptenone4 CHO.CH:C~H.CMe:A:C~le.CH:CH.CHO + C8HI40Bixin dialdehyck.1 HO,C. CH: CH. CMe:A: CMe- CIX CH- C0,Hfl-2Vorbixi.n69 R.Kuhn and A. Winterstein, Natwwiss., 1933, 28, 527 ; A., 954.70 R. Willstatter and H. H. Escher, 2. physiol. Chem., 1910, 64, 47; A.,1910, i, 330150 ORGANIC CHEMISTRY.-PART I.bonds were symmetrically disposed and that it was built up fromtwo phytol residues.71 The fission of &he lycopene chain by chromicacid to yield first lycopenal 72 and then bixin dialdehyde,6* the latterof which is convertible into the corresponding dibasic acid, estab-lishes a t once the correctness of Karrer's formula, since the dibasicacid proves to be identical with the hydrolysed geometrical isomerideof ordinary bixin, i.e., p-norbixin or cis-norbixin.Carotene is obtained from numerous sources : optically activepreparations may be obtained from carrot, palm oil, #orbus uucu-paria, Aesculus hippocastanurn, and inactive preparations fromwinter spinach, grass, and stinging nettle.73 Carotene has beenfound t o contain three distinct substances, of which two (a- andp-carotene) contain eleven double bonds and, like vitamin A,promote growth.a-Carotene, m. p. 187" (corr.), is optically activeand can be completely separated from p-carotene by filtration ofits ligroin solution through calcium hydroxide or lime, the p-formbeing adsorbed in the dark reddish-brown upper layer and thea-form in the lower yellow layer (Tswett's method of chromato-graphic analysis). The two forms differ in melting point, solu-bility, and optical activity and in the position of their absorptionNow, since a-carotene yields by ozonolysis both geronicacid and isogeronic acid, its constitution is without doubt correctlyrepresented by (XXXIII).75/CMe,-CH,\~ : C H .C ~ e ~ * = C M e * ~ ~ : C H r y 2 HO,C/ CH2a-Carotene (XXXIII.) isoGeronic acidp- Carot ene (XXXIV. ) Geronic acidFor p-carotene, m. p. 183" (corr.), Karrer and his collaboratorsproposed the symmetrical formula (XXXIV), although a t the time71 Ann. Reports, 1932, 29, 122.72 R. Kuhn and C. Grundmann, Ber., 1932, 65, [B], 898; A., 1932, 749.73 R. Kuhn and E. Lederer, Ber., 1931, 64, [B], 1349; A,, 1931, 959; 2.physiol. Chern., 1931, 2Q0, 246; A., 1931, 1421.74 p. Karrer and 0. Walker, HeZv. Chim. Acta, 1933, 16, 641; A., 805; P.Karrer, 0. Walker, K. Schspp, and R. Morf, Nntwe, 1933, 132, 26; A., 805.7 5 p.Karrer, R. Morf, and 0. Walker, Helv. Chim. Acta, 1933, 16, 975; A.,1150 ; compare P. Karrer, A. Helfenstein, H. Wehrli, and A. Wettstein, ibid.,1g30,13, 1084; A., 1930, 1422; P. Karrer and R. Morf, ibid., 1931, 14, 1033;A., 1931, 1299FBRMER. 151less than one molecular proportion of geronic acid had been derivedfrom it by oxidation.76 Careful thermal degradation of P-caroteneyielded toluene and m-xylene, which doubtless arose by cyclisationof fragments of the polyene chain (:CN-CH:CH*CMe:CH*CH: and:CH*CMe:@H*CH:CH*CMe:). It was known, however, that zea-xanthin, whose constitution (according to the evidence available)differed from the proposed one for P-carotene oiily by the presenceof hydroxyl groups in the ionone rings, yielded not only toluene andm-xylene but also a small proportion of 2 : 6-dimethylnaphthalene.@-Carotene, on testing, was found also to yield 2 : 6-dimethyl-naphthalene.But there is plain evidence that vitamin A, whichstands biologically and chemically close to @-carotene (see p. 159),yields by cyclisation and dehydrogenation, not 2 : 6-, but 1 : 6-di-methylnaphthalene ; furthermore, the 1 : 6-compound might wellbe expected to arise by the cyclisation and dehydrogenation of aF-ionone ring and its adjacent carbon atoms. Consequently theformation of 2 : 6-dimethylnaphthalene by thermal degradationprobably arises in the mannerc \/(/)-(C4H6)-cH CH\ // ,A\/ //CH ?Me _3 //v\MeC CH CH I II f /\n/ CH CH--(C4H6)--/-and this probability is rendered almost a certainty by the fact thattrans-crocetin, which contains no carbon rings, has now been foundto yield 2 : G-dimethylnaphthalene.I n this case a large part (12carbon atoms) of the conjugated chain of @-carotene must bejdentical (although not necessarily as regards geometrical con-figuration) with that of crocetin.Further striking testimony to the correctness of Karrer's formul-ation of @-carotene is supplied by the fact that the substance canbe oxidised in two stages to yield first a diketone (XXXV) by thefission of one ring 77 and then a tetraketone (XXXVI) by the fissionof both.'* The tetraketone yields with zinc and acetic acid a di-Thecourse of oxidation in polyenes depends on the number of double bonds, thelength of chain, and other not yet understood circumstances, and the quantita-tive evidence afforded in the case of 8-carotene is insufficiently significant topreclude the possibility that the molecule has similar ends.\ '> /76 R.Kuhn and A. Winterstein, Ber., 1933, 66, [B], 429; A., 387.7 7 R. Kuhn and H. Brockmann, Ber., 1933,66, [B], 1319; A., 1297.7 8 Idem, &id., 1932, 65, [B], 894; A., 1932, 749; 2. physiol. Chem., 1932,The preferential attack of oxidising agents a t the 213, 1 ; A., 1933, 195.terminal double bonds of the carotene system is a remarkable feature152 ORGANIC CIIEM1STRY.-PART I.hydro-derivative (XXXVII) which closely resembles dihydrorhodo-xanthin (p. 156) in its eiiolic properties. Like p-carotene, butc)-(c22H26)-c*'L' \/ \/Me Me co-/ MeSemi-p-carotinone (XXXV.) p- Carotinone (XXXVI.)M e 4 0 0 [CH,] ,~CMe,~CO~CH,~CH:CMe-B-CMe:CH.CH,.CO.CMe~.[CH,],~COMe(XXXVII.)unlike p-carotinone, semi- @-carotinone shows strong growth-promot-ing action, as might be expected if, as R. Kuhn and H. Brockmannbelieve, 1 mol. of @-carotene is the precursor of 2 mols. of vitamin Ain the animal body; 79 for, if only one ring system of p-carotenesuffers fission, 1 mol. of vitamin A should still be capable of arising(C40H56 + 2H,O -> 2C2,H,,0). In this connexion it may bementioned that a @-carotene monoxide,*O prepared by the action ofperbenzoic acid on p-carotene, and a hydroxy-p-carotene (C,0H5,03),obtained by the action of chromic acid on @-carotene,78 bothpromote growth, and there is little doubt that in the formation ofboth derivatives only one ring system has been attacked.y-Carotene, m.p. 178" (corr.), which appears to constitute aboutone thousandth part of ordinary carotene,sl is successfully separatedfrom the a- and the p-form by adsorption on aluminium oxide. Like@-carotene, it is optically inactive, but it differs from both the otherforms in taking up, not eleven, but twelve molecules of hydrogenand in yielding on degradation only 0-85 mol. of acetone. Invarious properties (especially the position of its absorption bands)it takes a position between @-carotene and lycopene and its structurei s best represented by (XXXVIII), that is, as containing one of theopened-ionone rings characteristic of lycopene./ ~ ~ : ~ H * C M e : A : C i M e .C H - L< \\ / '3 y- Carotene (XXXVlIl.)y-Carotene strongly promotes growth to an extent nearly equalto that of a- or (3-carotene. Thus in hydrocarbons of the com-position C40H56 such biological activity appears not to be dependenton the number of double bonds, which may be 10, 11, or 12, ornecessarily t o be lost when the molecule i s modified; '8 an essential79 R. Kuhn and H. Brockmann, KZin. Woch., 1933,12,972; A., 1212.80 H. v. Euler, P. Karrer, and 0. Walker, HeZu. (7h.i.m. Acta, 1932, 15, 1507.81 R. Kuhn and H. Brockmann, Ber., 1933, 66, [B], 407; A., 431. Thea-, 15% ; average composition of investigated carotene preparations is :/3-, 85%; and y-, 0.1%FARMER. 153feature, however, according to R.Kuhn and H. Brockrnann,appears to be the presence of at least one carbon ring of P-iononetype, but other special conditions may possibly need fulfilling.Xanthophyll or Phytoxanthin Group (Lutein, Violaxanthin,Z euxanthin, Ph y salien, Taraxanthin, Pucoxant hin , Capsant hin,Pluvoxanthin, Kryptoxanthin) .-For the C,,-carotenoids which con-tain oxygen, P. Karrer and A. Notthafft propose the generic term“ phytoxanthins,” whilst they reserve the name “ xanthophyll ” forthe substance C40HS602 derived from green leaves ; 82 this substance,however, is designated “ lutein ” by R. Kuhn, A. Winterstein,and E. Lederer.83 The substances of the group, so far as is deter-mined, contain hydroxylated ionone rings, and frequently in theplant the hydroxyl groups are esterified by fatty acids.To lutein or leaf-xanthophyll, which not only occurs in the greenleaves and yellow petals of various plants but constitutes one of thecolouring matters of egg-yolk, P.Karrer and A. Zubrys * haveassigned the formula, (XXXIX), thus representing it to be adihydroxy- derivative of a- carotene. Since its perhydro- derivativeXanthophyll (XXXIX.) /\/ \/O=C)-[CH2]2*CHMe*C*CHMe*[CH2]2--~~\-0 v- cro,\ FL.1yields a diketone (presumably XL) on oxidation with chromicacid, there is no doubt that the hydroxyl groups are of secondarytype.Violaxanthin, which occurs in pansy, laburnum, nettles, and horscchestnut leaves, is an allied substance containing four oxygenatoms, of which three appear to be present as hydroxyl groups.The constitutions of other known compounds of the xanthophyllgroup, vix., fucoxanthin (c4$35606), flavoxanthin (C,H,,O,),capsanthin (C35H5003), and taraxanthin (C,,,H,603), have not yetbeen determined.Zeaxanthin accompanies lutein in egg-yolk and occurs with rhodo-xanthin in the ripe fruits of yew (Taxus baccata).It has recentlybeen formulated as a dihydroxy-@-carotene (XLI).85 In the redcalyces and berries of the Physalis species it occurs as its dipalmitate,to which the name physalien was originally giveh.82 Helv. Chim. Acta, 1932, 15, 1195; A., 1932, 1256.83 2. physiol. Chern., 1931, 197, 141; A., 1931, 885.84 Helv. Chim. Acta, 1933, 16, 977; A., 1150.85 R. Kuhn and H. Brochann, Ber., 1933,66, [B], 828; A,, 716154 ORGANIC CHEMISTRY .-PmT I.\/ \/OH<7--CH:CHCMe:A:CMe*CH:CH~OH -&\ Zeaxanthin (XLI.)Plavoxanthin, C40H5603, isolated from Ranunculzcs acer, 86 was thefirst xanthophyll found to contain an odd number of oxygen atoms.A second pigment of the same type, kryptoxanthin, Cp0H560,occurs in esterified condition as a main constituent of the dye inthe red calyces and berries of the Physalis species from whichphysalien was originally isolat'ed.Previously it had not beendifferentiated therein from p-carotene, which it closely resemblesand accompanies; now, however, it has been isolated by chromato-graphic analysis, but cannot be distinguished €rom p-carotenespectroscopically. It contains a hydroxyl group and, since ittakes up 11 mols.of hydrogen, without doubt contains two carbonrings. In colour, tendency to become adsorbed, and yield of aceticacid on complete oxidation with chromic acid it occupies a positionbetween p-carotene and zeaxanthin; accordingly R. Kuhn and C.Grundmann suggest a formula (XLII) comprising half the p-caroteneand half the zeaxanthin molecule. 87 I n optical activity, however,/\ ,' \/C\-CH:CH*CMe:A:CMe*CH:CH-/DOH -< \? Kryptoxanthin (XLII.)there appears to be an anomalous relationship between krypto-xanthin on the one hand and 13-carotene (inactive) and zeaxanthin(laevorotatory) on the other, for any rotation which is observableis extremely small. According to formula (XLII) kryptoxanthincontains one (unmodified) 13-carotene residue and according toKuhn and Brockmann's views should therefore show half thevitamin-A activity of p-carotene and differ from zeaxanthin, whichhas no such residue. Actually, kryptoxanthin adds to the existinglist of naturally occurring pro-vitamins (a-, p- and y-carotene) afourth example.Axafrin.-Azafrin, which gives rise to many magnificent colourreactions, especially with mineral acids, is extracted from azafranroot.88 It has been shown to be a carotenoid carboxylic acidcontaining seven conjugated double bonds and four oxygen atoms-the latter probably tertiary.The C,,-formula first suggested86 R. Kuhn and H. Brockmann, 2. physkd. Chem., 1932,213, 192 ; A., 1933,8 7 Ber., 1933, 66, [B], 1746.6 8 R. Kuhn, A. Winterstein, and H. Roth, Ber., 1931, 64, [B], 333; A.,329.1931, 492FARMER.155represented the nearest methyl group to the carboxyl group asattached t o the &carbon atom. But since azafrin yields m-xylene,m-toluic acid, and toluene on thermal degradation, this methylgroup must be attached either to the y- or to the E-carbon atom.The C,,-formula has now been replaced by a C,,-formula (C2,Hb20q)(XLIII) which agrees more closely with new analytical results andpermits the representation of the azafrin molecule as derived fromcarotene by the loss of thirteen carbon atoms (Le., one completeionone skeleton); the first methyl group then occurs in the y-po~ition.~gGeronic acid m-Xylene m-Toluic acid7 1 7 ---_H 2 8 OH)-CH:C€€-i-CMe: CH*CH:CH CMe:CH-+CH:CH*CH:CMe* CH:CHCO,HH, Me(0H) Axafrin (XLIII.)Hz Me2 .1.CO-CH:CH-CMe:CH-CH:CH*CMe:CH.CH: CH*CH:CMe*CH:CH* CO,HH 2 c M e Azafrinone (XLIV.)By careful oxidation of azafrin, azafrinone (XLIV) is produced,which takes up seven molecules of hydrogen easily and two morewith difficulty; but in this oxidation the optical activity charac-teristic of azafrin is lost and its original presence must be attributedto the asymmetry of an adjacent pair of carbon atoms (* * in XLIII)in the carbon ring.Azafrin which has been reduced to the pointof saturation of its seven ethylenic bonds yields on oxidation withlead tetra-acetate a compound identical with reduced (tetra-decahydro-)azafrinone ; moreover, azafrin yields with perma>nganategeronic acid. Thus it is evident that the hydroxyl groups of azafrinare differently placed from those which form the characteristicfeature of the xanthophyll group; indeed azafrin is a di-tertiaryglycol in which the hydroxyl groups probably occupy trans-positions.The methyl ester of azafrin, like that of various conjugated di-estersand di-ketones (not mono-), is capable of smooth reduction by zincand acetic acid to the dihydro-stage ; and like the dihydro-derivativesof crocetin, bixin, rhodoxanthin, etc., the sodium or potassiumderivative is intensely coloured and reverts to the parent polyeneby the action of air.Rhodoxanthin.-The red fruits of the yew (Taxus bacmta) yielda very stable blue-black crystalline pigment, CMH,oO,.Thepoverty in hydrogen shown by this formula agrees with the presenceof an exceptionally large number of double bonds in the azafrin89 R.Kuhn and A. Deutsch, Ber., 1933, 66, [B], 883; A., 711156 ORGANIC CHEMISTRY.-PART I.molecule-a characteristic reflected in the especially long-waveposition of the absorption bands. The pigment contains no esterifi-able hydroxyl groups, but yields a dioxime ; moreover, spectro-graphic evidence indicates that the oxime-forming carbonyl groupsare conjugated with the unsaturated chain of the molecule andchemical evidence shows that rhodoxarithin represents a new typeof polyene pigment, viz., a polyene ketone.gQRhodoxsnthin is the most unsaturated of the known carotenoidsand takes up easily twelve molecules of hydrogen and with greaterdifficulty a further two molecules.This behaviour correspondsto the saturation of twelve ethylenic linkages, followed by thereduction of two carbonyl groups ; and since the saturated diketoneproduced in the first of these stages has the composition C40H,402and the saturated diol finally produced has the compositionC40H,802, two carbon rings must be present as represenked in (XLV).O = ( - ) = C H * C H : C ~ ~ e * ~ * C ~ ~ ~ : C ~ n C ~ ~ = C \ - - o - L =/- --+ 2 I1\/ \// Rhodoxanthin (XLV.)<>=o -+/ Dihyclrorhodoxanthin (XLVI.)DisodiodihydrorhodoxantJbin (XLVII.)The passage from rhodoxanthin to its dihydro-derivative (XLVP)is marked by a lightening in colour as in the case of bixiii andcrocetin. Spectroscopically the dih ydro-compound resembles p-carotene and zeaxanthin to the point of confusion; and since thedihydro-compound and its dioxime agree with one another spectro-scopically, it can be deduced that the addition of two hydrogenatoms has interrupted the conjugat'ion of the carbonyl groups withthe polyene chain.Like dihydrocrocetin dimethyl ester, methyl-dihydrobixin?l and the above-mentioned dihydro-P-carotenone,dihydrorhodoxant hin gives with alcoholic soda an intensely colouredsodio-derivative (XLVII), which reverts by the action of air to t)heparent polyene.Vitamin A .-Certain naturally occurring polyene pigments,especially @-carotene, are capable of promoting growth in animals,thus acting as sources of vitamin A. Those carotenoids whichcontain one or two unmodified @-ionone rings appear to suffer90 R.Kuhn and H. Brockniann, Ber., 1933, 66, [B], 828; A., 716.91 Ann. Repwts, 1932, 29, 121FARMEB. 157degradation in the animal organism to vitamin A (see p. 152).Vitamin A itself is found in the liver and blood serum of animalsand is reported to occur. in butter; 92 the richest concentrations,however, are found in the livers of sea-fish, but there is an enormousdifference in the vitamin A potency of the liver oils from differentspecies and also in that of different specimens of oil from a singlespecies.93P. Karrer and his collaborators have founds4 that the liver oilof Hippoglossus hippoglossus, a variety of halibut rich in vitamins,95is a suitable material for obtaining highly active preparations ofvitamin A.Such preparations can be distilled in part at reducedpressure without decomposition, and, since they yield geronicacid on ozonolysis, contain, like p-carotene, the P-ionone residue ;furthermore the yield of geronic acid increases with the Lovibondfigure for the amount of vitamin A present. The molecular weightof vitamin A, determined by the Rast method, is about 320, andfrom the nature of t,he colour reactions and t,he extent to which thepreparations (a) yield acetic acid (from the :CH*CMe: groups) onoxidation and ( b ) undergo catalytic hydrogenation, it is deduced thatvitamin A has a polyene constitution analogous to that of carotene.Experiments with a still more active preparation from the liver-oil of a variety of mackerel (Scombresox saurus) revealed that thevitamin is an alcohol, capable of esterificstion, which after purifica-tion by fractional adsorption on alumina has the formula C,,H,,Oor, less probably, C2,H,20.96 Vitamin A obtained from bothScombresox and Hippoglossus was a pale yellow viscous oil whichprobably contained a small proportion of related substances butnothing of entirely foreign nature, nor any considerable proportionof sterols or vitamin D. From considerations relating to (a) theyield of geronic acid 97 and acetic acid obtained by oxidation of the92 R.A. Morton and I. M. Heilbron, Biochem. J., 1930, 24, 870; A., 1321.93 For details regarding the methods of determination and the spectro-graphic criteria of vitamin A, see J. C. Drummond and R. A.Morton, ibid.,1929,233, 785; A,, 1929, 1202; R. A. Morton, I. M. Heilbron, and F. S. Spring,ibid., 1930, 24, 136; A., 1930, 380; A. Chevallier and P. Chabre, ibid., 1933,27, 298; A., 540.94 P. Karrer and H. V. Euler, Natumiss., 1931, 19, 676; A., 1031, 1195; P.Karrer, H. Morf, and K. Schopp, Helv. Chim. Ac~u, 1931, 14, 1036, 1040; A.,1931, 1463.06 E. Poulson, Struhlenther., 1929, 34, 648; A., 1930, 1070; S. Schmidt-Nielssen, Biochem. J., 1929, 23, 1153; A., 1930, 255.96 P. Karrer, R. Morf, and K. Schopp, ZOC. cit.; P. Karrer, 0. Walker, K.Schopp, and R. Morf, Nature, 1933, 132, 26; A., 805.97 On the basis of the yields of geronic acid obtained in analogous decom-positions, the preparation is calculated t o have contained SO-SO% ofvitamin A158 ORGANIC CHEMISTRY .-PBRT I.Scornbresox preparation, (b) the molecular weight (300-320), (c)the amount of hydrogen absorbed catalytically, and (d) the com-position of the reduction product, the C,,-formula (XLVIII) wastentatively assigned to vitamin A by P.Karrer, R. Morf, and K.Schopp, but the C,,-formula (XLIX) was not entirely excluded.H,?>[CH:CH*CMe:CH] ,*CH,*OHH, Me Vitamin A (XLVIII.)HAP2\ [CH:CH*CMe:CH],*CH:CH*CH,*OH(XLIX.)* 2 < J - - H, MeAccording to (XLVIII), vitamin A represents exactly half the P-carotene molecule with the terminal carbon atom combined withthe elements of water.Karrer and Morf subsequently synthesised a compound, describedas " perhydro-vitamin-A " (L), by reducing the easily obtainedtriene-acid (LI) and extending the carbon chain of the reducedacid by a series of simple operations.This compound is reportedto be identical with fully hydrogenated natural vitamin A.H2~~€€2]2*CHMe*[CH2]3*CHMe*[CH2]2*OHH, €€Me (La)H2 Me2\-CH:CH*CMe:CH*CO,H(LI-). 2 c / H, MeI. M. Heilbron, R. A. Morton, J. C. Drummond, and their col-laborators 98 have purified vitamin A concentrates from halibutliver-oil at a pressure below 0.00001 mm. : a t this pressure thevitamin distils without decomposition a t 137-138" as a pale yellowviscous oil of approximately the composition C,,H,,,O, and ofmolecular weight 320 rf 15 ; there is little spectrographic differencebetween samples so purified and those pursed by the chromato-graphic method. The same authors have carried out an exhaustivespectrographic examination of vitamin A preparations from variousmammalian and fish-liver oils with the object of arriving at aconclusion respecting the homogeneity of the vitamin product.The absorption spectra do not afford entirely conclusive evidenceof homogeneity, and although observations are impressively in98 Bwchem.J., 1932, 26, 1178; A., 1932, 1174FARMER. 159favour of tbe structural representation (XLVIII) for vitamin A,they are not as yet conclusively so; moreover, from the standpointof analytical data, the failure to prepare crystalline derivatives ofvitamin A and the indication over a very large number of ultimateanalyses of some slight degree of variability from the C20H300formula, render any claim to have isolated vitamin A in a state ofpurity at the present stage somewhat insecure.Partial confirmation of Karrer's formulation of vitamin A hasbeen obtained by degradation.A vitamin A concentrate, whendehydrogenated with selenium at 300-330", gave a good yieldof 1 : 6-dimethylnaphthalene, thus showing that the concentratecontained a substance whose constitution as far as the fourteenthcarbon atom must be identical with that shown in (XLVIII).99natural Flavin Dyes (Lyochromes) .A number of naturally occurring colouring matters which appearto belong to the same group of water-soluble nitrogenous dyeshave been isolated from different sources and by different workers.P. Ellinger and W. Koschara have isolated from whey three crystal-line dyes of red or reddish-yellow colour which are rich in nitrogenand oxygen and show strong yellow-green fluorescence in aqueoussolution.1 They suggest that these three dyes are related to twooxidation-ferments, obtained by 0.Warburg and W. Christian,2and to others contained in urine. R. Kuhn and his collaboratorshave independently described two dyes of nitrogenous water-soluble type?" One of these, constituting the colouring matter ofegg-albumen, displays in solution strong green fliiorescence ; theother, isolated from whey, appears to be different from Ellingerand Koschara's preparation and displays strong green fluorescence.The name " flavins " is suggested for the new group.The dye from white-of-egg (180 mg. from about 10,000 eggs),which has been termed " ovoflavin," 3b can be extracted by alcoholsand purified by adsorption on fuller's earth; it has approximatelythe composition C,HlaO3N2, but the double formula C,,H,,O,N,or even C,,H,oQ6N4 may be correct.It resembles spectroscopicallyg9 I. M. Heilbron, R. A. Morton, and E. T. Webster, Biochem. J., 1932,26, 1194; A., 1932, 1174.P. Ellinger and W. Koschara, Ber., 1933, 66, [B], 315, 808; A., 298, 847.Compare €3. Bleyer and 0. Kallmann, Biochern. Z . , 1925,155,54; A., 1925, i,457. Ellinger and Koschara originally proposed the name " lyochromes "for dyes of the new type.2 Biochem. Z . , 1932,254,438; A., 1932, 1285.3 R. Kuhn, P. Gyorgy, and T. Wagner-Jauregg, Ber., 1933, 66, [B], (a)p, 317, (b) p. 576, (c) p.1034; A., 298, 622, 8471 60 ORGANIC CHEMISTRY.-PART I.a dye, ClzHl2O2N4, obtained by Warburg and Christian from yeast:and is very stable towards mineral acids and oxidising agents butis decomposed by alkalis. It is reduced (decolorised) by reducingagents and regains its colour on shaking with air, the change,flavin =+ leucoflavin, depending on oxygen transference.The dyes of this new group offer a, strong contrast in composition,solubility, fluorescence, sensitivity towards acids, alkalis, andoxidising agents, and in biological activity to the carotenes (lipo-chromes). The crystalline orange-brown dye from whey, termedlactoflavin3c (nearly 1 g. from 5,400 litres of whey), was obtainedby Kuhn and his collaborators in their attempts to concentratethe vitamin B, from this source.It has the composition C,,H,,O,N,and resembles ovoflavin closely in decomposition point and inabsorption spectrum. Both ovoflavin and lactoflavin appear tobe related to vitamin B,, and Kuhn and his co-workers state thatthe crystalline lactoflavin not only represents the most activepreparation of vitamin B, yet obtained, but is probably identicaltherewith.6 Lactoflavin yields with sodium bisulphite, sodiumhydrosulphide, zinc dust and acid, or hydrogen and platinum, aleuco-derivative which (a) reverts to the original dye or (b) passesby successive illumination, oxidation, and treatment with coldalkali through two well-defined intermediate stages into a crystallinecompound, C1,H1,0&4 (lumilactoflavin).The latter when boiledwith alkali suffers loss of urea and yields a crystalline compound,C1,H1203N2, which is in turn decomposed by heat into carbondioxide and a crystalline substance, Cl1H1,ON2. From a con-sideration of these changes the partial formulation (LII) is tentativelyput forward.,CH,*OHHO-CHPeroxides and Oxidation Processes.AJdehydes and ketones yield peroxides which hitherto have beenknown only in multimolecular form. They are similarly constitutedto cyclic polymeric aldehydes, except that the oxygen atoms ofthe latter are wholly or partly replaced by the peroxide group;the simplest example of such alkylidene peroxides is ethylidene4 Naturwiss., 1932, 20, 688; A., 1932, 1164; Biochern. Z., 1933, 257, 492;A., 424.5 R.Kuhn, H. Rudy, and T. Wagner-Jauregg, Ber., 1933,66, [B], 1960FARMER. I61peroxide (LIII) . Simple and multimolecular ozonides representmixed peroxidic compounds of a related type : indeed unimolecular$XtMe*O*O-$lHMe ,O'OOH (LV.) OH CHMe b H M e\ / 1 0, /o (LVII.) E::lnCHMe (LIII.1 J.CH, CH, 0 >o/o-o, CHMe.0 *O*CHMe HO*CH,*O*O-CH,-OH (LVIII.)HO*CH,*O*OR (LIX.1 G H M e O*O*CHMe(LVI.1\O/ (LIV.)(as also multimolecular) ethylene ozonide (LIV) belongs to a classof cyclic acetals of which A. Rieche and R. Meister have nowobtained a definite example in mono-perparaldehyde. In theformation of this substance, water is split off from bishydroxyethylperoxide (LV) (2 mols.), a cyclic peroxide (LVI) thus being pro-duced which in constitution is a dimeric ozonide.This " syntheticdimeric ozonide " has in fact properties intermediate between thoseof true uni- and multi-molecular butylene ozonide and indeed itrepresents an intermediate stage between the two, but is not sostable as either. There arises, however, in addition, monomericbutylene ozonide, which is identical with that obtainable by ozonis-ing butylene. The dimeric ozonide appears to break down toethylidene peroxide when kept in a vacuum, but yields on cautiousdistillation mono-perparacetaldehyde (LVII). The latter com-pound is a peroxidic analogue of paraldehyde and its formation isprobably due t o a cracking process in which ethylidene peroxide isThe peroxides of the formaldehyde series hitherto known arebishydroxymethyl peroxide (LVIII) and hydroxymethyl alkylperoxides (LIX), obtained by adding formaldehyde to hydrogenperoxide and to alkyl hydrogen peroxide respecti~ely.~ The formerof these peroxides reacts with two further molecules of formaldehydeto give the bishydroxymethyl derivative, (OH-CH,*O*CH,*O-),,which by prolonged a,ction of phosphoric oxide yields pertrioxy-split off.o<cH,*O~CH,* CH2*0*CH2'8 O=CH,*O CH,*O mO*CH,HaC<O.CH~. /J '<CH2* 0 0 CH,>'(LX.1 (LXI.) (LXII.) -methylene (LX), or by more intense action of the same reagent,tetraoxymethylene diperoxide (LXI) .* Both of these cyclic6 Ber., 1932, 65, [B], 1274; A., 1932,1114. Ann. Reports, 1931,28, 86.8 A. Rieche and R. Meister, Ber., 1933, 66, [B], 718; A., 692; A. Rieche,Angew. Chern., 1932, 45, 441; A., 1932, 831.REP.-VOL.XXX. 162 ORGANIC CHEMISTRY.-PART I.peroxides are explosive, the pertrioxymethylene yielding on ex-plosion trioxyrnethylene and a compound which is possibly per-tetraoxymethylene (LXII).Acetaldehyde with oxygen or air rapidly gives a substance whichin water has the properties of a per-a~id.~ According to H. Wielandand D. Richter lo the reaction between aldehydes and molecularoxygen consists in the first stage, under all observed conditions,in the primary formation of a per-acid : the two reactants directlyunite and it is the true aldehyde form (not the hydrate) whichsuffers autoxidation. The final stage in the autoxidation consistsin interchange between per-acid and aldehyde, yielding acid(*CO,H + *CHO -+ 2 *CO,H), but the course of this reactionremains uncertain. Since peracetic acid and acetaldehyde do notreact in the dry state, the reaction occurring between them inpresence of water probably does so because the Der-acidhydrogen acceptor towards the hydratedThe slow oxidation of the aldehyde whichaldehydi :--+ 2R*CO,Htakes place in aacts asmixtureof dry perbenzoic acid and dry- acetaldehyde is, however, probablydue to the decomposition of an additive compoundbut in general the aldehyde hydrate appears to be responsible forthe rapid oxidation of aldehydes in presence of water. I n theslow oxidation of acetaldehyde with ammonium persulphate, the-0.0. bridge probably acts as hydrogen acceptor.The unsatisfactory results frequently obt,ained in the oxidationof aldehydes to acids by hydrogen peroxide and alkali are attributedby J.von Braun and W. Keller l1 to the slowness of the changesinvolved in (a) the formation of a primary additive productOH*CHR*O*OH or OH*CRR*O*O*CHR*OH, (b) its conversion byloss of water, or dissociation, into R*CH<b, and ( c ) the final trans-formation into R*CO,H, as a result of which opportunity is affordedfor the excess of hydrogen peroxide to attack any incompletelystable radical R. Oxidation with molecular oxygen is too slow forpreparative purposes and is not greatly accelerated by ferrousiron : finely divided manganese dioxide (even 1/4000 mol.), how-* W. H. Hatcher, E. W. R. Steacie, and F. Howland, Canadian J .Res.,lo Annalen, 1932,405,284; A., 1932,722.l1 Ber., 1933, 66, [B], 215; A., 259.(BzO*O*CHMe*OH --+ BzOH + Ac*OH),01031,& 648; A., 1932, 253FARMER. 163ever, or, preferably, potassium permanganate is a suitable acceleratorwhich retains its activity over long periods. The manganesecatalyst probably owes its activity to its interaction with thecarbonyl group, whereby an unstable additive compound is pro-duced which renders the aldehyde more sensitive to oxygen; butwhen an olefin is added to the reaction mixture, the oxidation ofthe aldehyde is retarded, because the olefin is then more favouredby the catalyst than is the aldehyde. Oxidation of acetaldehydeby acidified hydrogen peroxide at 95" yields largely methane andhydrogen, and that of propaldehyde largely ethane and hydrogen.12The mechanism for such reactions proposed by H.Wieland l3 isdisputed.F. Fichter and S. Lurie 1* have now obtained synthetically, bythe action of hydrogen peroxide on ethyl IEevulate, the ester ofthe same bimolecular ketoperoxide, [C02H*CH2*CH,*CMe<O'as was first isolated by C. Harries from the ozonolysis products of 0.1,'rubber and later formulated by R. Pummerer and his collaborator^.^^A higher homologue of this compound derived from c-keto-n-octoicacid has also been obtained.16 Pichter and Lurie have tried byvarious methods to prepare a true carbonyl peroxide containingthe group *CO*O*O*CO*, but without success.16, l7 Whether suchperoxides are formed at any stage during the action of hydrogenperoxide on acids, by the electrolysis of the salts of monobasicacids, or by the oxidising action of potassium persulphate on thesalts of acids, is uncertain, but at all events they do not survive.S.S. Medvedev and E. N. Alexeeva have endeavoured to applythe original method of Baeyer and Villiger to the preparation ofthe higher alkyl peroxides and have obtained by the action ofhydrogen peroxide and alkali on diisopropyl sulphate, isopropylhydrogen peroxide.18PrPO*O*Ba*Q.0*Prs,3H2Q,which is convertible by terephthalyl chloride into the peroxidicester C,H,( CO*O*OPrP),. Attempts to prepare peroxidic compoundsof the camphoric acid series have been successful, and percamphoricacid has been applied to the determination of unsaturation.lgThis compound yields the salt12 S.Bezzi, Gazzetta, 1933, 63, 345; A., 1036.13 Ber., 1921, 54, 2353; A., 1921, i, 889.14 Helv. Chim. Acta, 1931, 14, 1445; A., 1932, 43.15 Ann. Reports, 1931, 28, 85.16 Helu. Chim. Acta, 1933, 16, 885; A., 807.17 F. Fichter and L. Panizzon, ibid., 1932, 15, 996; A., 1932, 929.18 Ber., 1932, 65, [B], 131; A., 1932, 363.19 N. A. Milas and A. McAlevy, J . Amer. Chm. SOC., 1933, 65, 349; A.,279; N. A. Milas and I . s. Cliff, ibid., p. 352; A., 279164 ORGANIC CHEMISTRY.-PAFiT I.Much interest has been shown in the nature and mechanism ofautoxidation processes. A remarkable example is afforded by theautoxidation of the oxido-triphenylpropyl alcohol (LXIII) to theperoxide (LXIV).*O This reaction proceeds in air which containsa little hydrogen chloride (not in acid-free air) and appears toinvolve first cleavage to benzaldehyde and diphenylethylene oxide,then autoxidation of the benzaldehyde, and finally synthesis ofPh*CH*CH*CPh, Ph.CH.0 *O *CH*CPh,(LXIV.)OH I '6 (LXIII.) I \/OH 0the peroxide from the diphenylethylene and the autoxidationproduct of benzaldehyde.Another interesting example is thatof tetrahydronaphthalene, which by the prolonged passage of aira t 75" yields a peroxide, apparently of formula (LXV) : it is sug-gested that the peroxide is derived from a tetrahydronaphthaleneof the form (LXVI) via the form (LXVII).21(LXV.) (LXVI.) (LXVII.) 0 (LXVIII. )According to K. Bodendorf,22 conjugated cyclic dienes areoxidised by oxygen without a catalyst much more rapidly than aremono-eth ylenic compounds, and the products are multimolecularperoxides of type (LXVIII).The reaction is autocatalytic, butthe catalytically active substance (" inductor ") formed duringthe reaction is unstable. Thus, a small amount of freshly oxidiseda-terpinene very greatly increases the rate of oxidation of a secondbatch, but the isolated peroxide, or oxidised material which hasbeen kept for some time, is inactive. Similarly, interruption ofthe oxidation leads to decreased velocity when oxidation is resumed,the decrease being the greater the longer is the interruption. Theinductor from a-phellandrene catalyses the oxidation of cc-phel-landrene, but not that of a-terpinenc, whereas the inductor fromor-terpinene catalyses the oxidation of both m-terpinene and ct-phellandrene.The last is regarded as an example of true inducedoxidation, which probably plays an important part in biologicalreactions. The inductor is assumed to be a moloxide. cc-Terpinene20 E. P. Kohler and E. M. Nygaard, J . Amer. Chem. Sac., 1933, 55, 310;21 H. Hock and W. Susemihl, Ber., 1933, 66, [B], 61; A,, 163.22 Arch. Pharm., 1933, 271, 1; A., 279.A., 271FARMER. 165and a-phellandrene both yield undistillable (multimolecular)peroxides.The autoxidation of linoleic acid is catalysed by numerous basiccompounds, by dihydroxyacetone and related compounds, bycarotene and related compounds (not bixin), by metal salts, and byvarious other materials; the autoxidation of oleic acid is similarlycatalysed by various basic compounds, but the autoxidation ofboth substances may be inhibited by the addition of certain othercorn pound^.^^ Ferrous iron catalyses the autoxidation of phenolswith the production of black humus-like products, but this onlyapplies when the hydroxyl groups are not replaced by alkyl groups.24The autoxidation process is not accelerated by ferrous iron unlessthe conditions necessary for the formation of hydrogen peroxideare fulfilled, and the hypothesis is put forward that the catalyticinfluence of the iron is exercised in promoting the action of initiallyformed hydrogen peroxide.The autoxidation of methyldihydro-bixin proceeds in the presence of amines to the absorption of tenatoms of oxygen per molecule without coming to a definite end;with caustic soda, rapid absorption of two atoms of oxygen takesplace, followed by a much slower consumption of oxygen.25 Thecarotenoids, including vitamin A, themselves act as catalysts 26in promoting the autoxidation of fatty acids but not of glycerides,and in the presence of hemin they promote a considerable degreeof autoxidation of neutral unsaturated oils ; when, however, theyare employed with other substances, other effects are observed.Studies of the mechanism of autoxidation of a-ketols have beenmade by A.Weissberger,Z7 and of the autoxidation products ofpiperitone by W. Treibs.28The existing evidence that peroxides and ozonides are effectivecatalysts for polymerisation of both hydrocarbons and aldehydesis strengthened by the experiments of 5.B. Conant and W. R.Peterson 29 on the polymerisation of aliphatic aldehydes and isopreneunder high pressures (12000 atm.). Here peroxide catalysis is25 W. Franke, Annalen, 1932,498,129; A., 1932,1112; P. Rona, R. Asmus,24 A. Bach, Ber., 1932, 65, [B], 1788; A., 1933, 155.25 R. Kuhn, P. J. Drumm, M. Hoffer, and E. F. Moller, Ber., 1932, 65, [B],26 W. Franke, 2. physiol. Chem., 1932,212,234; A., 1933, 49.27 Ber., 1932, 65, [HI, 1815; A., 1933, 161; see also A. Weissberger, H.Maim, andE. Strasser, ibid., 1929, 62, [B], 1942; A., 1929, 1301.28 Ibid., 1930, 63, [B], 2423; A., 1931, 94; ibid., 1931, 64, [B], 2178; A.,1931, 1299; eb seq.2Q J . Amer. Chem. SOC., 1932, 64, 628; A,, 1932, 367; see also J.B. Conantand C. 0. Tongberg, ibid., 1930, 52, 1659; A., 1930, 735.and H. Steineck, Biochem. Z., 1932, 250, 149; A., 1932, 1003.1785; A., 1933, 52166 ORGANIC CHEMISTRY.-PART I.essential to polymerisation (very minute proportions of catalystare sufficient) and the effect of the increased pressure is only toaccelerate the reaction. The rate of polymerisation of n-butaldehydeparallels the amount of oxygen which has been absorbed before thepressure treatment commences; and whatever the nature of theactual catalyst may be it is considered not to he an intermediatebetween the aldehyde and the acid formed therefrom. Polymeris-ations of this kind can be catalysed by added peroxides (e.g.,benzoyl peroxide) or greatly hindered by anti-oxidants such ashexaphenylethane and it is supposed that they result from a seriesof chain reactions initiated in the liquid by the spontaneous de-composition from time to time of the single molecules of peroxidiccatalyst.The accelerating effect of great pressure is accordinglydue to the orienting of the molecules of aldehyde or isoprene intomore compact bundles in which longer reaction chains can bepropagated by the decomposing catalyst. The accelerating effectof temperature is attributed to the more frequent peroxide decom-positions which occur when the temperature rises, and the efficiencyof a given peroxide as catalyst depends on its instability. Thepolymeric aldehydes formed by these polymerisation processescannot be stabilised.An important catalytic effect of peroxides has recently beenobserved in additive reactions : 3O traces of peroxidic materialpresent as impurities in ally1 bromide, vinyl bromide, and propylene(or even small quantities of added peroxides) reverse the normaldirection of addition of hydrogen bromide to these substances.A number of new oxidising agents have been brought into usefor specific purposes.Lead tetra-acetate has been successfullyemployed by R. Criegee 31 and subsequent workers for the oxidationof ethylenic substances and the oxidative fission of a-glycols ;periodic acid has been applied to the oxidation of hydroxy-acids,sugars, polyhydric alcohols, and a- and 8-glycerophosphoric acids ; 32an iodo-silver benzoate complex is reported by C.Pr6vost 33 to beserviceable for the oxidation of ethylenic compounds to a-glycols ;and selenium dioxide has been employed in the oxidation of30 M. S. Kharasch and F. R. Mayo, J . Arner. Chem. SOC., 1933, 55, 2468,2521 ; A., 805; M. S. Kharasch, M. C. McNab, and F. R. Mayo, ibid., p. 2531 ;A., 805.3 1 Annalen, 1930, 481, 263; A., 1930, 1278; Ber., 1931, 64, [B], 260; A.,1931, 461.32 L. Malaprade, Compt. rend., 1928, 186, 382; A., 1928, 269; Bull. SOC.chim., 1928, [iu], 43, 683; A., 1928, 867; P. Fleury and J. Lange, C m p t .rend., 1932,195, 1395; A., 1933, 376; J . Pharm. Chhn., 1933, [viii], 17, 313;A., 591 ; P. Fleury and R. Paris, Compt. rend., 1933, 196, 1416; A., 696.Cmtpt. rend., 1933, 196, 1129; A., 711FARMER.167aldehydes and ketones containing the group *CH,*CO- to dicarbonylcompounds 34 and of olefins to aldehyde~.~5Ascorbic Acid.In 1928 A. Szent-Gyorgyi isolated from adrenal cortex, oranges,and cabbages a highly reactive crystalline substance, C6H806,which on accouut of its acidic character and its resemblance tothe carbohydrates in reducing power and colour reactions wasnamed hexuronic a ~ i d . ~ 6 This substance, which is widely dis-tributed in plants and animals, was found to possess strong anti-scorbutic properties 37 and subsequently aroused great interestas the possibility became apparent that it might represent theantiscorbutic factor of vegetable and animal materials (vitamin C)in pure crystalline condition. Its constitution and behaviourhave been extensively investigated, but owing to the complexityof the problem presented by its biological activity its uniquenessas an antiscorbutic agent has not been finally ~stablished.~~Szent-Gyorgyi’s compound is not a member of the uronic acidclass and has been renamed ascorbic acid.39 It is a weak acidwhich does not lactonise, and since it gives salts of the type C6H706Mit cannot itself be a lactone of an acid C6HI0O,.That at leastfive of the six carbon atoms are present as an unbranched chainis shown by the fact that it yields furfuraldehyde quantitativelyon treatment with boiling hydrochloric acid; 40 that it is enolicand probably contains no free aldehyde group is shown by theintense colour developed with ferric chloride and by the failure togive a colour with Schiff’s reagent, respectively.With phenyl-hydrazine it yields a diphenylhydrazone, and with other ketonicreagents analogous derivatives.40, 41 Its most characteristic prop-erty, however, and one that is probably largely responsible for34 H.. L. Riley, J. F. Morley, and N. A. C. Friend, J . , 1932, 1875; A,, 1932,833.35 H. L. Riley and N. A. C. Friend, ibid., p. 2342; A , , 1932, 1108.36 Biochem. J . , 1928, 22, 1387 ; A., 1929, 98.37 J. L. Svirbely and A. Szent-Gyorgyi, Nature, 1932, 129, 576, 690; A.,1932, 548, 657; Biochem. J., 1932, 26, 865; A., 1932, 886; ibid., 1933, 27,279; A., 541; T, W. Birch, L. J. Harris, and S. N. Ray, Nature, 1933, 131,273; A., 433; J. Tillmans, P. Hirsch, and R. Vaubel, 2. Unters. Lebensm.,1933, 65, 145; A., 433.38 For a summary of the evidence regarding the relationship t o vitamin C,see A. Szent-Gyorgyi, Nature, 1933, 131, 225; A., 433.3* W. N. Haworth and A. Szent-Gyorgyi, ibid., p. 24; A., 196.40 E. G. Cox, E. L. Hirst, and R. J. W. Reynolds, ibid., 1932, 130, 888; A.,*I P. Karrer, H. Salomon, K. Schopp, and R. Morf, Helv. Chim. Acta, 1933,1933, 100.16, 181 ; A., 490168 ORGANIC CHEMISTRY.-PART I.its biological activity, is its ability to undergo reversible oxidation.It is immediately attacked in cold neutral or acid solution by aqueousiodine, ozone, silver nitrate, copper acetate, and permanganate,and in alkaline solution by gaseous oxygen and sodium hypoiodite.etc. etc. 266 ORGANIC CHEMISTRY .-PART 111.Attempts to prepare other optically active triarylmethyl deriva-tives from the active thiolacetates were invariably unsuccessful.This implies that the intermediate carbonium ion (IV) is opticallyunstable. Further, the optical iiistability of the free radical (V)was shown by the gradual loss of optical activity when Z-phenyl-diphenylyl-a-naphthylmefhylthioacetic acid was allowed to reactwith triphenylmethyl in the presence of mercury. The rotationfaded to zero after about 30 days.The resolution of two compounds showing molecular dissymmetryis noteworthy. S. E. Janson and (Sir) W. J. Pope 29 have announcedthe resolution of diaminospirucycloheptane (VI) by means of d- andZ-camphor- 8-sulphonic acids, the magnitude of the molecular rotatorypower of the dihydrochlorides, & 30°, leaving no doubt asto the genuineness of the phenomenon. C. S. Gibson and B. Levin 30have resolved dZ-spirobis-3 : 5-dioxan-4 : 4'-di(phenyl-p-arsonic acid)(VII) by means of d- and Z-nor-4-ephedrines. The two enantio-morphous acids had [a]5461 & 24" in aqueous solution. The acetalnature of this substance is shown by the complete racemisationwhich occurs on conversion into the bis(dich1oroarsine).Finally, T. T. Chu and C. S. Marvel 31 have furnished an attractiveand conclusive proof of the unsymmetrical nature of the azoxy-group by the use of optical properties. H. King 32 proposed to solvethe problem by the resolution of rnethyl- 2 : 2'-azoxy-4 : 4'-bis-dimethylaminodiphenylmethane (VIII), and S. C. Hussey, C. S.Marvel, and F. D. Hagen 33 attempted to make ethyl-2 : 2'-azoxy-diphenylmethane-4 : 4'-dicarboxylic acid (IX) but found that theproduct was polymeric.2s Chem. and Ind., 1932, 316; A , , 1932, 508.30 Proc. Roy. SOC., 1933, A., 141, 494; A . , 1177.3 1 J . Amer. Chem. Soc., 1933, 55, 2841; A., 945.32 J . SOC. Chern. Ind., 1930, 49, 281.33 J . Amer. Chem. Soc., 1930, 52, 1122; A., 1930, 774KING. 267Chu and Marvel have now prepared meso- and &I-a-p-azophenyl-butyric acids (X) and have resolved the dl-compound. On oxidationof the meso- and the dl-compound with hydrogen peroxide, two azoxy-H (I)B=N<\:CO,H0 4 Co2Ha5<>=Nr\ko,H C O p C ‘-4 Et I E t (XI.) (X.) Et I Etcompounds (XI) were produced, both of which were resolved intooptically active forms. If the azoxy-group had been symmetrical,the original meso-acid should have given an unresolvable meso-azoxy -compound.HAROLD KING

ISSN:0365-6217    

DOI:10.1039/AR9333000133

出版商:RSC    

年代:1933

数据来源: RSC

摘要:

ANALYTICAL CHEMISTRY.THE continued application of physical methods t o analyticalproblems is reflected in the large number of communications appear-ing during the period under review, depending mainly upon physicalmeasurements. Opportunity is here taken to deal with one or twoof these methods in some detail. Special attention has also beengiven to advances in gas analysis, since the methods now availablefurnish means for the more accurate and rapid analysis of gaseousmixtures. It is hoped that the short accounts given in this Reportwill achieve its object of placing before the reader the presentposition of a few chosen aspects of analytical chemistry; thereferences are full.Polarographic Method-Since we think that the " polarographic "method of Heyrovsky and his co-workers should find wider appli-cation as a method of determining various cations, a short summaryof the recent work in this field may be useful.The general theory ofthe method and its application were described some time ag0.lIt need only be stated here that the expression upon which themethod is based is iii form similar to that for concentration cells,namely, x1 - x2 = RT/nP. In C,/C2, where x1 is the so-calleddeposition potential of a molar solution referred to the normalcalomel electrode, x2 being the voltage of a more dilute solution ofthe same substance; x1 is the point of inflexion of the current-voltage curve and is observed by a sudden upward surge in thecurve, tables being available for its value for various cations. x2 isthe position of inflexion on dilution, and since the value xl-nXzdepends on the logarithm of the concentrations the displacementfor a univalent cation on dilution from normal to decinormal isabout 58 millivolts only at ordinary temperatures.For this reason,the point of inflexion does not move much on dilution, although thepolarisation current makes a sudden jump at this point. If im-purities are present, as for example traces of lead, zinc, or iron incommon salt, these can be detected from the position of the onsetof the " surge " in the curve, since each cation has its own depositionpotential. The height of the surge as shown on the current-1 See especially J. Heyrovskf, Rec. trav. chim., 1925, 46, 488; Bull. SOC.chim., 1927, 41, 1224; 2. anal.Chem., 1933, 91, 206; Mikrochent., 1933, 12,2sFOX. 269voltage graph can be used to measure the quantity of the impurityprovided the apparatus is calibrated for this purpose. The de-position potentials must be sufficiently separated to make a dis-tinction possible, otherwise the method can only give the totalquantity of the overlapping cations. In such cases various devicesmay be employed to distinguish the cations, such as rendering thesolution alkaline, or forming complex ions with potassium cyanide.In general the method is very delicate, cations in concentration ofaboutBriefly, the apparatus used consists of a cathode formed by finedrops of mercury charged negatively, and a very large anode formedby a pool of mercury charged positively, upon which is placed thesolution whose concentration is to be determined. A galvanometerof fair sensitivity is inserted in series with the cathode. Thevoltage on the anode is varied continuously by means of a rotatingpotentiometer from, say, 0 to 2 or 4.With the large mercuryanode, a condition of non-polarisability is in practice secured. Themovements of the galvanometer mirror are recorded, or photo-graphed on a rotating drum for rapidity, and plotted against thevoltage referred to the normal calomel electrode.An application of the polarographic determination to alkalimetals has been made using tetramethylammonium hydroxide asthe reference liquid, since the deposition potential of NMe, is - 2-6volts, that of lithium being - 2.033 and of sodium and potassiumvery near - 1-87 volts.It is not possible to determine the last twometals separately, but the estimation of the first in the presence ofthese alkalis is secured. Other metallic ions, provided they can beprecipitated or converted into some complex, do not interfere withthe determination of these alkalis. For example, calcium may beprecipitated as phosphate, magnesium i s hydroxide, and aluminiumconverted into an aluminate. If the separate proportions of sodiumand potassium are desired, then the total alkali sulphate or chlorideis determined gravimetrically and the usual calculation made.Consideration of the reactions occurring at the dropping mercurycathode shows that it is available for dealing with electro-reductionprocesses, and even the nitrate and nitrite anions may be estimated.It has been suggested3 that the strong field in the neighbourhoodof the cathode breaks up these ions in the first place into suchpowerfully charged atoms as N5+, which presumably are then ableto attract more anions towards the cathode surface, so replenishingthe supply.This is assumed to occur as well as diffusion, withoutwhich no continuous measurements could be made. I n neutral ora V. Majer, 2. anal. Chm., 1933, 32, 321; A., 583.3 M. Tokuoka, Coll. Czech. Chem. Coinrn., 1932, 4, 444; A., 1932, 1221.g.-equiv. per litre being detectable270 ANALYTICAL CHEMISTRY.alkaline solutions reduction takes place only if bi- or ter-valentcations are present in the liquid.For example, in a solution ofN/10-lanthanum chloride the reduction potential is - 1.2 voltsreferred to the normal calomel electrode, and for nitrate eightfaradays are required to give ammonia. Again, nitrite is de-composed by acids with the formation of nitric oxide, which iscapable of easy electro-reduction, the increase of current in acidsolutions occurring at - 0.77 volt. Very small concentrationsof nitrate and nitrite can be determined ; * for instance, it is claimedthat quantities as small as 1 part of nitrous acid in 5 x lo7 parts ofan explosive may be estimated, and much less than this may bedetected.An interesting application of the electro-reduction is the deter-mination of nicotine in the presence of pyridine and ammoniumsalts, as little as one part in ten million parts of tobacco beingdetectable.That the method is capable of more extended use isshown by the determination of cystine and cysteine in the hydrolysisproducts of proteins from a comparison of the current-voltagecurves with those of known quantities of the substances in solutionsof cobalt chloride. Up to about 24 mg. per litre is stated to be thebest concentration for accurate work.Magneto-optical Method.-In the Annual Reports for 1930 a briefstatement of the magneto-optical method for the detection ofcations was given, and since then a number of papers has appeareddealing with the application of the method to the detection ofisotopes, and in a few cases to the determination of cations. With-out entering into discussion of the underlying principles of the‘‘ time lag ” in the Faraday effect, it appears to be established thatthe procedure is capable of detecting, and in many cases estimating,certain cations even when these are present in extremely minutequantities.For example, calcium has been determined where largeamounts of anions such as nitrate or phosphate and even whenmagnesium is p r e ~ e n t . ~ The method hitherto adopted was to alterthe length of the circuit to the cell containing the solution undertest, so that the “ time lag ” was counterbalanced. It is now foundthat by modifying the polarising system, it is possible to pass thecurrent in the same direction through the cells containing the carbondisulphide and the solution under test, and by altering the settingof an analysing Nicol prism to determine the angle of rotation4 G, Semerano, Biorn.Chirn. ind. appl., 1932, 14, 608; A., 243.5 Ibid., and {bid., p. 1093.6 R. BrdiEka, Coll. Czech. Chern. Cornm., 1933, 5, 238; A., 964.7 E. R. Bishop and C. B. Dollins, J . Amer. Chem. SOC., 1932, 54, 4586;A,, 137FOX. 27 1necessary to produce a minimum of light intensity.8 The apparatusis calibrated by means of solutions of known concentrations ofcalcium chloride, allowance being made for the isotopes. Theadvantages claimed for this method of determining calcium arethat it can be used for lower concentrations than any other methodavailable, and that other substances do not interfere with thedetermination. Since the angle of rotation is stated to be the samefor the same quantity of calcium in any of the compounds studied,it would appear that the method is available for calcium in any kindof material provided a clear solution of the calcium compound canbe obtained.The process is somewhat limited in its application,not so much on the grounds of accuracy as from the fact that twoobservers are required, and that much practice is necessary to attainskill in working the apparatus, apart from the strain on the eyes ofthe observers. As the method is claimed to be available for thedetermination of as little as about 4 x g. of calcium per litre, itwill be seen that it has possibilities not possessed by any otheranalytical process hitherto proposed.Spectroscopic Methods.-The investigations reported are, ingeneral, applications, extensions, or suggested improvements ofthe technique of modern practice. For quantitative work onemission spectra of the elements, usually metals, designed to esti-inate the proportion of a particular constituent, the methodscommonly employed finally depend for their accuracy on two verygeneral considerations : (1) determination of the intensity ofspectrum lines as shown by equality of a photographic image of thelines from alloys or compounds of known composition, or com-parison of the relative intensity of the lines visually or photo-graphically, (2) evaluation of the intensity of the lines by use ofthe logarithmic wedge sector.In (1) difficulties arise from thenecessity of obtaining uniformity in the production of arc or sparkdischarges if direct comparison on each occasion with a set of stand-ard alloys is to be avoided.Methods for producing uniform con-ditions of excitation of arc and spark spectra are again discussed,and useful reviews of the conditions necessary for obtaining re-producible results are described in some d e t d 9 In view of thedifficulties of obtaining exactly reproducible discharges, it is desir-able, if possible, to reproduce the luminosity of the spectral lines,for in this case the conditions of excitation of the spectrum aredefinite. One way of doing this is t o arrange conditions of excitation8 E. R. Bishop, C. B. Dollins, and I. G. Otto, J . ,4rraer. Clbem. Soc., 1933, 55,9 0.Feussner, Arch. Eisenhiittenw., 1932-1933, 6, 651; F. Waibel, 2.4365.Metallk., 1933, 25, 0 ; 0. Feussner, ibid., p. 73; A., 800272 ANALYTICAL OHEMISTRY.in such a way that an arc line of an element of the alloy or metalunder investigation is equal in intensity to that of a spark line ofthe same element, the two lines, arc and spark, being close togetheron the photographic plate. The electrical conditions of the circuitcan then be varied in order to ascertain to what extent variationsmay be permitted without altering the apparent equality of in-tensity of the two lines. It is true that the preliminary search forsuch lines is tedious and the labour involved is great, but it is alsoapparent that the luminosity of the spectrum for any particularalloy is thereby fixed.This scheme has been employed with muchsuccess by Gerlach and his co-workers.10 One of the best studiesinvolving (2) is that of F. Twyman and F. Simeon,ll the limitationsof the method, and particularly the relation of intensity of illumina-tion and blackening of the plates, being dealt with adequately.One essential condition for the application of the process is that theintensity of a spectrum line of one constituent of an alloy relative t othat of a line of a second constituent should increase as the pro-portion of the first constituent increases, a condition which does notalways necessarily follow with alloys. A variation of the moreusual methods, which seems to be capable of extension beyond thecases given, namely, silicon in steel, chromium in nickel and barium-nickel alloys, has been described in some detail.12 Arcs are used,and such a value of the arc current is chosen that a small changein current is accompanied by only a slight alteration in the relativeintensity of the spectral lines.Furthermore, an intensity cali-bration pattern is photographed on the same plate as the spectrum,through a step diaphragm, thus providing a correct method forphotographic photometry . This procedure renders unnecessarythe considerations of the values of the plates and whether or notthe lines under examination fall on straight portions of the character-istic curve of the plates. The relative densities of any two lines tobe compared, e.g., Ba 4554.0 and Ni 4546.9, are then determined bya microphotometer from a comparison with the density of thecalibration pattern. Once a calibration curve for the range of asmall constituent of an alloy is drawn, all similar alloys can beexamined accurately by reference to the curve, for the ratio of thelogarithms of the relative intensities of a pair of spectral lines isplotted against the proportion of the minor constituent. Here againi t is to be n6ted that the curves of relative intensities can only be10 See W.Gerlach and E. Schweitzer, “Die Chemische Emissionsspek-l1 Trane. Opt. SOC., 1930, 31, 169.tralanelyse.”0. S. Duffendack, R. A. Wolfe, and R. W. Smith, I n d . Eng. Chem.(Anal.), 1933, 5, 226; a., 920FOX. 273used for one type of alloy and that if mother large constituent isadded to the alloy a new curve is necessary.Better results can beobtained if the relative intensity of the pair of lines is referred tothe proportion of the primary constituent of the alloy and not to thealloy as a whole. The logical method of calculation is atomicproportions of the constituents of the alloy.An interesting application of the spectrographic method 13 is thequantitative study of the coprecipitation of cations with bariumsulphate, a subject of great analytical interest for the accuratedetermination of sulphates. Carbon arcs were used and thesewere impregnated with solutions prepared from the precipitates,the quantity of coprecipitated cation being determined by com-parison with prepared standards.Calcium and strontium are co-precipitated, but not beryllium, magnesium, or zinc, presumablyin accordance with the rule that adsorption is likely to occur onpolar crystals with ions which form slightly soluble compoundswith the oppositely charged ions of the crystal lattice. The ruleseems to break down for cadmium iodide, which is adsorbed onbarium sulphate, and it is suggested that this is due to a preferentialadsorption of the iodide ion and also of a CdI," ion which is pre-sumed to be formed from the slightly dissociating cadmium iodidein aqueous solutions.For the more qualitative methods for the detection of variouselements several investigations are recorded. An intermittent orflame arc is described l4 whereby such small amounts as 2 x lo-* g.of arsenic or lo-' g.of tellurium can be detected in solids. Beryl-lium may be detected in minerals after decomposition with hydro-fluoric acid and conversion into chlorides, if the solution is excitedby means of a high-frequency alternating arc of high potential.15The lines at 3131 and 2349A. are used, the first line detecting aslittle as O . O O O l ~ o in certain conditions. It has been found thatcertain elements exercise a marked damping on these emission lines,and that the damping increases in the order K, Na, Ba, Ca, Mg, Al,Si, an order following the ionisation potentials of these elements.A difficulty in the detection of some of the commoner elements inthe minute quantities which can be found spectroscopically is thepreparation of electrodes free from the elements sought for.If,for example, as little cadmium as lWIO g. can be found, then theelectrode of silver used must clearly be of special quality.1613 L. Waldbauer and E. St. C. Gantz, I n d Ertg. Chern. (Anal.), 1933,5,311;14 E. Riedl, 2. anorg. Chem., 1932, 200, 356; A., 134.1 5 W. Kemula and J. Rygielski, Przemyst Chem., 1933, 17, 89; A., 798.16 W. Spiith, Monatsh., 1932, 81, 107; A., 1932, 1220.A., 1133274 ANALYTICAL CHEMISTRY.FZuorescence.-A useful device for work on fluorescence of solu-tions is designed to eliminate the troublesome spurious reddish-violet fluorescence sometimes seen when measurements of the truefluorescence colours are needed. This spurious fluorescence isthe result of internal reflexions in the apparatus of the ultra-violetand far red regions of the exciting radiation.The interposition ofa filter screen consisting of a solution of potassium nitrite is effectivein suppressing the ultra-violet part of the spurious reflexions,permitting of more accurate examination of the scattered radiati0n.l'The use of a screen of the nitrite is likewise effective in eliminatinga large proportion of ultra-violet radiation when a mercury lamp isemployed as the exciting source in studying the Raman effect.X-Ray Spectrography .-Two important communications in X-rayspectrographic methods call for brief mention. The effect of thesize of the particles and the necessity for employing very finepowders in quantitative X-ray chemical determinations has beendiscussed and investigated, and it is shown that the intensity of thediffracted ray may be decreased by irregularities in the surface ofthe target.l8 The molybdenum and the tungsten content of igneousrocks have been shown to be about 1.5 x and 6.9 xrespectively.For determining the minute proportions here con-sidered, the two elements are recovered from the mineral by pre-cipitation with sodium hydroxide on a barium base, and the chloridesare volatilised from the precipitate by meaiis of a stream of hotchlorine gas. The sublimed chlorides are then dissolved in acid,and molybdenum sulphide is coprecipitated with a substantialproportion of copper sulphide, the copper being subsequentlyremoved by electrolysis. Molybdenum may then be determinedcolorimetrically.In the filtrate from the molybdenum sulphide,tungsten is estimated by the X-ray spe~trograph.1~Staleness of Fish.-The range of application of electrometricmethods is now so extensive that it is hardly surprising to find itapplied to such a problem as estimating the relative freshness ofhaddock, and it will probably in due course be extended to otherkinds of fish. At one end of the freshness scale the nose is doubtlessan effective organ for the detection of loss of freshness, and at thisstage further investigation is not informative. There are twochanges in the decomposition of the proteins of fish which occLlrafter rigor mortis begins to fade. In the first stage, due to enzymeaction, autolysis begins and the proteins hydrolyse to polypeptides,peptones, and amino-acids.These substances are then brokenl7 E. Grunsteidl, Mikrochern., 1933, 13, 183; A., 800.l8 G. R. Fond, J. Amer. Chern. SOC., 1933, 55, 123, 127; A., 242.lS G . von Hevesy and R. Hobbio Z. onorg. Chern., 1933,212,136; A., 688FOX. 275down further by bacterial decomposition in the second stage toammonia, amines, indole, skatole, hydrogen sulphide, and othersubstances. Many of these are basic, so that this decomposition isaccompanied by a rise in pH value. The second decomposition alsooccurs to some extent during autolysis, and the method employedis to titrate a solution prepared from the fish, first to pH 6.0 and thento p , 4.3-the range of pH where the buffer effect of the inorganicphosphates is least.The amount of acid used in the first titrationshould be a measure of the degree of secondary bacterial decomposition,while the acid used in going from pH 6 to p H 4-3 should be in inverseproportion to the extent of the protein hydrolysis. The method oftest is to use about 5 g. of finely minced flesh, shake it vigorouslywith about 100 ml. of water, and add excess of quinhydrone. Witha platinum electrode and saturated calomel cell, titration is carriedout using O.0165N-hydrochloric acid. The amount of acid requiredto alter the original voltage of the fish extract to the voltage requiredfor pH 6.0 and then the amount for the step to pH 4-3 are recorded,protein errors being disregarded. A discussion of the results andthe significance of the variations in the quantity of acid used isdescribed in detail.*OUltimate Analysis.-The ter Meulen method of determiningoxygen directly in organic substances has been found to give poorresults on many occasions, and this has clearly been due to theemployment of relatively inactive catalysts so that methane failsto be formed from the oxides of carbon, or to sublimation or toorapid volatilisation of the substance.A reinvestigation of themethod 21 shows that satisfactory results can be obtained with allkinds of substance, ranging from oxalic acid to anthraquinone,provided active catalysts are used and suitable precautions taken.Asbestos-supported catalysts are undesirable, since the asbestosgives up water deposited on them very slowly, and the success of thedetermination depends mainly on the proportion of water formed.The catalysts recommended are platinum-coated quartz in finegranules or thoria-coated reduced nickel.Reported results forsubstances so far examined containing only carbon, hydrogen, andoxygen are good. J. J. F.Physicul Properties of Solutions.-The simple oxygen-containingorganic compounds are of considerable commercial importanceand, as in many countries revenue interests are also involved,these substances have attracted and still attract much analyticalattention.20 M. E. Strtnsby and J. M. Lemon, Ind. Eng. Chem. (Anal.), 1933,5,208.21 W. W. Russell and J. W. Fulton, ibid., p. 384276 ANALYTICAL (IHEMISTRY.In the course of analysis, the compounds are usually obtainedin the form of aqueous solutions.The simplest way of dealingwith these is by physical methods, e.g., determination of specificgravity, refractivity, etc., and for solutions of single substances, oneof these will suffice. It frequently happens, however, that absoluteknowledge of individual character is unknown, and in such casesit is necessary to obtain proof by agreement or non-agreementbetween two or more physical properties, possibly coupled withchemical evidence. To give a simple example, the refractivity of adilute alcoholic distillate may agree with the ethyl alcohol strengthas indicated by specific gravity ; this may, however, be mere coinci-dence unless the absence of methyl alcohol is assured, since thislow-refracting substance may otherwise be associated with a com-pensating quantity of a compound of higher refractive power.The physical properties of mixed aqueous solutions are notstrictly additive in character ; nevertheless, if concentration is nottoo great, e.g., up to lo%, such properties as specific gravity andrefractivity of the substances under consideration are sufficientlyadditive that allowance can be made for the presence of compoundsdetermined by other methods.In the following account, which is intended to cover mainly thedevelopments during recent years, occasional reference is made toearlier investigations.Alcohols.-Of the lower aliphatic alcohols, methyl, ethyl andisopropyl are of first practical importance, n-propyl being relativelyinsignificant.Several sets of derivatives have been prepared,e.g., phenylcarbamates (m. p.'s 47", 51", go", and 57" respectively)and p-xenylcarbamates (m. p.'s 127", 119", 138", and 129"),22p-iodoxenylenecarbamates (m. p.'s 191", 200", -, and 189"),23p-nitrophenylcarbamates (m. p.'s 179", 129", 116", and 115"),%3 : 5-dinitrobenzoates (m. p.'s 107", 92", 122", and 73"),25 and a-naphthylcarbamates (m. p.'s 124", 79", 78", and 105"). Thep-nitrobenzoates can be isolated from dilute solutions of thealcohols.26There exist many very delicate tests for form-aldehyde, and the majority of tests for methyl alcohol dependupon its preliminary oxidation to this compound; those in whichthe process is carried further to formic acid or carbon dioxide are22 G.T. Morgan and A. E. J. Pettet, J., 1931, 1126.z3 S. Kawai and K. Tamura, Sci. Papers Inst. Phys. Chem. Res. Tokyo,24 R. L. Shriner and R. F. B. Cox, J . Arner. Chem. Soc., 1931, 53, 1601;25 G. B. Malone and E. E. Reid, ibid., 1929, 51, 3424; A., 1930, 58.36 H. Henstock, J., 1033, 216.MethyZ aZcohoZ.1930,13, 270; A., 1930, 1159.A., 1931, 709ELLIS. 277generally less sati~factory.~~ Since the acetaldehyde produced onoxidation of ethyl alcohol interferes with many of these tests, theearly applications of such methods to mixed alcohols were of littlevalue. The process proposed by G. Denighs 28 was designed to meetthe conditions imposed by these circumstances, and the methodwas later applied q~antitatively.~~ Recent workers have found itadvantageous to replace the sulphuric acid originally employed byphosphoric a~id.~O Some of these also give lists of substances whichhave been found to interfere and of those which do not; to theformer class must be added dimethyl ether.31 Many of the inter-fering substances, e.g., glycerol and pectin, are readily separatedby distillation.Mixtures of methyl and ethyl alcohol may be evaluated rapidlyby dilution with water t o a total alcohol concentration of not morethan 40%, which can then be ascertained from the specific gravity,since equal concentrations (by weight) of the two alcohols in waterhave practically the same density ; the refractivity of the mixturesthen serves to give the individual amounts of the alcohols.32 Ifacetone and its homologues also be present, they may be removedby refluxing with paraformaldehyde in the presence of alkali, theexcess of aldehyde being removed by oxidation with Fehling'ssolution prior to di~tillation.~~The use of 5 : 5-dimethyldihydroresorcinol (methone, dimedon)for fixing and characterising aldehydes, though not new,34 hasattracted considerable attention in recent years : 35 the formationof methylenedimethone has been utilised for the detection of verysmall quantities of methyl alcohol in ethyl alcohol.36 The mixed27 F.R.GeorgiaandR.Morales, I n d .Eng. Chem., 1926,18,304; B., 1926,381.28 Compt. rend., 1910, 150, 832; A., 1910, ii, 461.29 C. Simmonds, Analyst, 1912, 37, 16; A., 1912, ii, 208; G. C. Jones,ibid., 1915, 40, 218; A., 1915, ii, 493.30 F.R. Georgia and R. Morales, Zoc. cit., ref. (27); R. M. Chapin, J . I n d .Eng. Chem., 1921, 13, 543; A., 1921, ii, 598; L. 0. Wright, ibid., 1927, 19,750; A., 1927, 687; H. Patzsch, Pharm. Ztg., 1932, 77, 1191; B., 43;Berg, ibid., p. 1262; B., 91; H. Jeglinski, ibid., 1933, '78, 77; B., 181.31 B. F. Dodge, I n d . Eng. Chern. (Anal.), 1932, 4 , 2 3 ; A., 1932, 250.32 F. S. Mortimer, I n d . Eng. Chem., 1927, 19, 635; A., 1927, 687; R. H.Field, M. W. Fairn, and J. M. Macoun, J. Soc. Chem. Ind., 1931, 50, 28311;B., 1931, 963; J. M. Macoun, ibid., p. 2 8 1 ~ ; B., 1931, 963.33 R. W. Hoff and J. M. Macoun, Analyst, 1933, 58, 749.34 D. Vorlander, Ber., 1897, 30, 1801; A., 1897, i, 513; C. Neuberg and(&I.) E.Reinfurth, Biochem. z., 1920,106, 281; A., 1920, i, 914; W. Stepp,ibid., 1922, 127, 13; 130, 578; A., 1922, ii, 403, 793.35 D. Vorlander, 2. anal. Chem., 1929, 77, 241, 321; A., 1929, 924, 949;G. Klein and H. Linser, Mikrochem., 1929, Pregl Fest., 204; A., 1929, 1292;L. Kofler and H. Hilbck, ibid., 8, 117; A., 1930, 940.36 0. Mohr, ibid.: p. 154; B., 1930, 651278 ANALYTICAL CHEMISTRY.aldehydes obtained by oxidation as in the Denighs process aredistilled into ammonium chloride solution ; evaporation nowexpels the acetaldehyde, and the formaldehyde, regenerated fromthe hexamethylenetetramine, is converted into its compound withmethone. Photomicrographs of the distinctive precipitates fromformaldehyde and acetaldehyde are on record.37Ethyl alcohol. This, the most important of the compoundsunder consideration, is most frequently determined by physicalmethods, e.g., specific gravity and refractivity. For this purpose,other substances must be eliminated, and for many of such of theseas are not removed by distillation, the petroleum spirit separationwas devised by T.E. Thorpe and J. H01mes.~~ Amongst compoundswhich are, at most, only partially removed by this method aremethyl alcohol, isopropyl alcohol, and acetone ; these are consideredin other sections.Of chemical methods, the process most used is that dependingupon oxidation with chromic acid; among the applications of thisprocess may be mentioned the determination of alcohol in urine,particularly in relation to the amount of alcoholic beverage con-s~rned,3~ in alcoholic liqu~rs,~O and in breath.41 s.G . Liversedge 42holds that the results, for small amounts of alcohol, are comparablewith those derived from specific gravity, upon which they providea useful check.A somewhat different oxidation process is the quantitativeapplication of the reagent (dichromate in nitric acid) devised byH. Agulhon43 for the detection of small quantities of alcohol andaldehyde in acetone; a positive reaction is given by primary andsecondary alcohols, negative by tertiary alc0hols.4~ The colourchanges of the reagent itself are used to indicate the progress ofthe reaction, whose most important application is to the deter-mination of ethyl alcohol in acetone.45isoPropyl alcohol.A rapid method of determining the proportionof isopropyl alcohol in ethyl alcohol depends upon the varying37 H. Leffmann and C . C . Pines, Bull. Wagner Free Inst., 1929, 4, 15; A.,1929, 1042.as J., 1903, 83, 314.40 L. Semichon and M. Flanzy, Ann. Falsif., 1929, 22, 139; B., 1929, 449;ibid.,p. 414; A., 1929,1266; ibid., 1930,23,347; B., 1930,1001; E. Kostuk,Bull. Assoc. Chern. Sucr., 1930, 47, 231 ; B., 1930, 1087.J. Evans and A. 0. Jones, Analyst, 1929, 54, 134; A., 1929, 601.41 L. Smith, Svensk Kern. Tidslcr., 1931, 43, 83; A., 1931, 990.42 Analyst, 1931, 56, 595; A., 1931, 1267.43 Ann. Chim. analyt., 1912, 17, 50; B., 1912, 255.44 Cf. W. R. Fearon and D. M. Mitchell, Analyst, 1932, 57, 372; A., 1932,45 E. C. Craven, J . SOC. Chem.Ind., 1933,52,239~; B., 696.718ELLIS. 279solubility of sodium hydroxide in the mixture; the water contentshould be adjusted to not greater than An apparatus andprocedure have been devised for the isolation of pure anhydrousalcohol from extremely dilute solutions.47isoPropy1 alcohol is gradually extending its scope as a substitutefor ethyl alcohol, not only in analytical operation^,^^ but in numerouscommercial preparations. Apart from a few direct colour reactions,such as those with piperona149 or m-nitrobenzaldehyde 50 and sul-pliuric acid, the majority of methods for detecting as well as fordetermining this alcohol depend on a preliminary chromic acidoxidation, tests being completed on the acetone formed thereby.Qualitatively, this compound may then be detected by means ofnitroprusside 51 or by condensation with o-nitrobenzaldehyde toi n d i g ~ t i n , ~ ~ though it has been pointed out that certain othercompounds containing the group -COMe also respond to the latter.%This test, due to PenzoldtYu has been applied quantitatively in theanalysis of mixtures of acetone, ethyl alcohol, and isopropylalcohol; 55 conditions are given for the chromic oxidation of thealcohols to acetic acid and acetone, respectively, which may bedetermined in the distillate.Another method of determiningacetone is that involving the use of hydroxylamine hydrochloride 56which is quite satisfactory in the presence of isopropyl alcohol.57On the whole, either of the above methods is preferable to Messinger'sprocess in the presence of methyl, ethyl, or isopropyl alcohols,since these consume a certain amount of iodine.58 Optimum46 I?.M. Archibald and C. M. Beamer, Ind. Eng. Chem. (Anal.), 1932, 4,18; B., 1932, 250.47 A. 0. Gettler, J. B. Niederl, and A. A. Bonedetti-Pichler, J. Arner. Chem.SOC., 1932, 54, 1476; A., 1932, 958.4t3 Cf. H. A. Schuette and M. P. Smith, Ind. Eng. Chenz., 1926, 18, 1242;B., 1927, 117; H. A. Schuette and L. E. Harris, J. Ainer. Pharm. ASSOC.,1926,15,166; B., 1927,118.49 G. Reif, 2. Unters. Lebensm., 1928, 55, 204; B., 1928, 686; J. Wiihrer,€'harm. Ztg., 1930, 75, 845; B., 1930, 9SO.50 T. Boehm and K. Bodendorf, Arch. Pharm., 1930, 268, 249; U., 1930,631.5 1 D. Henville, Analyst, 1928, 53, 416; B., 1928, 726; 0.Noetzel, Z .Unters. Lebensm., 1927, 53, 388; B., 1927, 668.53 R. Raw, J. SOC. Cliem. Ind., 1932,51,276~; A., 1932,1235.53 R. J. W. Le FBvre and J. Pearson, ibid., 433T; A., 377.54 Deut. Arch. klin. Med., 34, 127.5 5 C. A. Adams and J. R. Nicholls, Analyst, 1929, 54, 2 ; B., 1929, 162.5 6 Cf. 0. Noetzel, loc. cit., ref. (51) ; M. KrajEinovic, Arhiv Hemiju, 1932,5 7 W. H. Simmons, Perf. & Essent. Oil Rec., 1927, 18, 168; B., 1927, 504.5 8 L. F. Goodwin, J . Arner. Chern. Soc., 1920, 42, 39; B., 1920, 280;6, 161 ; B., 11.H. A. Cassar, Ind. Eng. Chern., 1927, 19, 1061; A . , 1927, 1100280 ANALYTICAL UHEMISTRY.conditions for the detection of acetone, acetaldehyde, ethyl alcohol,and lactic acid by the iodoform test have been pre~cribed.~~Miscellaneous.-Individually, many of the chlorinated aliphatichydrocarbons, e.g., chloroform,60 can be determined in terms of thechloride formed by saponification.This method cannot be appliedto mixtures of chloroform with certain other members of the group ;to such, the colorimetric test of S. Lustgarten with @-naphthol canbe applied quantitatively.G2 Amongst some microchemical testsfor saccharin may be noted in particular that with copper andpyridine; 63 this reagent has also been applied to caffeine, theo-bromine, and theophylline64 and to compounds of the veronalfamily; G5 for the last group, the formation of the complex sub-stances has been adapted quantitatively in toxicological work.66Reference was made at some length in last year's Report to theapplications of organic compounds to inorganic analysis ; thisimportant class of reagent continues to attract attention, includingin particular 8-hydroxyquinoline.This compound serves a8 aprecipitant for tungstete 67 from acetic solution; in the presenceof stannic salts, the precipitation should be effected in the presenceof ammonium oxalate.68 The presence of oxalate is also recom-mended in the precipitation of niobium.6g Among the other rarermetals which have been investigated are gallium 70 and indium.71The utility of the reagent in recovering aluminium after removalof, or in the presence of, other metals or of phosphate has again beendem~nstrated.~~ The effect of p , on the precipitation of magnesium,zinc, cobalt, nickel, copper, and molybdenum irom acetate solutions5a I.M. Korenman, 2. anal. Chem., 1933, 93, 335; B., 821.6o G. D. Beal and C. R. Szalkowski, J. Amer. Pharm. Assoc., 1933, 22,61 Monatsh., 1882, 3, 715; A., 1883, 243.62 W. G. Moffitt, Analyst, 1933, 58, 2.63 M. Wagenaar, Mikrochem., 1932, 11, 132; Pharm. Weekblad, 1932,69, 614; A., 1932, 763, 867; J. J. L. Zwikker, ibid., 1933, 70, 551; A.,732.540 ; B., 764.64 L. Rosenthaler, Pharm. Zentr., 1933, 74, 288 ; A., 732.6 5 J. J. L. Zwikker, Pharm. Weekblad, 1931, 68, 975; A., 1931, 1328.6 6 J. Peltzer, Chm.-Ztg., 1933, 57, 816.6 7 S. Halberstadt, 2. anal. Chm., 1933, 92, 86; A., 366.6 8 A. Jilek and A. RygQnek, Coll. Czech. Chem. Comm., 1933, 5, 136; A.,584.P. Sue, Compt. rend., 1933,190, 1022; A., 585.L.Moser and A. Brukl, Monatsh., 1929, 51, 76; W. Geilmann andF. W. Wrigge, 2. anorg. Chem., 1932,209, 129; A , , 43.71 Geilmann and Wrigge, loc. cit.72 J. Haslam, Analyst, 1933, 58, 270; A., 688; R. Lang and J. Reifer,2. anal. Chem., 1933,93,161; A., 799; G. Balanescu and (Mlle.) M. D. Motzoc,ibid., 1932, 91, 188; A., 138ELLIS 281has been carefully examined, and certain separations have beenbased on the findings under the specified conditions.73 Calciumoxalate does not interfere with the bromometric titration of themagnesium compound ; the freshly ignited mixture of oxidesobtained on ignition of the mixed precipitates yields its lime to 30%saccharate s0lutions.7~7-Iodo-8-hydroxyquinoline-5-sulphonic acid has been used as areagent for the colorimetric determination of ferric Othersubstituted derivatives have been found of value qualitatively ;for instance, the 5 : 7-dibromo-compound reacts specifically forvanadium in fairly concentrated nitric acid solution, provided ferriciron has been removed,77 whilst certain azo-derivatives are of valuein detecting bivalent mercury 78 and magnesiu~n.~~ The crystallinecompound given by hydroxyquinoline-iodide reagent serves todetect bismuth a t a dilution of one part in a million.80Precipitation of (Cu en,)(HgI,) in the presence of tartrate permitsseparation of mercury from numerous other heavy metals.81 Thecomplexes (Cu pn,)(AgI,), and (BiI,)[Co en(SCN),] serve in therapid macro- and micro-determination of silver and bismuthrespectively.82In the separation and determination of copper and nickel bymeans of salicylaldoxime, for which details of procedure are given,83it is not necessary to isolate the reagent.s4Addition of l-nitro-2-naphthol to a hot solution containing dilutesulphuric acid precipitates cobalt salts but not those of nickel oriron.85 Accurate results in the determination of cobalt by means ofdinitrosoresorcinol are stated to be fortuitous, the precipitateli3 H.R. Fleckand A.M. Ward, Analyst, 1933,58,388; A., 922.74 J. C. Redmond, Bur. Stand. J . Res., 1933, 10, 823; B., 749.7 5 A. C. Shead and R. K. Valla, Ind. Eng. Chem. (Anal.), 1932, 4, 246; d.,7 6 J. H. Yoe, J . Amer. Chem. SOC., 1932, 54, 4139; A., 43.7 7 G. Gutzeit and R. Monnier, Helv.Ghim. Acta, 1933, 16, 239; A.,78 Idem, ibid., p. 233; A., 479.70 G. Gutzeit, R. Monnier, and R. Bachoulkova-Brun, Arch. Sci. phys.80 R. Sazerac and J. Pouzerguos, Compt. rend. SOC. Biol., 1932, 109, 79;81 G. Spacu and G. Suciu, 2. anal. Chem., 1933,92,247; A., 478.82 G. Spacu and P. Spacu, ibid., 1932, 90, 182; A., 42; 1933, 93, 260;83 H. L. Riley, J., 1933, 895.84 S. Astin and H. L. Riley, ibid., p., 314.85 A. Herfeld and 0. Gerngross, 2. anal. Chem., 1933, 94, 7; A., 1025.8 6 W. R. Orndorff and M. L. Nichols, J . Amer. Chem. SOC., 1923,45,1439;1932, 588.479.nat., 1933, [v], 15, Suppl., 203; A., 1133.A., 139.A., 924.A., 1923, ii, 684282 ANALYTICAL CHEMISTRY.readily adsorbing reagent and alkali.8' A similar effect is recorded *'in the precipitation of cadmium by hexamethylenetetramine ally1iodide ; 89 the use of p-naphthaquinoline in the presence of iodideand of phenyltrimethylammonium iodide has been successfullyapplied to the determination of cadmium in spelter and zincores.g1A detailed investigation of the beliaviour of small amounts ofnickel in rock analysis indicated the desirability of determiningthis metal separately and of making allowance for the amountretained in the R,O, precipitate; for very small quantities, a-furildioxime may advantageously replace dirnethylgly~xime.~~On account of solubility, nickel dimethylglyoxime should be filteredcold and washed with cold water ; 93 the optimal pH for precipitationis 7.5--8.1.9* Nickel dicyanodiamidine 95 can be titrated withstandard acid.96 Zinc, cadmium, cobalt, nickel, and copper saltsare quantitatively precipitated from neutral aqueous solutions bysodium anthranilate.972 : 2'-Dipyridyl gives a red coloration with ferrous salts in slightlyacid solution : 98 this reagent has now been applied quantitatively,total iron being determined after reduction by s ~ l p h i t e .~ ~ Porthe colorimetric determination of aluminium with aurintricarboxylicacid, a modified procedure applicable to 0.0001 mg. of the metalhas been described-lThe insolubility of magnesium calcium hexamethylenetetramineferrocyanide is utilised in the quantitative separation of traces ofmagnesium for subsequent colorimetric comparison.2 Qualitatively,87 0. Tomi6ek and K.Komarek, Chem. Listy, 1932,26, 515; 2. anal. Chem.,1932,91,90; A., 1932, 1224.8 8 L. C. Hurd and R. W. Evans, Ind. Eng. Chern. (Anal.), 1933, 5, 16; A.,245.89 V. Evrard, Natuurwetensch. Tijds., 1929, 11, 191 ; A., 1930, 182.91 A. Pass and A. M. Ward, Analyst, 1933, 58, 667.92 H. F. Harwood and L. S. Theobald, ibid., p. 673; cf. B. A. Soule, J .s3 P. Nuka, 2. anal. Chern., 1932, 91, 29; A., 138.94 T. Nagai, Compl. Abs. Jap. Chem. Lit., 1932, 6 , 409.gg Cf. H. Grossmann and B. Schiick, Chesn.-Ztg., 1907, 31, 335; A., 1907,96 J. V. Dubsk$ and E. Hauer, Mikrochem., 1933,12,321; A., 365.9 7 H. Funk and M. Ditt, 2. anal. Chem., 1933, 91, 332; 93, 241 ; A., 244,98 F. Feigl and H. Hamburg, ibid., 1931, 86, 7 ; A., 1931, 1261.99 F. Feigl, P.Krumholtz, and H. Hamburg, ibid., 1932, 90, 199; A., 43;R. Berg and 0. Wurm, Ber., 1927, 60, 1664; A., 1927, 847.Amer. Chern. Soc., 1925, 47, 981 ; A , , 1925, ii, 603.ii, 582.924.H. Miiller, Mikrochem., 1933, 12, 307; B., 366.1 P. S. Roller, J . Amer. Chem. SOC., 1933, 55, 2437; A., 799.2 L. Debucquet and L. Velluz, Compt. rend., 1933,166,2006 ; A., 922OLASSTONE. 283hexamethylenetetramine has been used to separate iron, aluminium,and chromium from metals of the ammonium sulphide group.3Diphenylthiocarbazone (dithizone), which has been investigatedas a reagent for the detection of heavy has now been appliedto the microchemical determination of lead and of ~ o p p e r . ~ Chlor-ides may be titrated in dilute acid solution against mercuric nitratewith diphenyl-carbazide or -carbazone as indicator.6A sensitive test for gold with a-naphthylamine is described.'Alcoholic solutions of 2 : 3-diaminophenazine hydrochloride givecoloured precipitates with neutral solutions containing Group I1cations ; * the colorations produced by 1 : 2-diaminoanthraquinone-3-sulphonic acid with copper, cobalt, and nickel salts are due toadsorption compo~nds.~ Antipyrine and potassium iodide may beused to differentiate between ter- and quinque-valent antimony.1°B. A.E.Electrometric Methods in Analytical Chemistry.Pot entimetric Methods .-T he fundament a1 basis of potentio -metric, as of ordinary, titration is that in the vicinity of the end-point (or equivalence-point) there is a very rapid change in theconcentration of the titrant or of the substance titrated.Thepotentiometric method depends on the fact that the potentialacquired by a suitable indicator electrode depends on the concentra-tion of the ions with respect to which the electrode is reversible.At the equivalence-point there is a sudden change of electrodepotential, and the plot of the E.2M.F. of the cell, consisting of theindicator electrode and a reference (e.g., calomel) electrode, againstthe volume of titrant added-the potential-titration curve-shows a marked inflexion. In general, it is assumed that theslope of this curve, AE/Au, where AE is the change of potentialresulting from the addition of a definite volume Au of titrant, isa maximum exactly at the required end-point.This is, however,only strictly true in certain limited cases, vix., in the neutralisationof a strong acid and strong base, and in the precipitation of a salt3 L, Lehrman, E. A. Kabat, and H. Weisberg, J. Amer. Chem. Soc., 1933,55, 3509; A., 1133; cf. P. Ray, 2. anal. Chem., 1931, 86, 13; A., 1931, 1261.4 H. Fischer, Mikrochem., 1930, 8, 319; A., 1931, 328; Angew. Chem.,1933, 46, 442, 517; A., 799; H. Wolbling and B. Steiger, ;bid., 279; A., 798.5 H. Fischer and G. Leopoldi, Wiss. Ver6.ff. Siemens-Konx., 1933, 12, No, 1,44, 52 ; A., 923.6 J. V. Dubsky and J. Trtilek, Milcrochem., 1933, 12, 315; A., 364.7 H. Holzer and W. Reif, 2. anal. Chem., 1933, 92, 12; A., 366.8 T. Pavolini, Ind. chim., 1933, 8, 692; A , , 799.9 J. V. Dubskf and V.Bencko, 2. anal. Chem., 1933,94, 19; A , , 1025.10 P. Duq&noi,s, Compt. rend., 1933,197,339; A., 921284 ANALYTICAL CHEMISTRY.of symmetrical valency type.1 When a weak acid is neutralisedby a strong base, the pH at the point for which AE/Av is a maximumis less than at the equivalence-point by 0.651/K,/cKa, where c isthe acid concentration and Kw and Ka are the ionisation constantsof water and acid, respectively.2 The accuracy with which thepoint of maximum A.E/Av can be identified depends on the mag-nitude of the inflexion in the potential-titration curve ; the smallerare c and K, the smaller is the value of AElAv, and if cK, is lessthan 27K, no infiexion occurs in the curve.3 In precipitationreactions the inflexion will be before or after the true equivalence-point according as the ratio of the titrant ion to that of the titratedis greater or less than unity.4 The titration error is less, and themagnitude of the inflexion greater, the lower the solubility of theprecipitated salt; in some instances this is decreased by theaddition of an organic liquid.For oxidation-reduction titrationsthe equivalence-point and that of maximum AE/Av can theoreticallynever coincide, but the deviation is less the larger the equilibriumconstant of the reaction; the value of AE/Av a t the end-pointincreases with this constant and the concentration of the s~lution.~It is evident that, unless results of the highest accuracy are requiredor very dilute solutions are being analysed,s if an acid is reasonablystrong, a precipitate not too soluble, or an oxidation-reductionreaction fairly complete, the end-point of a titration may be foundfrom the point of inflexion in the potential-titration curve, orbetter from the plot of AE/Av against v , ~ provided Av representsepual volumes of titrant added.lO For precision work theE.D. Eastman, J. Amer. Chem. SOC., 1925, 47, 332; A., 1925, ii, 594;I. M. Kolthoff and N. H. Furman, “ Potentiometric Titrations,” 2nd Edtn.,1931, p. 19.a P. S. Roller, J. Amer. Chern. SOC., 1928, 50, 1; 1932, 54, 3485; A., 1928,262; 1932, 1101.Idem, ibid. ; see also F. Auerbach and E. Smolczyk, 2. physikal. Chem.,1924,110, 6 5 ; A., 1925, ii, 118.F. L. Hahn, M. Frommer, and R. Schulze, ibid., 1925, 133 390; A,,1928, 857; B. Cavanagh, J., 1930, 1425; A., 1930, 1142.E.Lange and E. Schwartz, 2. physikal. Chem., 1927, 129, 111; A., 1927,1029; S. Glasstone, “ The Electrochemistry of Solutions,” 1930, p. 347.6 J. A. Atanasiu, Compt. rend., 1926, 182, 519; A., 1926, 376; J . Chim.physique, 1926, 23, 501; K. Schwartz and C. Schlosser, Mikrochem., 1933,13, 18; A., 582.7 I. 31. Kolthoff and N. H. Furman, op. cit., p. 55; S. Glasstone, op. cit.,8 Compare E. Lange and R. Berger, 2. Elektrochem., 1930, 36, 980; A.,9 J. C. Hostetter and H. S. Roberts, J. Amer. Chem. SOC., 1919, 41, 1337;10 F. L. Hahn and 31. Frommer, Z. physikal. Chem., 1927, 127, 1; A.,p. 348.1931, 186.A., 1919, ii, 480.1927, 743GLASSTONE. 285equivalence-point may be determined by calculation from potentialsmeasured in its vicinity,l’ or from the values determined accuratelyat three or four points in the course of the titration.12 A graphicalmethod for determining the end-point on the potential-titrationcurve, which involves plotting its evolute,13 has been proposed,but it is doubtful if it has any advantages.14 Other special methodsfor finding the point at which AE/Av is a maximum are discussedlater in this Report.Indicator and Reference Electrodes.-The indicator electrodedepends on the nature of the titration: for neutralisation pro-cesses some form of hydrogen electrode is required.The usesand limitations of the hydrogen gas, oxygen, and quinhydroneelectrodes are well ~ I I O W ~ , ~ ~ although there have been some recentstudies of technique,16 but the glass electrode is worhh specialmention.The potential of this electrode follows that of the hydro-gen electrode very closely over the range of pH 0-9 and approxim-ately up to pH 12 ; l7 in very acid or alkaline solutions deviationsare observed,18 influenced sometimes by the cations in the solution.19l1 F. L. Hahn and M . Frommer, 2. physikal. Chem., 1927,127, 1 ; see alsoF. L. Hahn, M. Frommer, and R. Schulze, loc. cit.; F. L. Hahn, ibid., 1930,146, 363; A., 1930, 560; 2. anal. Chem., 1932, 87, 263; A., 1932, 353;F. b. Hahn and R. Klockmann, 2. physikal. Chem., 1930,146, 373 ; 1931,157,203, 206, 209; A., 1930, 560; 1932, 33, 33, 24.l2 B. Cavanagh, loc. cit., ref. (4) ; idem, J., 1928, 843, 855 ; A., 1928, 607 ;F.Fenwick, I n d . Eng. Chem. (Anal.), 1932,4,144; A., 1932, 241.l3 F. L. Hahn and G. Weiler, 2. anal. Chem., 1926, 69, 417; A., 1927, 124.14 I. M. Kolthoff, Rec. trccv. chim., 1928, 4’7, 397 ; A., 1928, 496.15 W. M. Clark, “ The Determination of Hydrogen Ions,” 1928; H. T. S.Britton, “ Hydrogen Ions,” 2nd Edtn., 1932 ; S. Glasstone, op. cit,l6 A. Thiel and 0. Schulz, 2. Elektrochem., 1930, 36, 408; A., 1930, 1009;L. Fletcher and J. 13. Westwood, J. Soc. Chem. Ind., 1930, 49, 2 0 1 ~ ; A.,1930, 1009; R. J. Best, J . Physical Chem., 1930, 34, 1815; A., 1930, 1124;J. L. R. Morgan, 0. M. Lammert, and M. A. Campbell, J . Amer. Chem. SOC.,1931, 53, 454, 597; A., 1931, 456; idem, Trans. Amer. Electrochem. Soc.,1932,61, 199; A., 1932, 471; J.L. R. Morgan and 0. M. Lammert, J . Amer.Chew,. Soc., 1931, 53, 2154; A., 1931, 914; 0. M. Lammert and J. L. R.Morgan, ibid., 1932, 54, 910 ; A., 1932, 699 ; I. M. Kolthoff and T. Kameda,ibid., 1931, 53, 821; A., 1931, 585; B. Grotik, Biochem. Z., 1932, 245, 61;A., 1932, 486; P. Jolibois and G. Fouretier, Cornpt. rend., 1932, 194, 1072;A., 1932, 486; L. Pincussen and J. Gorne, Biocibenz. Z., 1932, 249, 126; A.,1932, 924; W. Kordatski and P. Wulff, 2. anal. Chem., 1932, 89, 241; A.,1932,1013; P. L. du Nouy, Compt. rend., 1932,195,1265; A., 242; J. A. V.Butler and G. Armstrong, Trans. Paraday SOC., 1933, 29, 862 ; A., 1022.17 W. S. Hughes, J., 1928, 491; A., 1928, 370; D. A. MacInnes andD. Belcher, J . Amer. Chern. SOC., 1931, 53, 3315; A,, 1931, 1264.18 M.Dole, ibid., 1931, 53, 4260; A., 1932, 126, 1207;C. Buchbock, 2. physikal. Chern., 1931,156, 232; A., 1931, 1237.19 w. S. Hughes, ZOC. cit.; M . Dole, ZOC. tit.; F. Urban and A. Steiner,J . Physical Chem., 1931, 35, 3058; A., 1931, 1264; s. I. Sokolov and H. A.Passinski, Z . physikat. Chem., 1932, 160, 366; A., 1932, 915.1932, 54, 3095286 ANALYTICAL CHEMISTRY.The glass electrode can be used where other forms of the hydrogenelectrode fail,20 but it does not function satisfactorily in ethylalcohol, in acetic acidY21 or in the presence of gelatin.22 Lowmelting-point glass of high electrical conductivity 23 is generallyused in the form of a thin bulb or film,24 but the resistance maystill be of the order of 10-100 megohms.A quadrant electro-meter 25 or a ballistic galvanometer,26 in conjunction with a potentio-meter, or a similar galvanometer, standardised by means ofsolutions of known pHY27 have been used to measure the E.M.F.'sof glass-electrode cells ; the most convenient methods, however,involve thermionic valve circuits.28 The high resistance of the2o D. A. MacInnes and 3%. Dole, Ind. Eng. Chem. (Anal.), 1929, 1, 57; A.,1929, 673; L. W. Elder, Trans. Amer. Electrochem. SOC., 1930, 57, 383; A.,1930, 565; F. Hazel and C. H. Sorum, J . Amer. Chem. SOC., 1931, 53, 49;A., 1931, 304; C. Morton, Trans. Faraday Xoc., 1932, 28, 84; A., 1932, 342;H. T. S. Britton and R. A. Robinson, ibid., p. 531; A., 1932, 709; H. T. S.Britton and E. N. Dodd, J., 1932, 1940; A., 1932, 814; H.T. S. Brittonand (Miss) B. M. Wilson, ibid., p. 2550; A., 1932, 1207; G. 3'. Davidson,J . Text. Inst., 1933, 2 4 ~ , 185; B., 587; J. W. Ingham and J. Morrison, J.,1933, 1200; A., 1118; T. F. G. Hepburn, J . SOC. Leather Trades Chem., 1933,17, 268 ; A., 689.21 M. Dole, J. Amer. Chem. SOC., 1932, 54, 3095; A., 1932, 1207.R. J. Hartman and I. F. Fleischer, J . Physical Chem., 1932, 36, 1136;A., 1932, 700.23 W. S . Hughes, loc. cit.; L. W. Elder, J . Amer. Chem. SOC., 1929, 51,3266; A., 1930, 50; D. A. MacInnes and M. Dole, ibid., 1930, 52, 29; A.,1930,423.24 Idem, locc. cit.; G. B. Harrison, J., 1930, 1528; A., 1930, 1151; G. R.Robertson, Ind. Eng. Chem. (Anal.), 1931, 3, 5 ; A., 1931, 449; G. D. Grevilleand N. F. Maclagen, Trans.Faraday Soc., 1931, 27, 210; A., 1931, 801;M. R. Thompson, Bur. Stand. J . Res., 1932, 9, 833; A., 367; E. C . Gilbertand A. Cobb, Ind. Eng. Chem. (Anal.), 1933, 5, 69 ; A., 248 ; D. A. MacInnesand D. Belcher, ibid., p. 199; A., 689; J. W. Ingham and J. Morrison, Zoc.cit .25 (Mrs.) P. M. T. Kerridge, J . Sci. Instr., 1926,3,404; A., 1926, 1115; L. W.Elder, J . Amer. Chem. SOC., 1929, 51, 3266; A., 1930, 50; D. A. MacInnesand D. Belcher, ibid., 1931, 53, 3315; A., 1931, 1264.26 G. Jones and B. B. Kaplan, ibid., 1928, 50, 1845; A., 1928, 954;M. Dole, ibid., 1931, 53, 620; A., 1931, 456.2' C . Morton, J. Sci. Instr., 1930, 7, 187; A., 1930, 1009; this method isonly satisfactory if the temperature of the glass electrode is carefully con.trolled (private communication from Dr.C. Morton).38 W. C. Stadie, J . Biol. Chem., 1929, 83, 477; A., 1929, 1262; G. B.Harrison, loc. cit. ; G. Schwarzenbach, Helv. Chim. Acta, 1930, 13, 865 ; A.,1930, 1526; R. J. Fosbinder, J . Physical Chem., 1930, 34, 1294; A., 1930,883; R. J. Fosbinder and J. Schoonover, J. Biol. Chem., 1930, 88, 605; A.,1930, 1376; D. Dubois, ibid., p. 729; A., 1931, 58; F. Muller, 2. EleJctr.ochem.,1930, 36, 923; A., 1931, 43; idem, Z. physikal. Chem., 1931, 155, 451; A.,1931, 1129; idem, 2. angew. Chem., 1931, 44, 698; A., 1931, 1144; idem,Trans. Amer. Electrochem. SOC., 1932, 62, 117 ; A., 1932, 999 ; F. Muller anGLASSTONE. 287glass electrode may alter the characteristics of the valve and alsointroduce an error owing to the fall of potential due to the flowof grid current.The introduction of '' electrometer " valves withvery high grid to filament resistance and low grid current 29 permitsthe use of glass electrodes with resistances of the order of 1000megohms. In spite of its relative insensitivity, an accuracy of&O-OEi millivolt has been claimed with the electrometer triodevalve ; 30 this is probably greater than the accuracy with which theliquid junction in the cell can be reproduced. For automaticregistration purposes or to permit the use of robust instruments,the electrometer valve may be followed by further amplifyingvalves.31 Another procedure for overcoming the resistance errors,not requiring special valves, is to balance the circuit first withthe glass-electrode cell in one direction and then reversed, so thatthe grid circuit resistance remains constant; the mean of the twopotentiometer readings gives the correct E.M.F.of the cell.32Whatever circuit is used it is essential that the grid current shouldbe very small (less than 10-lo amp.), otherwise the electrode willpolarise and yield erroneous results. Some workers operate thethermionic valve at its " floating grid " potential; the grid currentis then zero and the anode current independent of the resistanceof the glass electrode.33 In this state the valve is, however, veryG. Meyer, 2. Elektrochern., 1932, 35, 418; A., 1932, 814; C. Morton, J., 1931,2977, 2983; 1932, 2469; A., 1932, 24, 1105; idem, J . SOC. Chem. Ind., 1931,50, 4 3 6 ~ ; A., 1932, 138; idem, J.Sci. Instr., 1932, 9, 289; A , , 1932, 1105;C. Morton and F. L. Best, J . SOC. Chem. Ind., 1933, 52, 6T; A., 366; S. E.Hill, Science, 1931, '73,529; A., 1931, 928; G. D. Greville and N. F. Maclagan,Zoc. cit., ref. (24); W. C. Stadie, H. O'Brien, and E. P. Laug, J . BioZ. Chern.,1931,91,243; A., 1931, 754; A, E. J. Vickers, J. A. Sugden, andR. A. Bell,C'henh. and Ind., 1932, 545, 570, 923; A., 1932, 828; H. M. Partridge andS. J. Broderick, Mikrochem., 1932, 11, 337; A., 1932, 1013; K. G. Comptonand H. E. Haring, Trans. Arner. Electrochem. SOC., 1932, 82, 195; A., 1932,1013; F. Rosebury, Ind. Eng. Chem. (Anal.), 1932, 4, 398; A., 1932, 1225;R. Nordba, Tidsskr. Kjeini Berg., 1933, 13, 62; A., 689; E. C.Gilbert andA. Cobb, Zoc. cit. ; A. S. McFarlane, J . Sci. Instr., 1933, 10, 208; A., 926.G. M. Metcalf and B. J. Thompson, Physical Rev., 1930, 36, 1489;G. B. Harrison, Zoc. cit., ref. (24); C . Morton, J., 1931, 2977; 1932, 2469; A.,1932, 24, 1105; C. Morton and F. L. Best, Zoc. cit., ref. (28); G. D. Grevilleand N. F. Maclagan, Zoc. cit., ref. (24); s. E. Hill, loc. cd., ref. (28); L. A.DuBridge, Physical Rev., 1931, 37, 392; F. Muller and G. Meyer, Zoc. cit.,ref. (28) ; F. Rosebury, Zoc. cit., ref. (28).30 G. D. Greville and N. F. Maclagan, loc. cit. ; A. S. McFarlane, Zoc. cit.31 C . Morton, J . , 1931, 2977; 1932, 2469; A., 1932, 24, 1105; J. J. Fox32 G. Schwarzenbach, Zoc. cit.; C . Morton, J., 1931, 2983; A., 1932, 24.33 R. J. Fosbinder; R.J. Fosbinder and J. Schoonover; D. Dubois, FMiiller, W. C. Stadie, H. O'Brien, and E. P. Laug ; H. M. Partridge and S. J.Broderick, locc. cit.and L. G. Groves, J . SOC. Chem. Ind., 1932, 51, 7T; B., 1932, 323288 ANALYTIC& CHEMISTRY.unstable, and slight positive deviations of the grid potential causea large flow of grid current and serious polarisation may result.34Various devices have been used to eliminate galvanometer drift ; 35one of the simplest and most effective is to connect a ballisticgalvanometer, acting as a null instrument, in series with a blockingcondenser across a resistance in the anode Accurateresults with the glass electrode are obtained by using a pure sodium-calcium-silicate glass, by avoiding leakages over the surface of theglass and in the grid circuit,37 and by providing adequate shielding.38The Sb/Sb,O, electrode, the antimony electrode, is attractingattention as a hydrogen-ion indicator; it is of the metal-metaloxide type39 and so should behave theoretically as an oxygenelectrode, its potential depending on the pE of the solution.Thereis generally sufficient oxide on the surface of cast antimony not torequire its addition to the solution,4° but some workers preferto add the oxide in ti crystalline form.*l Antimony deposited bythe electrolysis of its trichloride in acetone is said to give very satis-factory In the range of pH 2-7 antimony electrodescan behave reversibly, but in more alkaline solutions deviationsoccur; 43 they may still, however, be used for potentiometrictitrations.44 The antimony electrode is not readily poisoned 45and is finding increasing application,46 although it does not always34 C.Morton, Zoc. cit.35 G. Schwarzenbach; W. C. Stadie, H. O’Brien, and E. P. Lsug; A. E.36 C. Morton, J., 1931, 2977; A., 1932, 24.37 G. D. Greville and N. F. Maclagan, loc. cit. ; H . Kahler and F. de Eds,J . Amer. Chem. SOC., 1931, 53, 2998; A., 1931, 1143; B. S. Platt and S.Dickinson, Biochem. J., 1933,27,106!3 ; A., 1027 ; also private communicationfrom Dr. C. Morton.J. Vickers, J. A. Sugden, and R. A. Bell; F. Muller and G. Meyer, locc. cit.a * R. J. Fosbinder ; G. D. Greville and N. F. Maclagan, locc. cit.39 L. R. Parks and H. C. Beard, J. Physical Chem., 1933,37,821,822 ; A., 913.40 Idem, J .Amer. Chem. SOC., 1932, 54, 856; A., 1932, 420 (contains anexcellent summary).4 1 E. J. Roberts and F. Fenwick, ibid., 1928, 50, 2125; A., 1928, 1098;H. T. S. Britton and R. A. Robinson, J., 1931, 458; A., 1931, 585.42 I. I. Shukov and G. P. Avsejevitsch, 2. Elektrochern., 1929, 35, 349;A., 1929, 899; G. P. Avsejevitsch and I. I. Shukov, ibid., 1931, 37, 771; A.,1931, 1370.43 L. R. Parks and H. C. Beard, loc. cit.; T. Uemura and H. Sueda, Bull.Chem. SOC. Japan, 1933, 8, 1 ; A., 362.44 I. I. Shukov end V. M. Gortikov, 2. Elelctrochem., 1929, 35, 853; A.,1930, 6 0 ; H. T. S. Britton and R. A. Robinson, loc. cit.; B. B. Malvea andJ. R. Withrow, J . Amr. Chem. SOC., 1932, 54, 2243; A., 1932, 826;E. Vellinger, Compt.rend., 1932,194, 1820; A., 1932, 696; A. Roche and J.Roche, Arch. Phys. biol. Chirn.-phys., 1932, 9, 273; A., 363.45 M. Catenacci, L’Ind. sacc. Ital., 1931, 24, No. 8 ; idem, Int. Sugar J.,1932,34,185 ; A., 1932,586.46 F. Fenwick and €3. Gilman, J . Biol. Chene., 1929, 84, 605; A., 1930GLASSTONE. 289function sati~factorily.~~ A tungsten wire or rod also behaves asan oxygen electrode ; its potential varies with the pH of the electro-lyte, but not in a reversible manner. It has been used for thepotentiometric titration of a number of acids.4s Other metals, andeven non-metals, acquire more or less definite potentials, but theydo not always adjust themselves rapidly to the changing pa of thee l e c t r ~ l y t e . ~ ~ This has been utilised in the " bimetallic " systemsfor simplified potentiometric titration wherein one metal, or ilnon-metal, acts as reference electrode and the other as indicator.During the titration the E.M.P.of the cell remains almost constant,but a t the end-point one electrode, being more reversible than theother, responds more readily to the change of pH so that there isa marked E.M.P. change. The following bimetallic systems havebeen found suitable for acid-base titrations : platinum and graphite,iron and antimony-cadmium alloy, antimony and antimony-leadalloy, bismuth and silver, antimony and copper amalgam, copperand cupric oxide, tungsten and silicon, copper, nickel, cobalt, orsilver; 5O pure carborundum is said to make a very satisfactoryreference ele~trode.~l The polarised bimetallic system (vide infra)has also been used in neutralisation processes.52For oxidation-reduction reactions smooth platinum is almostinvariably the indicator electrode ; the calomel reference elect,rode122; J.C. Vogel, J . SOC. Chem. Ind., 1930, 49, 2 9 7 ~ ; R. J. Fosbinder, J .Lab. Clin. Med., 1931, 16, 411; A., 1931, 811; W. Bottger and L. vonSzebelledy, 2. Elektrochem., 1932, 38, 737; A., 1932, 1101; T. Gysinck,Arch. Suikerind. Ned.-Indie, 1932, 711 ; A., 1932, 1220; E. Vellinger, Chim.et Ind., 1933, Spec. No. 218; A., 1022.4 7 R. J. Fosbinder, loc. cit.; S. Bodforss and A. Holmqvist, 2. physikal.Chem., 1932, 161, 61; A . , 1932, 999; I. I. Shukov and J. A. Boltunov, J.Gen. Chem. (U.S.S.R.), 1932, 2, 407 ; A,, 1932, 1220.48 J.R. Baylis, Ind. Eng. Chem., 1923, 15, 852; H. C. Parker, ibid., 1925,17, 737; A., 1926, ii, 899; H. T. S. Britton and E. N. Dodd, J., 1931, 829;A., 1931, 699.49 J. 0. Gloss and L. Kahlenberg, Trans. Amer. Electrochem. SOC., 1928,54, 369; A., 1928, 1203; A. Banchetti, Gazzetta, 1932, 62, 999; A., 135.50 J. C. Brunnich, Ind. Eng. Chem., 1925, 17, 631; A,, 1925, ii, 711; R.M. Fuoss, Ind. Eng. Chem. (Anal.), 1929,1, 125; A., 1929, 1034; L. Kahlen-berg and A. C. Krueger, Trans. A m r . Electrochem. SOC., 1929, 56, 201 ; A.,1929, 1255; M. L. Holt and L. Kahlenberg, ibid., 1930, 57, 361; A,, 1930,724; A. J. French and J. M. Hamilton, Proc. Indiana Acad. Sci., 1931, 40,171; A., 1932, 586; A. Mazzuchelli, Cfazzetta, 1932, 62, 265; A., 1932, 586;A. Banchetti, ibid., 1932, 62, 1011; A., 135; W.Hiltner, Chem.-Ztg., 1933,53, 704; A., 1027; N. H. Furman and G. W. Low, J . Amer. Chem. SOC.,1933, 55, 1310; A., 572.61 B. Kamienski, Z. physikal. Chem., 1928, 138, 345; 1929, 145, 48; A.,1929, 144; 1930, 57.62 H. H. Willard and F. Fenwick, J . Amer. Chem. SOC., 1923, 45, 715; A.,1923, ii, 286; A. H. Wright and F. H. Gibson, Ind. Eng. Chem., 1927, 19,749 ; A., 1927,637.REP.-VOL. xxx. 290 ANALYTICAL CHEMISTRY.may be used, but the development of bimetallic systems oftenmakes this unnecessary. Although platinum rapidly acquires theoxidation-reduction potential of a solution, other metals, e.g.,palladium, gold, tungsten, and amalgamated gold, do not do so;the E.M.F. of a cell consisting of platinum and one of these metalsconsequently changes very rapidly in the vicinity of the end-pointof an oxidation-reduction titrati0n.~3 A further modificationobviates the use of a potentiometer, the end-point being indicatedby the maximum swing of a galvanometer in series with the celland a high resistance.54 For certain titrations i t is even possibleto use a cell consisting of two different platinum electrode^,^^ butmore reliable results are obtained if two similar electrodes arepolarised by a small current (ca.6 x amp.); the E.lI1.H. ofthe polarised cell, measured on a potentiometer, changes sharplyas the equivalence-point is attained.56 If one of the reactants orresultants of an oxidation-reduction process is an efficient de-polariser and the other not, the simple " dead-stop " end-pointmethod may be used.57 A very small E.fM.P.is applied to twoplatinum wires placed in the solution to be oxidised, e.g., thio-sulphate, so as just to balance the polarisation E.M.F. but to perinitno current to flow; on the addition of the oxidant, e.g., iodine,the current remains a t zero until the end-point, when the firstexcess depolarises the cathode and so current flows and the gal-va'nometer is deflected. In a recent modification the grid currentfrom a thermionic valve, which is given a positive bias, polarisesthe electrodes; the change of E.M.P. at the end-point is indicatedby a galvanometer in the anode circuk5*When the titration involves a precipitation reaction the indicator53 J.C. Hostetter and H. S. Roberts, Zoc. cit., ref. (9); H. H. Willard andF. Fenwick, J . Amer. Chem. Soc., 1922, 44, 2504; A., 1923, ii, 33; R. G. vanName and 17. Fenwick, ibid., 1925, 47, 9 ; A . , 1925, ii, 594; N. H. Furman,ibid., 1928, 50,268, 273, 755; A., 1928, 383, 499; I. Lifschitz and M. Reggiani,Gazzetta, 1931, 61, 195; A., 1932, 241.54 N. H. Furman and E. B. Wilson, J . Amer. Chem. Soc., 1928, 50, 277;A., 1928, 382.55 G. G. Reissaus, 2. anal. Chem, 1926, 69, 450; A , , 1927, 126; E. Miillerand H. Kogert, 2. physikal. Chem., 1928, 136, 437; A., 1928, 1203; Z. anal.Chern., 1928, 75, 235; A . , 1929, 42.56 H. H. Willard and F. Fenwick, J. Amer. Chem. Xoc., 1922, 44, 2516;1923, 45, 84, 928, 933; A . , 1923, ii, 33, 187, 436, 430; R.G. van Name andF. Fenwick, ibid., 1925, 47, 19; A., 1925, ii, 594.ii7 C. W. Foulk and A. T. Bawden, ibid., 1926, 48, 2045; A,, 1926, 927;(Miss) M. E. Pring and J-. F. Spencer, Analyst, 1930, 55, 375; A., 1930, 1011 ;see also B. F. Brann and M. €I. Clapp, J . Amcr. Chem. Xoc., 1920, 51, 39;A., 1929, 286.G8 J. L. Kassner, 1%. II. Kunze, and J. N. Chatfield, ibid., 1932, 54, 2278;A., 1932, 814; K. Masaki and 0. Hirabayashi, Bull. Chem. Soc. Japan, 1933,8, 245; A., 1131GLASSTOME. 291electrode is generally of the metal forming the insoluble salt, e.g.,silver or mercury, although certain inert materials can replacesilver in titrations with silver nitrate; 59 the reason for this isobscure.60 Bimetallic systems, both unpolarised 61 and polarised,62have been used to follow the course of precipitation processes.The formation of complex ions can also be adapted to potentio-metric titration, the indicator material being one of the elementspresent in the complex.63Simplijied Methods.-In addition to those mentioned for specialcases, a number are applicable to all types of titration. If a referenceelectrode is used having the potential the indicator electrode willacquire at the end-point, the E.M.P.of the cell will be zero a tthis point, which may be detected by a reversal of current indicatedon a gal~anometer.~~ The reference solutions may be made fromthe actual reactants, or the required potential may be obtainedin other ways; 65 the method is useful when many titrations ofthe same type have to be performed, one reference solution beingadequate.Another method is to oppose the E.H.P. of the titrationcell, consisting of indicator and reference (calomel) electrodes, byone equal to that the cell will have at the end-point.66 A furthermodification, applicable to acid-base and silver-halide titrations,involves the construction of a simple cell, with quinhydrone andsilver halide electrodes, which has zero E.M.P. at the requiredequivalence-point .6759 E. Muller, 2. Elektrochem., 1924, 30, 420; A., 1924, ii, 777.60 W. Bottger and B. M. Schall, 2. physikal. Chem., 1933,165,398 ; A . , 920.61 R. G. van Name and F. Fenwick, J. Amer. Chem. SOC., 1925, 47, 9 ; A , ,1925, ii, 594; J. A. Atanasiu and A. J. Velculesco, 2.anal. Chem., 1931, 85,120; A., 1931, 1260; W. Hiltner, loc. cit.62 H. H. Willard and F. Fenwick, J . Amer. Chem. SOC., 1923, 45, 645; A , ,1923, ii, 332; compare P. Dutoit and G. von Weisse, J . Chim. physique, 1911,9, 578; A . , 1911, ii, 1129.63 E. Miiller and H. Lauterbach, 2. anorg. Chem., 1922, 121, 178; A.,1922, ii, 403; I. M. Kolthoff, Rec. traw. chim., 1922, 41, 172; A., 1922, ii, 388.64 J. Pinkhof, Chem. Weekblad, 1919,16, 1163; A., 1920, ii, 120 (see I. M.Kolthoff and N. H. Furman, op. cit., p. 96); W. D. Treadwell, Helw. Chim.Acta, 1919, 2, 672; A., 1920, ii, 119; W. D. Treadwell and L. Weiss, ibid., p.680; A., 1920, ii, 119; T. Callan and% Horrobin, J. SOC. Chem. Ind., 1928, 47,3 2 9 ~ ; B., 1929, 154; B. L. Clarke and L. A.Wooten, Ind. Eng. Chem. (Anal.),1931,3, 402 ; B., 1931, 241.65 P. F. Sharp and F. H. MacDougall, J . Afner. Chem. SOC., 1922,44, 1193;A , , 1922, ii, 579; H. T. Beans and E. Little, Ind. Eng. Chem., 1925,17, 413;T. Callan and S. Horrobin, Zoc. cit., ref. (64).6 6 H. J. S. SandandD. J. Law, J . SOC. Chem. Ind., 1911, 30, 3; A., 1911,ii, 233 ; E. Muller, loc. cit. ; idem, " Die elektrometrische Massanalyse," 1923,p. 78; see also I. M. Kolthoff, Rec. Iraw. chim., 1928, 47, 397; A., 1928, 496.67 B. Cavanagh, J., 1927, 2207; A . , 1927, 1045; L. V. Wilcox, Incl. Eng.Chem. (Anal.), 1933, 4, 38; A., 1932, 242292 ANALYTICAL CHEMISTRY.The E.M.P. of a cell consisting of two identical electrodes im-mersed in two solutions, representing the titration system beforeand after the addition of a definite volume of titrant respectively,i s a direct measure of AE/Av, and should be a maximum at theend-point ; this is the principle of " differential titration," themaximum being detected by the deflexion of a galvanometer.68In the first applications, two beakers and burettes were used; 6oin the later modifications the electrodes are placed in the samevessel, but by means of special devices a small quantity of thetitrated solution surrounding one of the electrodes is temporarilykept from mixing with the bulk of the solution before each additionof titranL70 These methods are simple, do not require standardcell, reference electrode, or potentiometer, and are capable ofconsiderable accuracy.71 Thermionic-valve circuits for continuousreading of E.M.F.may be used for potentiometric t i t r a t i ~ n , ~ ~and these permit of a special application of the differential method.'SAfter each addition of titrant a compensating current is adjustedto bring the reading of the galvanometer in the anode circuit tozero ; the deflexion resulting from the addition of a definite volumeof titrant (0.05 c.c.) is then a direct-measure of AEJAv.The majority of reactions used in volumetric analysis, andmany others for which suitable indicators are not available, can6 8 B. L. Clarke and L. A. Wooten, J . Physical Chem., 1929, 33, 1468; A.,1929, 1410.69 D. C. Cox, J . Amer. Chem. SOC., 1925, 47, 2138; A., 1925, ii, 999.70 D. A. MacInnes and P. T. Jones, ibid., 1926, 48, 2831; A., 1927, 35;D.A. MacInnes and M. Dole, ibid., 1929, 51, 1119; A., 1929, 666; D. A.MacInnes and I. A. Cowperthwaite, ibid., 1931, 53, 555; A., 1931, 450; D.A. MacInnes, 2. physilcal. Chem., 1927, 130, 217; A., 1928, 36; W. A. Roth,2. Elektrochem., 1927, 33, 127; A., 1927, 533; T. Heczko, 8. anal. Chem.,1928, 73, 404; 1928, 74, 289; 1928, 75, 183; A., 1928, 726, 980, 1345; B.Kamienski, Bull. Acad. Polonaise, 1928, [ A ] , 33; A., 1928, 1345; E. Mullerand H. Kogert, 2. physikal. Chem., 1928, 136, 437, 446; A., 1928, 1203;N. F. Hall, M. A. Jensen, and S. E. Baeckstrom, J. Amer. Chem. SOC., 1928,50, 2217; A., 1928, 977; B. L. Clarke and L. A. Wooten, Zoc. cit., ref. (68);see also H. 13. Willard and A. W. Boldyreff, J . Amer. Chem.SOC., 1929, 51,471 ; A,, 1929,413.7 1 D. A. MacInnes, loc, cit.; D. A. MacInnes and M. Dole, Zoc. cit.7 1 K. H. Goode, J . Amer. Chem. Soc., 1922, 44, 26; 1925, 47, 2483; A.,1922,ii, 307; 1925, ii, 1196; J . 0pt.Soc. Amer., 1928,17, 59; A., 1928, 1109;D. F. Calhane and R. E. Cushing, Ind. Eng. Ghem., 1923, 15, 1118; W. D.Treadwell and C. Paoloni, Helv. Chim. Acta, 1925, 8, 89; A., 1925, ii, 595;H. Bienfait, Rec. trav. chim., 1926, 45, 166; A., 1926, 260; J. W. Williamsand T . A. Whitenack, J . Physical Chem., 1927, 31, 519; A., 1927, 434; C.G. Pope and F. W. Gowlett, J . SOC. Instr., 1927, 4, 380; A., 1927, 1049; seealso P. Jolibois and G. Fouretier, Compt. rend., 1932,194, 872 ; A., 1932, 486.73 C. Morton, Trans. Paraday Soc., 1928,24,14; B.L. Clarke, L. A. Wooten,and K. G. Compton, Ind. Enp. Chern. (Anal.), 1931, 3, 321 ; B., 1931, 849GLASSTONE. 293be studied potentiometrically ; a large number of processes hasalready been listed,74 and in addition titrations of the followingelements, compounds, or mixtures may be noted: borica strong acid in the presence of salts of weak nitrites,77lead,78 ferrous ions,79 osmium,sO hypophosphoric acid,s1 ferro-cyanide by mwcuric, nickel, cobalt, ferric, and silver salts,82 ironin fel~par,*~ platinum and gold,s4 carbon dioxide,s5 hypochloriteand chlorate,86 arsenite,s7 potassium,B* cyanamides and cyanides,s9very dilute solutions of silver, mercury, copper, chloride and iodide,g0chromate and ferricyanideYg1 hydrazine with bromate and iodine,92t ri- and tetra- t h i ~ n a t e s , ~ ~ fluorides,94 sulphat e ,95 ant i r n ~ n y , ~ ~ silver74 I. M.Kolthoff and N. H. Furman, op. cit.; N. H. Furman, Ind. Eng.Chem. (Anal.), 1930, 2, 213; Ann. Reports, 1930, 27, 227; 1931, 28, 210.7 5 L. V. Wilcox, loc. cit., ref. (67).7 6 K. Drewski, Rocz. Chem., 1932, 12, 112; A,, 1932, 472.7 7 E. Jimeno and J. Ibarz, Anal. Pis. Quim., 1932, 30, 128; A., 1932,487.78 A. V. Pamfilov and E. G. IvanCeva, J . Gen. Chem. (U.S.S.R.), 1931, 1,79 N. A. Schischakov, ibid., p. 1012; A., 1932, 575.81 L. Wolf, W. Jung, and L. P. Uspenskaja, 2. anorg. Chem., 1932, 206,125; A., 1932, 710.82 K. Masaki, Bull. Chem. SOC. Japan, 1932, 7, 188; J. A. Atanasiu andA. J. Velculesco, Bull. Chem. SOC. Romdne, 1933, 34, 71 ; A., 924.83 A.K. Lyle, J. Amer. Ceram. SOC., 1932,15, 334; B., 1932, 723.84 E. Muller and K. H. Tiinzler, 2. anal. Chem., 1932, 89, 339; A., 1932,1104; C. del Fresno and E. Mairlot, 2. anorg. Chem., 1933, 214, 73; A., 1134.8 5 P. W. Wilson, F. 0. Orcutt, and W. H. Peterson, I n d . Eng. Chem. (Anal.),1932,4, 357; A., 1932, 1222.86 B. Troberg, Z . anal. Chem., 1932, 91, 161; A., 135; A. Rius and V.Arnal, Anal. Pis. Quim., 1933, 31, 325; A., 1022.8 7 I. C. Schoonover and N. H. Furman, J. Amer. CherrL. SOC., 1933, 55,3123; A., 1023.88 B. P. Nicolski, and I. N. Lavrov, Proc. Lenin'grad Dept. Inst. Fevt., 1933,17, 45; A., 922.760; B., 1932, 259.W. R. Crowell, J. Amer. Chem. SOC., 1932, 54, 1324; A., 1932, 587.H. Sinozaki, J . SOC. Chem.I n d . Japan, 1933, 36, 145; A., 798.O0 R. Spychalski, Rocz. Chem., 1933, 13, 236; A., 798; K. Schwarz, Mikro-chem., 1933, 13, 6; A., 581; K. Schwarz and C. Schlosser, ibid., p. 18; A.,582; K. Schwarz and T. Kantor, ibid., p. 225; A., 799; R. Flatt and A.Boname, Bull. SOC. chim., 1932, [iv], 51, 761; A., 1932, 1009; R. Tronstad,K. Norske Vidensk. Selsk. Forhandl., 1931, 4, 20 ; A., 42.81 C. del Fresno and E. Mairlot, 2. anorg. Chem., 1933, 212, 331 ; A., 800;idem., Anal. Pis. Quim., 1933, 31, 122; A., 924.92 0. Stelling, Scensk Kem. Tidskr., 1933, 45, 4; A., 363.98 F. Ishikawa and T. Murooka, Sci. Rep. T6hoku Imp. Univ., 1932, 21,94 N. Allen and N. H. Furman, J . Amer. Chem. SOC., 1933, 55, 90; A., 242.95 J. A. Atanasiu and A. J. Velculesco, 2.anal. Chem., 1932, 90, 337; A.,96 W. Pugh, J., 1933, 1; A., 246.527; A,, 363.363294 ANALYTICAL CHEMISTRY.c hloricle-t hiosulphat e mixtures, chromic acid, 98 iron, vanadium,and nickel in steels,99 zinc,l and free alkali in phenoxides.2Conductometiic Titrations.-Many reactions involving the form-ation of a sparingly soluble or feebly ionised product, viz., neutralis-ation, precipitation, replacement of a weak by a strong acid orbase, and complex formation, may be followed by measurementof cond~ctivity.~ The end-point is marked by a change in thedirection of the conductivity-titration curve. If the volume ofa dilute solution does not change during the titration, this curveshould consist, in general, of two straight lines intersecting at theend-point.In order to minimise the volume change, the titrant,preferably added from a micro-burette, should be 10-20 times asconcentrated as the solution titrated ; for accurate work a correctionmust be a ~ p l i e d . ~ The end-point is obtained graphically, and sono special precautions need be taken in its vicinity. Solutions asdilute as 0400PN may be litrated with reasonable accuracy pro-vided other electrolytes are absent ; 5 unless carbon dioxide canbe excluded it is necessary to apply a correction in the titrationof weak acids and bases.6 To estimate the precise position of theend-point, the two conductivity lines should cut as acutely aspossible, and the titrants are chosen to this end.7 In the neutralis-ntion of an acid of I<, about 10-4-10-7, a sharper angle is obtainedby titrating with ammonia than with a strong base.s Strong andvery weak (K, ca.10-7-10-11) acids should be neutralised by astrong base, although the intersection of the titration curves doesnot always occur exactly at the equivalence-point .g Moderatelystrong acids (K, ca. 10-2-104) do not give sharp intersections andspecial methods must be used to determine the end-points; lo one97 A. Petit, Bull. SOC. chim., 1932, [iv], 51, 1312; A., 137.98 E. Muller and G. Haase, 2. anal. Chem., 1933, 91, 245; B., 271; N. I.9g P. Dickens and G. Thanheiser, Arch. Eisenhiittenw., 1932-33, 6, 379;1 V. F. Stefanovski, J . Gen. Chem. (U.S.S.R.), 1931,1, 991; B., 1932, 553; . Tananaev, J . Appl. Chem. Russia, 1932, 5, 86; B., 1932, 1084.2 V.A. Kargin and M. I. Usanovich, ibid., p. 458; B., 215.3 1. M. Kolthoff, Ind. Eng. Chem. (Anal.), 1930, 2, 225; A., 1930, 1142.4 E. C. Righellato and C. W. Davies, Trans. Paraday SOC., 1933, 29, 429;5 See, however, G. Jander and H. Schorstein, Angew. Chem., 1932, 45,6 W. Poethke, 2. anal. C h e p , 1931, 86, 45; A., 1931, 1256.* 33. D. Eastman, J . Amr. Chem. SOC., 1925, 47, 332; A., 1925, ii, 594.9 I. M. Koltho€f, Z. anorg. Ckena., 1920, 111, 1 ; A., 1920, ii, 420.10 Idem, ibid.Chlopin, Vestn. Metalloprom., 1932, 19, NO. 3, 74; B., 432.B., 430 ; W. Bohnholtzer, 2. anal. Chem., 1932, 87, 401 ; B., 1932, 430.fA., 242.701 ; A., 1932, 1222.I. M. Kolthoff, loc. cit., ref. (3)GLASSTONE. 295of the best is to titrate partly with ammonia and then with sodiumhydroxide.ll The presence of alcohol does not affect the end-point in acid-base titrations.12 I n precipitation reactions, errorsmay arise because of (a) solubility of the precipitate, ( b ) its speedof formation, (c) its uncertain composition ; the first two difficultiesmay sometimes be overcome by the addition of alcohol,13 and thelast by titrating a hot s01ution.l~The experimental methods have been simplified by recent develop-ments of technique.Thermionic-valve o s ~ i l l a t o r s , ~ ~ silent inoperation, have replaced the induction coil, and for titrations notrequiring very high accuracy A.C. mains may be used; l6 the soundin the telephones may be increased by amplifiers,17 or the telephonecan be replaced by a galvanometer in conjunction with a thermal-junction,18 or with galena,lg carborundum,20 or copper-copper oxide 22 rectifiers.The Wheatstone bridge can be arrangedso as to read directly the conductivity of the titrated solution,23but for many purposes the use of a bridge can be avoided; theA.C. is passed through the experimental cell and the current strengthl1 E. C. Righellato and C. W. Davies, Zoc. cit., ref. (4).l2 W. Poethke, 2. anal. Chem., 1931, 86, 399; A . , 1932, 135.l3 I. M. Kolthoff, ibid., 1922, 61, 171; A., 1922, ii, 452.l4 H. S. Harned, J . Amer. ChemSoc., 1917, 39, 252; 1917, ii, 272.l5 R. E. Hall and L. H. Adams, ibid., 1919, 41, 1515; A . , 1919, ii, 490;H. ulich, 2. physikal. Chem., 1925, 115, 377; A., 1925, ii, 671; J.W. Wool-cock and D. M. Murray-Rust, Phil. Mag., 1928, [vii], 5, 1130; A., 1928, 712;G. Jones and G. M. Bollinger, J . Amer. Chem. SOC., 1929, 51, 2407; A., 1929,1161; E. Gotte and W. Schramek, 2. Elektrochem., 1931, 37, 820; A . , 1931,1387; W. Hiltner, Chem. Fabr., 1931,389,398; A., 1931,387; W. Muchlinsky,ibid., pp. 462, 469; A , , 1932, 138; E. Denina and G. Sella, Ind. chinb., 1932,7, 986; A., 1932, 999.l6 W. D. Treadwell and C. Paoloni, Helv. Chinz. Acta, 1925, 8, 89; A.,1925, ii, 595; T. Callan and S. Horrobin, J. SOC. Chem. Ind., 1928, 47, 329.1.;B., 1929, 154; T. T. Potts, Paper Trade Review, 1931, 95, 1037; Proc. Tech.Sect. Paper 1ialcers’ ASSOC., 1931, 12, 156; B., 1932, 177.17 R. E. Hall and L. H. Adams, Zoc.cit., ref. (15); G. Jones and G. 31.Bollinger, Zoc. cit., ref. (15).18 G. Jander and 0. Pfundt, 2. anorg. Chem., 1926, 153, 219; A . , 1926,700; 2. Eklctrochena., 1929, 35, 206; A,, 1929, 652; “Die visuelle Leit-fiihigkeitstitrationen,” 1929 ; G. Jander, 2. angew. Chem., 1929, 42, 1037 ;R., 1930, 51.G. Jander and E. Manegold, 2. anorg. Chem., 1924, 134, 283; A . , 1924,ii, 496.2o T. Callan and S. Horrobin, loc. cit.z1 W. D. Treadwell and C. Paoloni, loc. cit., T. Callan and S. Horrobin,loc. cit.; J. A. C. Toegan, Nature, 1930, 126, 504; A., 1930, 1375; W.Muchlinsky, Zoc. cit.21 F. L. Hahn, 2. E’bktrochem., 1930, 36, 989; A . , 1931, 189.z3 E. C. Righellato and C. W. Davies, loc. cit., ref. (4)296 ANALYTICAL CHEMISTRY.measured by means of a galvanometer and rectifier.24 The gal-vanometer readings are plotted directly on the conductivity-titration graph.The Dionic water tester has been used withsuccess in a number of tit ration^.^^ Special cells have beendesigned,26 and in view of the simplification of procedure, con-ductometric methods should find increasing applications in analyticalwork. The most important processes studied in recent years arethe titration of : acids and bases of various strength^,^' mixturesof acids,28 mixtures of sodium hydroxide and carbonate,29 salts ofweak acids and ba~es,~O carbonic phosphoric thesalts of heavy meta1sY33 phenols,34 alkal0ids,3~ anions formingsparingly soluble mercuric, silver, lead, or barium cationsyielding slightly soluble sulphates, chromates, oxalates, ferro-cyanides, ferric yanides, or ni troprussides ,37 moly bdat e, t ungst at e,and thallium,38 ferrocyanide by zinc,39 fatty acids,*O zinc chloride,and cyan ate^.^^ s.G.24 W. D. Treadwell and C. Paoloni, loc. cit.; T. Callan and S. Horrobin,2 5 N. Rae, J., 1931, 3143; A., 1932, 135.2o K. T. S. Britton and W. L. German, ibid., 1930, 1249; A., 1930, 860;J. M. Preston, ibid., 1931, 1827; A., 1931, 1026.27 I. M. Kolthoff, 2. anorg. Chem., 1920, 111, 1; A., 1920, ii, 420; E. C.Righellato and C. W. Davies, loc. c i t . ; idem, Trans. Faraday Xoc., 1933, 29,437; see also P. Hirsch, 2. anal. Chem., 1926, 68, 160; A., 1926, 700.28 I. M. Kolthoff, 2. anorg. Chem., 1920, 111, 28; A., 1920, ii, 421; E. C.Righellato and C.W. Davies, locc. cit.29 I. M. Kolthoff, 2. anorg. C?mm., 1920, 112, 155; A,, 1920, ii, 705; M.Aum6ras and J. Marcon, Bull. SOC. chim., 1932, [iv], 51, 1594; A., 363.30 I. M. Kolthoff, 2. anorg. Chem., 1920, 111, 97; A., 1920, ii, 501.31 Idem, ibid., 1920, 112, 165; A., 1920, ii, 705.32 Idem, ibid., p. 165; A., 1920, ii, 705; (Miss) J. C. Lanzing and L. J.van der Wolk, Rec. trav. chim., 1928, 48, S3; A., 1929, 284.33 I. M. Kolthoff, 2. anorg. Chem., 1920, 112, 172; A., 1920, ii, 709.34 Idem, ibid., p. 187; A., 1920, ii, 711; 0. Pfundt and C. Junge, Ber.,35 I. M. Kolthoff, 2. anorg. Chem., 1920, 112, 196; A., 1920, ii, 781.36 Idem, 2. anal. Chem., 1922, 61, 229, 336, 369, 433; A,, 1922, ii, 681,656, 781, 864; G. Jander, 0. Pfundt, and H.Schorstein, 2. angew. Chem.,1930, 43, 507; A., 1930, 1142; I. M. Kolthoff and T. Kameda, Ind. Eng.Chem. (Anal.), 1931, 3, 129; A., 1931, 699; 0. Pfundt, Angew. Chem., 1933,48, 200; A,, 582.37 I. M. Kolthoff, 2. anal. Chem., 1923, 62, 1, 97, 161, 209, 214, 216; A.,1923, ii, 88, 88, 256, 260, 256, 257.38 E. Rother and G. Jander, 2. angew. Chem., 1930, 43, 930; A., 1930,1548.39 G. Jander, 0. Pfundt, and H. Schorstein, loc. cit.40 G. Jander and K. F. Weitendorf, Angew. Chm., 1932, 45, 705; A., 49.4 1 0. Pfundt, ibid., 1933, 46, 218; A., 582.loc. cit. ; T. T. Potts, locc. cit.1929, 62, 616; A., 1929, 586AMBLER. 297Gas Analysis.Micro-Gas A.%alysis.--This term is best restricted to processesin which a very small amount of gas is analysed, and should not beapplied to those where a relatively large volume of gas is tested forminute traces of a constituent.Processes for the analysis of volumesof the order of I m1.l differ little in method from those for largerquantities; for samples of 0.1 ml. or less, however, a specialtechnique is required. I n some earlier methods2 for quantities ofthis magnitude, the gas was measured at low pressures by meansof a McLeod gauge; such processes are, however, being supplantedby the use of instruments of extremely simple Po~M,~, 4, 59 6, 7 inwhich 0.02--0.1 ml. of gas is measured in a vertical or horizontal 7graduated capillary tube of approximately 0.5 mm. bore, andtransferred for absorption, through a syphon tube, to small invertedtest-tubes standing in a mercury trough; the gas does not comein contact with a tap or other possible source of leak during theanalysis.The gas is measured in the dry state, instead of its beingsaturated with water vapour, and in most of the proce~ses,~, 59 6, 7solid absorbents are used, introduced into the absorption vesselsin the form of beads fused on the end of platinum wire. Oxygenis absorbed with solid phosphorus, washed with water and alcohol,and dried; 7 water vapour with fused phosphoric oxide; andammonia with potassium hydrogen sulphate. Carbon monoxideis absorbed with silver oxide (precipitated, pressed moist, and thendried; no carbon dioxide is pr~duced).~, Where no suitablesolid absorbent is available, beads of sintered glass or of kaolinand ground pottery,? impregnated with liquid reagent, are foundsatisfactory; sulphuric acid, for example, is used in this way forthe absorption of ethylene.6 Combustion is effected by explosioninitiated by a spark between specially designed electrodes, or, inthe case of hydrogen, by heating the outside of the tube with aflame.5 The paramount factor is the design and construction of1 E.g., D.S. Chamborlin and D. M. Newitt, Ind. Eng. Chem., 1925, 17,621; H. R. Ambler, Analyst, 1929, 54, 517.2 I. Langmuir, J . Amer. Chem. SOC., 1912, 34, 1310 ; P. A. Guye and F. E.E. Germann, J. Chim. phpique, 1916, 14, 194; H. M. Ryder, J . Amer. Chem.SOC., 1918, 40, 1656; C. H. Prescott, ibid., 1928, 50, 3237.3 L. Reeve, J . , 1924, 125, 1946.4 J.A. Christiansen, J . Amer. Chem. SOC., 1925, 47, 109; J. A. Christiansen6 F. E. Blacet and P. A. Leighton, I n d . Eng. Chem. (Anal.), 1931, 3, 266.6 F. E. Blacet, G. D. MacDonald, and P. A. Leighton, ibid., 1933, 5, 272.7 J. S. Swearingen, 0. Gerbes, and E. W. Ellis, ibid., 1933, 5, 369.8 A. Gautier, Compt. rend., 1898, 126, 171.and J. R. Huffmann, 2. anal. Chem., 1930, 80, 435.K 298 ANALYTICAL CHEMISTXY.apparatus so as to avoid gas-locks; details of such technique aregiven in the papers citedqSWithout personal experience, it is difficult to assess the accuracyof these methods as compared with that of macro-methods, but fromsome of the results given,6 the attainable accuracy seems to be ofthe same order (O-lyo). Manipulation, moreover, appears to beneither difficult nor slow ; indeed, the latest investigators find thatan analysis takes less time than with the Orsat or Burrell (elaboratedprecision Orsat) apparatus.There is, therefore, a case for the useof micro-technique even where plenty of sample is available, butexperience will be needed to decide on this point. I n the Reporter'sopinion, the elimination of liquid reagents is in itself likely toincrease accuracy, and it is undoubtedly easier, other things beingequal, to work with a small instrument and a small quantity ofgas. There are, of course, some processes, e.g., fractional com-bustion, which have not yet been made amenable to micro-technique.Fractional Combustion Methods.-J. G. King and L. J. Edg-combe lo have made a thorough study of the copper oxide method,and have shown that hydrogen and carbon monoxide are com-pletely oxidised by copper oxide at 280" and methane unaffected;the carbon monoxide can be accurately determined by measuringthe volume of carbon dioxide produced.This work seems tore-establish the process as a precise method after some ratherdiscouraging results by other workers.11 Homologues of methaneare slightly attacked a t 280°, and, if these are present in considerableamount, carbon monoxide should be removed by absorption, andthe hydrogen oxidised at 230-250". I n a method used by theReporter l2 for the determination of small proportions of methane(0.003 yo and upwards) in combustible'gases, hydrogen and carbonlaonoxide are oxidised by copper oxide; any residual traces ofthese are detected and removed by a check fractional combustionwith oxygen in the presence of platinum wire a t very dull red heat.The methane is then burned by raising the temperature of the wireto bright yellow heat.I<.A. Kobe and E. J. Arveson13 have investigated the use ofplatinised silica-gel for the catalytic fractional combustion of hydrogenand carbon monoxide; at 300" these are found to be fully oxidised,and methane unaffected. Hydrogen alone is oxidised at loo", butat this temperature the catalyst is poisoned by carbon monoxide.9 Particularly ref. ( 5 ) .10 34'uel Research Technical Paper No. 33, 1931.l 1 E. Ott and E. Scherb, 2. anal. Clwm., 1926, 68, 233; E. Scherb, &~s-urd Wasserfach, 1924, 67, 391.12 H.R. Ambler, Analyst, 1031, 56, 635.13 Ind. Eng. Chem. (Anal.) 1933, 5, 110AMBLER. 299If this process is found to be reliable, and the catalyst to retain itsactivity without frequent regeneration,14 the method is likely to beof great value for precise analysis, especially if higher paraffins arefound to be unaffected.Catalytic oxidation of carbon monoxide a t ordinary temperatures,by means of metallic oxide preparations (" Hopcalite ") l5 has beenapplied to the design of apparatus for continuous indication oftraces of carbon monoxide in air.16 The heat produced by theoxidation of the carbon monoxide is indicated by means of a, thermo-couple and galvanometer graduated in terms of carbon monoxideconcentration ; 0.01 yo can readily be detected. Further processesfor the determination of carbon monoxide by the iodine pentoxidemethod have been published.17, 18Analytical Absorbents.-The use of suspensions of iodine pentoxidein oleum for the analytical absorption of carbon monoxide l9 hasreceived further study,20,21 and it seems established that it isthoroughly reliable, provided that hydrocarbons other than methanebe absent.Suspensions of iodine in oleum also absorb carbonmonoxide but are less satisfactory.21Hydrobromic acid in glacial acetic acid has been recommendedfor the absorption of higher olefins in the presence of ethylene; 22and for hydrogen, a platinum preparation contained in a specialpipette, in contact with sodium chlorate solution.23 On the im-portant and debated question of the extent of the evolution ofcarbon monoxide when oxygen is absorbed by alkaline pyrogall01,~~l4 Compare E.Ott, Helv. Chim. Acta, 1924, 7, 886; Z . anal. Chem., 1926,68, 240; E. Scherb, Gas- und Wasserfach, 1924, 67, 391.l5 A. B. Lamb, W. C. Bray, and J. C. W. Frazer, J . Ind. Eng. Clzem., 1920,12, 213; A. B. Lamb, C. C. Scalione, and G. Edgar, J . Arner. Chem. SOC.,1922, 44, 738.l8 J . Sci. Instr., 1932, 9, 327; Electrician, 1933, 111, 309.l7 P. Borinski and H. Murschhauser, Chem. Fabrik, 1932, 5, 541 ; A. Bech,Chem. Abstr., 1933, 27, 5220.0. Pfundt, Chem. Fabrik, 1933, 6, 69; B., 265; G. Ljunggren andG. Frang, Svenslc Kern. Tidskr., 1932,44,279.P. Schlapfer and E. Hofmann, Monats Bull., 1927, 7, 293, 349; Chem.Zentr., 1929, i, 3013; B., 1930, 5.20 P.Schliipfer and H. Ruff, Monats Bull., 1929, 9, 5 ; E. Ott, Gas- undWasserfach, 1929, 72, 862; 1930, 73, 801 ; E. Dittrich, 2. angew. Chern.,1930, 43, 979; P. Schlapfer, ibid., 1931, 44, 170; H. A. J. Pieters, 2. anal.Chem., 1931, 85, 50.21 P. Schlapfer and C. Mosca, Monats Bull., 1932, 12, 205, 253, 286; ChemZentr., 1933, i, 268.22 V. Sorokin, A. Belikova, and 0. Bogdanova, J . Rubber Inst., RUGS., 1931,5, 26; Chem. Abstr., 1933, 27, 5680.23 E. Biesalski and H. Giehmann, Angew. Chern., 1932, 50, 767.24 See T. J. Drakeley and H. Nicol, J . SOC. Chern. Ind., 1925,44,457 ; 1929,48, 62300 ANALYTICAL CHEMISTRY aJ. S. Haldane and R. H. Makgill 25 have shown that with analysesof air in the Haldane apparatus, no measurable carbon monoxideis produced, provided the reagent (composition specified) has stoodfor 70 hours after being prepared, or alternatively, is heated for anhour at 100".For the direct absorption of nitrogen, heated lithiumis found satisfactory.26Miscellaneous Processes.-Processes have been described forthe determination of carbon dioxide in air,27 sulphur dioxide in air,28hydrogen cyanide in mixtures of carbon dioxide and air,29 smallquantities of oxygen 3O and iron carbonyL31 The palladium testfor carbon monoxide has been studied and its sensitiveness in-creased,a2 and the blood test, as modified by the use of the reversionspe~troscope,~3 has been used for determining very small pro-portions of carbon monoxide.34 The determination of ozone withpotassium iodide has been further examined by R.R ~ y s s e n , ~ ~and conditions determined where the titration is accurate withoutthe use of buffer solutions. Sources of error in the determinationof hydrogen have been examined by the Reporter.36 In the ex-plosion method, if the hydrogen content of the mixture explodeddoes not exceed 20%, no appreciable oxidation of nitrogen occurs.A little hydrogen may always remain unburnt; if the hydrogencontent is not less than lo%, this does not exceed 06y0 of the totalhydrogen. Below this limit, however, the proportion of unburnthydrogen may be high. In catalytic combustion with heatedplatinum wire, oxidation of nitrogen is insufficient to produceerror. The platinum wire method has been used also for thedetermination of small proportions of oxygen,37 by combustion with25 Analyst, 1933, 58, 378; B., 737.26 M.Trautz and K. F. Kipphan, 2. anal. Chem., 1929,78,350; H. Copaux,Bull. SOC. chirn., 1932, 51, 989; J. H. Severyns, E. R. Wilkinson, and W. C.Schumb, Ind. Eng. Chem. (Anal.), 1932, 4, 357.2 7 W. M. K. Martin and J. R. Green, ibid., 1933, 5, 114; A., 478; M. D.Thomas, ibid., 1933, 5, 193; L. W. Winkler, 2. anal. Chem., 1933, 92, 23;B., 366.28 M. D. Thomas and R. J. Cross, Ind. Eng. Chem., 1928, 20, 645; S. W.Griffin and W. W. Skinner, ibid., 1932, 24, 862.29 H. L. Cupples, I n d . Eng. Chern. (Anal.), 1933, 5, 50; A., 243.ao M. Mugdan and J. Sixt, Angew. Chem., 1933, 46, 90.81 E. Pohland and W.Harlos, 2. anal. Chem., 1932, 90, 193; A., 43.32 W. Ackermann, Chem-Ztg., 1933, 57, 154; Nature, 1933, 131, 441.33 H. Hartridge, Proc. Roy. SOC., 1923, [A], 102, 575; R. C. Frederick,34 See H. Hartridge, Nature, 1933, 131, 654.35 Natuurwetensch. Tijds., 1933, 15, 125; A., 921 ; see also A. Juliard andAnalyst, 1931, 56, 561.S. Silberschatz, Bull. SOC. chirn. Belg., 1928, 37, 205; A , , 1928, 978.H. R. Ambler, Analyst, 1930, 55, 436, 677.Ibid., p. 677AMBLER. 301hydrogen. If only hydrogen and inert gases are present withoxygen, the process is accurate to O.Olyo on the total gas. In thepresence of oxides of carbon and methane, the process is accurateA new method for the determination of nitrous oxide has beenadumbrated by L. M e ~ e r , ~ ~ based on the observation that whenthis gas is absorbed by active charcoal a t temperatures above O",oxygen is permanently retained, only the nitrogen being removable.If such a process proves satisfactory, it will be of great value, con-sidering the complete dearth of direct methods of determination ;no details are, however, given.L.C. McNair and H. C. Gulls9 determine combustible vapoursin air by burning them in a special apparatus by means of platinumwire, and measuring the volume change by a sensitive devicecompensated against changes in temperature and barometricpressure. The Reporter has devised a combustion vessel for thecontrolled slow combustion of This is used in a methodfor the direct determination of nitrogen in combustible gases, thesebeing burned in a stream of electrolytic oxygen, the excess of whichis afterwards absorbed, together with the resulting carbon dioxide,by alkaline pyrogallol.Other instruments which have been described are a compensator 41for constant-volume gas burettes and a portable apparatus 42 forgeneral gas analysis in which an accuracy of 0.1% is attained withsmall weight and bulk by the application of a special principle ofgas-measurement.In this method both volumes and pressures aremeasured, the measured pressure being used to correct the volumesto standard (atmospheric) pressure. For analyses of 0.5 % accuracy,. the pressure measurement may be omitted. Gases can also bemeasured a t constant volume ; the volumes dealt with are about10 ml. J.J. Fox and L. G. Groves43 have designed a recordinginstrument for small traces of sulphur dioxide in flue gases; someautomatic instruments for the determination of carbon monoxidehave been mentioned.16, l8Substances whose determination had received little attentionhitherto are the vapours of lead tetraefhylju amyl alcohol andto 04yo.38 Naturwiss., 1932, 20, 791.39 Analyst, 1932, 57, 159.40 H. R. Ambler, J . Sci. Instr., 1931, 8, 18; Analyst, 1931, 56, 804.dl Ibid., p. 374.42 Ibid., p. 369.43 J. SOC. Chem. Ind., 1932, 51, 7 ; B., 1932, 323.44 Ministry of Health, Final Report of the Departmental Committee on EthylPetrol, 1930302 ANALYTICAL CHEMISTRY.acetate 45 and ethylene oxide,46 the oxides of chlorine:’ and theproducts of the thermal decomposition of phosgene.48Analysis of Hydrocarbon Mixtures.-Mixtures of homologoushydrocarbons, particularly paraffins, have in the past presented avery difficult analytical problem.Chemical separation is out ofthe question, and the usual practice has been to burn the mixedgases with oxygen, and from the data so obtained (i.e., contraction,carbon dioxide produced, and oxygen consumed) to express themixture as C%H2n+2, with an empirical value for n. A recentprocess which gives some information as to the identity and pro-portions of the different constituents which make up this averagen is to allow some of the gas to diffuse out through a standard orificeand determine n before and after this process.49 J. D ~ b o i s , ~ ~ usingthe Union portable gas ~alorimeter,~l determines the calorific valueof gases to obtain information as to the nature of the “heavyhydrocarbons ” which are usually determined together in ananalysis.The only exact method, however, for mixtures of unknownqualitative composition, or those containing more than two homo-logues, is separation by a physical process, such as fractionation ofthe liquefied gas.A process for the fractionation of complexmixtures was described in 1915 by G. A. Burrell, F. M. Seibert, andI. W. Robertson 52 and has since been improved by W. E. Stockingsand G. W. Himus; 53 the introduction, however, of vacuum-jacketed fractionating columns % has made the separation almostas simple as the separation of liquids in an organic laboratory.Thevarious constituents distil over almost completely separated, andare identified by their boiling points and measured by collection ofeach constant-boiling fraction in a previously evacuated vessel ofknown volume, connected to a manometer. For binary andternary mixtures of known qualitative composition, a “ short-cut ” method has been devised for further speed and simplification ; 5545 I. M. Korenman, J. Appl. Chent. RUSS., 1931, 4, 940; A., 1932, 632.4 6 W. Deckert, Angew. Chem., 1932, 45, 559; B., 1932, 1010.4 7 J. W. T. Spinks, J. Amer. Chem. SOC., 1931,53, 3015.48 A. Stock and W. Wustrow, 2. anorg. Chem., 1931,195,129; A., 1931,326.49 R. A. Baxter and L. J. Beckham, J . Arner. Chem. Soc., 1933, 55, 3926.50 Przemyst Chem., 1931, 15, 390; B., 1932, 247,61 See A. Blackie, J. Sci. Instr., 1930, ‘7, 84.52 U.S. Bur. Mines, Tech. Paper No. 104, 1915.53 “ Fuel Testing,” 1932, p. 195.54 F. E. Frey and W. P. Yant, Ind. Eng. Chem., 1927, 19, 21, 492; H. S ,Davis, Ind. Eng. Chem. (Anal.), 1929, 1, 61 ; H. J. Lucas and R. T. Dillon,J. Amer. Chem. SOC., 1928, 50, 1460; W. J. Podbielniak, Ind. Eng. Chew.(Anal.), 1931, 3, 177; 1933, 5, 119, 172; W. E. MacGillivray, J . , 1932, 941,5 6 R. Rosen and A. E. Robertson, Ind. Eng. Chem. (Anal.), 1931, 3, 284AMBLER. 303the temperatures are noted a t which certain definite percentages ofthe total gas have been evaporated off under standardised conditions,and from these data the percentages of the constituents are obtainedfrom empirically constructed graphs. For simpler mixtures ofhydrocarbons 56 and for fractionation processes to be used in con-junction with chemical analyses,57 simpler processes have beendescribed.A further method for complex hydrocarbon mixtures, which seemsto give accurate separation, is selective absorption by cooled char-coal.58 The constituent gases are desorbed successively from thecharcoal and can be collected separately; the changes in thecomposition of the emergent gases are observed by means of a gas-interfer~rneter.~~ G. Kuhn 6o has described a somewhat similarprocess, wing silica-gel.Miscellaneous Physical Methods of Analysis.-The year 1933has seen the appearance of an authoritative text book61 on thethermal conductivity method. This method is of great and in-creasing applicafion, particularly for control work in technicalprocesses, since it is usually automatic and recording, and is ideallysuited to “ flow ” analyses. The “ viscosity-effusion bridge ” 62is an instrument with a wide field of possible application, workingon the principle of the relation between the density and viscosityof a gas. Small differences in this relation between the gas undertest and a comparison gas are indicated on an oil-manometer. Themethod is particularly sensitive for gases of high density and lowviscosity, such as organic vapours; for instance, 1% of hexanevapour in air produces a movement of 150 mm. on the manometer.The design of the instrument allows it t o be used with corrosivegases.M. Lambrey 63 finds that some bands in the absorption spectrumof nitric oxide are.greatly strengthened by the presence of anothergas, and can thus detect nitric oxide at a, partial pressure of 0.00256 E. 31. J. Mulders and F. E. C. Scheffer, Rec. trav. chim., 1930, 49, 1057;A., 1931, 54; C. A. L. Horstmann and F. E. C . Sclieffer, Rec. trav. chim.,1932, 51, 153; A., 1932, 247; 0. J. Walker and S. N. Shukla, J . , 1931, 368.57 H. S. Davis and J. P. Daugherty, I n d . Eng. Chem. (Anal.), 1932, 4, 193.58 P. Schuftan, “ Gasanalyse in der Technik,” 1931, p. 69; see also E.Berl and 0. Schmidl;, 2. angew. Chem., 1923, 36, 247.(Lord) Rayleigh, Yroc. Roy. SOC., 1896, 59, 201.60 2. angew. Chem., 1931, 44, 757; B., 1932, 7.61 H. A. Daynes, “ Gas Analysis by Measurement of Thermal Conductivity,”6 3 I. Fagelston, Brit. Pat., 330799 (1’321)) ; B., 1‘330, 799 ; Iwd. Chem., 1932,63 Compt. rend., 1931, 139, 857.1933.8, 57304 ANALYTICAL CHEMISTRY.mm. in one atmosphere of hydrogen. E. J. B. Willey and S. G.F ~ o r d , ~ * using the potassium photoelectric cell, which is mostsensitive in that region of the spectrum where absorption by nitro-gen peroxide is strongest, are able to determine 0.001% of thisgas. R. Ruyssen 65 uses the potassium photoelectric cell for thedetermination of ozone. Researches by A. Gatterer 66 indicate thefeasibility of quantitative determination of some gases by emissionspectra.H. R. A.H. R. AMBLER.B. A. ELLIS.J. J. Fox.S. GLASSTONE.64 Proc. Roy. Soc., 1932, [A], 135, 166.6 5 Natuurwetensch. Tijds., 1933, 15, 6 ; Bull. Acad. roy. Belg., 1932, [v],6 6 Physikal. Z . , 1932, 33, 64.18, 1085; A., 362

ISSN:0365-6217    

DOI:10.1039/AR9333000268

出版商:RSC    

年代:1933

数据来源: RSC

摘要:

BIOCHEMISTRY.A SECTION has been devoted this year to certain aspects of thebiochemistry of bacteria. The extreme diversity of the chemicalactivity of these organisms renders impossible any approach to acomprehensive review of recent investigations. The more practicalaspects of bacteriology in relation to industry, pathology, agri-culture, sanitation, etc., receive attention elsewhere. It hasseemed more appropriate to consider certain selected topics, whichby reason of their fundamental or theoretical character are suit-able for inclusion here. Moreover, since no reference to bacteriahas appeared for some time in these Reports, the period underconsideration has been extended to cover the last three or fouryears. A short section on algze has similarly been introduced.In the section dealing with animal biochemistry the subjectsreviewed this year are the highly important new fermentationschemes of Embden and Meyerhof, deaminisation by survivingtissues, metabolic aspects of methionine, vitamin A , vitamin C(ascorbic acid), the physiology of vitamin B,, carcinogenesis bypure hydrocarbons, synthetic estrogenic substances, the cestringroup, and the new group of the Jlavin pigments.The morechemical aspects of the subjects italicised are dealt with in thealiphatic division of the organic chemistry report.PLANT BIOCHEMISTRY.Biochemistry of Certain Bacteria.The Nitrogen Assimilation of Azotobacter.-The relationshipbetween the nature of the nitrogen assimilation process and thegrowth of these important organisms has presented a fascinatingproblem to both the bacteriologist and the agriculturalist.It haslong been recognised that in the presence of nitrates the fixationof free nitrogen by Axotobacter is retarded or entirely prevented.The more precise information of the detail of this effect which hasbecome available during the last two or three years is to a largeextent due to the work of D. Burk and H. Lineweaver. Theirearlier investigations showed that the addition of small but increas-J . Bact., 1930,19, 389; A., 1930, 1219306 BIOCHEMISTRY.ing amounts of combined nitrogen to otherwise nitrogen-free mediarapidly reduced the amount of elementary nitrogen fixed. Fixationceased altogether when the medium contained 0.6 mg. of availableN per C.C.The growth and respiration rates of the organism alsoincreased rapidly with t,he supply of combined-nitrogen, reaching amaximum with 0.5-1.0 mg. of N per C.C. and declining slowly asthis value was exceeded. As far as could be ascertained, the principalphysiological functions of the bacteria were unaltered whether theassimilated nitrogen was free or combined. The nitrogen contentof the dry matter of the cells was likewise unaffected.It would appear that for normal metabolic processes Axotobacterutilise combined nitrogen if available and will resort to fixationonly when the alternative source fails. This view is confirmed,and more light thrown on the relative assimilability of variousnitrogenous compounds by J. E. Fuller and L.F. Rettger.2 Theeffect of a range of nitrogenous compounds varying in complexityfrom simple nitrates and ammonium salts to simple amides andamino-acids (urea, glycine, aspartic and glutamic acids) and morecomplex compounds such as tyrosine, nucleic acid, etc., on nitrogenfixation by, and the reproduction of, Axotobacter was comparedwith those in cultures from which all combined nitrogen was ex-cluded. I n general, the simpler compounds were the more easilyassimilated, produced the greater growth increases, and, at thesame time, caused the heavier reduction in the amount of freenitrogen fixed. Similar effects were observed by L. G. Th~mpson.~The clear differentiation between the course of N fixation and thegrowth process in Axotobacter is further emphasised in the nutri-tional requirements.Thus in media containing the customaryminerals (K, Mg, PO4”’, SOa”, etc.) and combined nitrogen littlecalcium seems necessary for optimum growth and normal meta-bolism.* During the fixation process calcium is necessary inconsiderable amount and becomes an important factor controllingthe amount of nitrogen fixed.5 The interesting point is broughtout by this investigation that strontium, but no other element,can replace calcium in this f~nction.~,The first product of nitrogen fixation so far identified is ammonia,and this appears only when an adequate energy (carbohydrate)source is available. If the organism is supplied with combinedforms of nitrogen, these are decomposed with the production ofSoil Sci., 1931, 31, 219; B., 1931, 557.J.Agric. Ites., 1932, 45, 149; A., 1932, 1169.I). Burk and H. Lineweaver, Arch. Mikrobiol., 1931, 2, 155; A , , 1931,D. Burk, Proc. 2nd Internat. C’ong. Soil Sci., 1932, 3, 67; A., 96.4 M. Schrtider, Zentr. Bakt. Par., 1932, 11, 85, 177; A., 1932, 306.1334POLLARD AND PRYDE. 307ammonia, which is then utilised. The deamination processapparently occurs only in the absence of carbohydrates (S. P.Kostytschev and A. Schelo~mov).~The examination of the effects of varying partial and totalpressures of oxygen and nitrogen on the fixation process, by Burk,8discussed in an earlier Report,g initiated a series of investigationsof the mechanism of the stimulatory action of humus, and later,of other substances on the activities of Axotobacter. In experimentson the respiratory exchange, K.IwasakalO showed that the re-productive process is associated with a steady increase in therespiratory rate, whereas the fixation process takes place withconstant respiratory conditions. A low oxygen pressure favoursfixation and inhibits growth, and respiratory activity. Humicmatter increases the respiration rate. The fixation of nitrogen isalso accelerated by molybdenum salts (Na,MoO,),ll, l2 but themechanism of the stimulation differs from that produced by soilextracts. Both materials increase the proportion of nitrogenfixed per unit of sugar consumed, but neither affects the ratio,nitrate assimilated : sugar consumed. Soil extract acceleratesthe growth rate of the organism, and in this respect, is effectivewhether free or combined nitrogen is being assimilated.Theaction of the molybdate, however, is confined to increasing theefficiency of the fixation process. That the growth-acceleratingeffect is a function of some organic constituent of soils is confirmedby the observation that the ash of soil extracts acts in a preciselysimilar manner to, and with approximately the same intensity as,the sodium molybdate. The activity of the latter apparentlydiffers from that of uranium (and in some cases of thorium), which,as previously reported by K. Hirai,13 stimulates both nitrogenfixation and growth. The action of humus is stated by J. Voicuand E. Lungulescu l4 to depend on its influence over the utilis-ation of energy materials by Axotobacter as well as on the actualfixation process.Thus, small amounts of humus accelerate theoxidation of sucrose but retard that of glucose. Larger proportionsin general retard oxidation to extents which vary with the natureof the carbohydrate present.7 2. physiol. Chem., 1931,198,105; A., 1931,986; Bull. Rcad.Sci. U.R.S.S.,8 J . Physical Chem., 1930, 34, 1174; A., 1930, 1068.S Ann. Reports, 1930, 242.10 Biochem. Z., 1930, 226, 32; A., 1930, 1622.11 H . Bortels, Arch. Mikrobiol., 1930, 1, 333; A., 1931, 1095.12 L. Birch-Hirschfeld, ibid., 1932, 3, 341 ; A., 1932, 1169.13 Proc. 1st Internat. Cong. Soil Sci., 1928, 3, 154..1 4 Bul. ChinL. SOC. Rodney 1930, 12, 71, 82; A . , 1930, 1478.1931, 661 ; A., 1931, 1459308 BIOCHEMISTRY.In a further investigation, Burk, Lineweaver, and G.K. Horner15associate the stimulatory action of humus with its iron constituents,since artificial iron-free humus is without action. Moreover theiron salts of certain organic acids (citrate, tartrate, malate) and, to aless extent, ferric sulphate may produce similar effects, the order ofefficiency being humus > organic salts of iron > ferric sulphate.Under cultural conditions such that natural humus fails to stimulate,the iron salts are also ineffective. It is indicated that the mainten-ance of an appropriate proportion of soluble iron in the medium is animportant factor in the production of stimulative effects.No confirmation was obtained by these workers of theories, fre-quently expressed, that the effect of humus results from its supposedaction in increasing the availability of nutrient materials, in detoxi-cating metabolic producls, or in improving the physical conditionof the substrate.Burk’s “ iron ” theory falls into line with observ-ations of many early workers and with that of J. Olsen,16 whoascribed the activity of Bottomley’s “ Bacterised peat ” to thepresence of iron compounds rather than to that of auximones.Bacterial Decomposition of Xugars and their Derivatives.-Thefermentation of various sugars has long formed the basis of testsused in characterising bacterial species. The consideration offermentability by bacteria as a property of sugars which might becorrelated with molecular structure is less commonly discussed.Bacteria as a class exhibit a specificity in this respect which is welldefined with regard t o individual sugars, but less easy to elucidatein reference to groups of sugars having common structural charac-teristics.An interesting variation of this theme is provided by thework of A. I. Kendall and C . E. Gross l7 in which the relative ease offermentability of sugars is compared with that of their correspondingsimple derivatives. Thus, while glucose is in general more readilyfermented than mannose, the reduction products sorbitol andmannitol show the reverse order of fermentability. Oxidation ofglucose to gluconic acid does not greatly restrict its utilisation bybacteria, whereas mannonic acid is attacked by relatively few organ-isms.The simple glucosides were rarely utilised, @-methylglucosidebeing more generally attacked than a-methylglucoside. a-Methyl-mannoside was very resistant. Alkylated sugars such as themethyl and isopropylidene derivatives of glucose, fructose andmannose, tetramethyl mannose, and y-trimethyl xylose were rarely,if at all, acted on by bacteria. Investigations of a somewhatsimilar character are more recently recorded by S. A. Koser andl6 Cornpt. rend. Trav. Lab. Carlsberg, 1930, 18, 1 ; B., 1930, 434.l7 J. Infect. Dis., 1930, 47, 249; A., 1931, 264.1 5 soil s~i., 1932,33,413, 455; B., 1932, 782POLLARD AND PRYDE. 309F. Saunders.l* A series of simple derivatives, selected as retainingthe essential spatial arrangement of atoms, were subjected to theaction of a series of bacterial species, and, incidentally, of two yeasts.Here again the refractory character of the glucosides is shown,d- a-met hylmannoside, Z- p- met hylarabinoside, and d - p - met hyl-xyloside being unattacked by any of the organisms which fermentedthe corresponding simple sugars, and a-methylglucoside beingfermented only in a few instances.On the other hand the intro-duction of the methyl group into the direct carbon chain as inrhamnose and fucose did not markedly reduce fermentability. Ofthe heptosa, a-glucoheptose and a-glucoheptulose were not utilisedby any of the bacteria examined. The hexose, sorbose, having asimilar spatial arrangement to a-glucoheptulose, was fermented onlyby a few species.Glucosamine and gluconic acid were attacked bythe majority of glucose-fermenting organisms, though not alwayswith equal readiness. It is significant that neither compound wasacted on by the yeasts. The sulphur derivative, glucose ethylmercaptal, and the acetyl derivatives of glucose and xylose provedunfermentable. The relative ease of fermentation of glucosamine isalso recorded by F. Lieben and L. LOwe,lg and the resistant characterof several glucosides towards certain organisms which attack thesimple sugars by S. Forssman.20 I n another paper, S. A. Koser andF. Saunders 21 show that a number of organisms ferment Z-arabinosemore readily than the d-form, although in a few cases the d- but notthe Z-form was utilisable.It is generally indicated that the spatialarrangement of atoms or the length of the carbon chain is by nomeans always a controlling factor in the fermentability of sugars bybacteria, even when the activities of a single species only are con-sidered.Production of Acetic Acid by Bacteria.-The very varied and oftencomplex transformations brought about by bacteria between carbo-hydrates, alcohols, acids and the intermediate products, togetherwith their control by the adaptation of cultural conditions to favouror inhibit the activity of individual bacterial enzymes, form thesubject of an ever increasing number of investigations. I n this, asin all other forms of the biochemical changes in bacteria, the im-portance of small variations in the primary or secondary nutrients,of the chemical form in which these are supplied, of the temper-ature, hydrogen-ion concentration and oxidation-reduction potentialof the media, of respiratory relationships and of the stage of develop-18 J .Bact., 1933, 25, 32; A,, 639; also ibid., p. 475.19 Biochem. Z . , 1932, 252, 70; A., 1932, 1066.20 Ibid., 1933, 264, 231; A., 1083.21 Proc. SOC. Exp. Biol. Msd., 1932, SO, 218; A., 1932, 1206310 BIOCHEMISTRY.ment of the cultures is repeatedly emphasised. Considerableattention attaches to the mechanism of the production of aceticacid, more especially to the final stage, aldehyde + acid.The dismutation theory initiated by Neuberg and Windisch in1926 postulated the change of two molecules of aldehyde t o onemolecule of acid plus one molecule of alcohol.The latter is furtherattacked to produce more aldehyde, thus obscuring the true courseof the change, when examined analytically.Later H. Wieland and A. Bertho 22 presumed the direct dehydro-genation of aldehyde to be the dominant reaction, CH,*CHO [orCH,CH( OH),] -+ CH,*CO*OH + H,O. The dehydrogenation ofisopropyl alcohol completely to acetone is cited as a, parallel action.23Further confirmation of the dismutation theory was furnished byE. 25 who established the occurrence of a keto-aldehydemutase in a number of species of acetic acid bacteria. K. Tanaka 26considered the dismutation and dehydrogenation of acetaldehyde astwo phases of activity of the same enzyme. Dehydrogenation withoxygen as the hydrogen-acceptor is more rapid in acid than inneutral media, the reverse being the case with dismutation.Thatdisrnutation can take place under aerobic conditions has further beenestablished by Simon.,'Kojic Acid.-Certain species of acetic acid bacteria which producegluconic acid from glucose, also yield kojic acid from fructose ormannitol. Among these are varieties of B. hoshigaki and, lessactive in this respect, B. xylinoides. According to T. Takahashiand T. Asai 28 the acid is formed only from fructose and mannitol,not from compounds containing less than six carbon atoms. Themechanism is represented as a progressive oxidation.-I€YH,*OH - H ~ O FH,*OH -~H,o (iH,*OH +- !.To [yH*OH], + o-+ F0 t oCH,*OH [(p*OHI,7Jo IG O O H 1CH,*OHCHKojic acid22 Annalen, 1928, 467, 95; A., 1929, 219; also A.Bertho and K. P. Basu,ibid., 1931, 485, 26; A., 1931, 394.z3 A. Bertho, ibid., 1929, 474, 1; A., 1929, 1492.24 Biochern. Z . , 1930, 224, 253; A . , 1930, 1477.25 C. Neuberg and E. Simon, ibid., 1932, 253, 2 2 5 ; A , , 1932, 1169.26 Acta Phytochim., 1931, 5, 239; A., 1932, 94.27 Biochem. Z . , 1931, 243, 401; A . , 1932, 196.28 Proc. Imp. Acad. Tokyo, 1032, 8, 364; A . , 189; also Zenir. Bakt. Par.,1933,11, 88, 286; A., 1206POLLAR,D AND PRYDE. 31 1It is of interest that kojic acid production by Aspergillus takesplace from simple compounds by an entirely different process(see p. 318).Xulphur Bacteria.-Some brief reference seems appropriate tothis widely distributed class of organisms, which are characterisedby the utilisation of relatively large proportions of elementarysulphur or its simple compounds.Only a very small percentage of the assimilated sulphur isrequired for purely nutritional purposes ; the major part serves inmost cases as an agent rendering available to the organism energyfrom other sources. The importance of these organisms in naturalwaters and water supplies, in sanitation, and in their relation tosoil fertility is well known.More recently their significance ingeological process is being explored. The possibility is also indicatedthat certain classes of sulphur bacteria are associated with, if notdefinitely concerned in, some forms of stone decay (Paine, et aZ).29Apart from these more specific activities, perhaps the chiefpoint of recent fundamental interest is the elucidation of the complexrelationship between the assimilation of hydrogen sulphide andthe respiratory process in these organisms.This relationship issomewhat obscured at times by the apparent ability of the bacteriato adopt an alternative metabolic process-so-called “ pleoener-gism.” 30 Many instances are recorded in which the normal develop-ment of the bacteria occurs in the complete absence of sulphur;e.g., the purple sulphur bacteria grow under anaerobic conditionsand in the presence of organic matter without any source of oxidis-able sulphur and in this condition utilise radiant energy.31- 32The final product of oxidation of sulphur under normal conditionsis sulphate, but the process occurs in stages which include theformation of thiosulphate, thionic and sulphite.Inter-mediate-stage oxidations may be intensified by use of particularspecies or by variation of growth conditions. As is to be anticipated,the reaction of the medium is an important factor influencingthe nature of the changes, and is utilised by some workers as abasis of classification. The particular stage or type of oxidationwhich is dominant in media of different reaction serves to dis-tinguish a number of main groups of organisms.34 For more29 S. G. Paine, l?. V. Linggood, F. Schimmer, andT. C. Thrupp, Phil. Trans.,1933, By 222, 97 ; B., 1933, 628.30 “ Sulphur Bacteria,” D. Ellis (Longmans), p. 47.3 1 C. B. van Niel, Arch.Mikrobiol., 1931, 3, 1 ; A., 1932, 884.33 F. M. Muller, ibid., 1933, 4, 131; A., 1207.33 Penta- and tetra-thionic acids were isolat’ed from soil cultures by G.34 I. P. Langc-Pozdeeva, ATkh. B i d . Naulc, 1930,30, 189; A., 1930, 1622.Guittonneau and J. Keilling (Compt. land., 1932,195, 679; B., 1932, 1128)312 BIOCHEMISTRY.detailed classification of the increasing number of species isolated,morphological characters, pigmentation, and the presence or other-wise of sulphur accumulations in the cells are considered.According to van Nie135 the development of the purple andgreen sulphur bacteria is largely controlled by the concentrationof hydrogen sulphide in the medium and in turn by its reaction.The metabolism is represented as a true photosynthetic processconforming to the equation CO, + 2H,S = CH,O + H,O + 2s.The sulphur is deposited within or without the cells and in turnreduces more carbon dioxide ; the complete equation becomingX O , + H,S + 2H,O + 2CH,O + H,SO,.I n the absence of sulphur the purple organisms utilise salts oforganic acids (lactate, pyruvate, acetate, malate, succinate, etc.).The work of F.M. Muller (Zoc. cit.) indicates that the preliminarytransformation of other acids to pyruvic acid is an essential stepin the building up of cell constituents. I n this case also the processmay be regarded as il particular instance of the general photo-synthetic rule. In another paper 36 Muller shows the similaritybetween the carbon assimilation of the red sulphur bacteria exposedto light and that of the higher plants.I n the presence of hydrogen-donors the organisms reduce carbon dioxide to formic acid andthence to formaldehyde. The red pigment is closely allied tocarotene and xanthophyll and appears to play a part in the assimil-atory function of these bacteria, similar to that of chlorophyll inplants.Carotenoid pigments are absent from the green sulphur bacteria,and in this species reduction of carbon dioxide occurs only in thepresence of hydrogen sulphide as a hydrogen-donor. Carotenoidsare associated with the utilisation of less active hydrogen-donors(e.g., water) than hydrogen sulphide.I n a recent monograph D. Ellis (op. cit.) gives a fairly comprehen-sive review of the morphology and chemical activities of the sulphurbacteria.Pigmentation and Luminescence in Bacteria.-Alt hough theformation of colour in bacterial cultures frequently serves a usefulpurpose in the identification or selection of organisms, little in-formation is available as to the nature of the pigments, their form-ation or purpose. The literature of the current year indicates arising interest in this aspect of bacterial biochemistry, and a briefreference to recent investigations seems desirable here, The verydefinite pigments of the sulphur bacteria (see above) have beenexamined spectrophotometrically but of their chemical compositionnot very much is known.The red pigment of B. prodigiosus has35 LOG. cit. ae Chern. WeekbZad, 1933, 30, 202; A., 429POLLARD AND PRYDE.313been more extensively studied. The production of the pigmentis largely influenced by growth conditions, particularly the natureof the carbon and nitrogen sources, and the reaction of the medium.37W. Moycho 38 indicates that the pigment is produced during auto-lysis of the cells in the presence of oxygen, and undergoes a colourchange (red --+ yellow) at p , 8.0 approx. In a series of papers3'. Wrede 39 and colleagues describe the isolation and examinationof the pigment prodigiosin, to which is ascribed the structure :R'--R MeQ-V-c==\/R 4 (l) 11 I (2) IThe position of the methoxy- group in ring(2) is uncertain. R + R' + R" = C,H1,. NH i /--IAn apparent inverse relationship between the active growthof bacteria and the formation of pigment is shown in an examin-ation of Grassberger's bacillus by G .Sandor and G. Rougebiefa40I n suitable media having a neutral or alkaline reaction the organismgrows rapidly, but for some time is colourless. When acid metabolicproducts have accumulated to lower t'he p E of the medium suffi-ciently, pigmentation is apparent. On the other hand, when theorganism is grown in slightly acid media, the red colour is apparentthroughout, but the culture develops only very slowly.The isolation of a highly toxic pigment from bacteria developingon coconut products is described by A. G . van Veen and W. K.Merten~.~l The yellow substance (m. p. 200") is dialysable, ampho-teric, and has a high nitrogen content. The minimal lethal dosewhen administered intraperitoneally to rats is less than 0.005 mg.The dependence of pigmentation on cultural conditions is demon-strated in a number of cases.For instance, s. Arakawa42 in anexamination of Axotobacter shows that, whereas A . chroococcum andA . vinelandii produce pigment in media containing the simplesugars, only the former organism is coloured when the carbonsource supplied is inulin, dextrin, or starch. I n both cases theaddition of potassium nitrate as a nitrogen source intensifies pig-mentation. Similarly, in the case of organisms producing fluor-37 G. Gorbach, Zentr. Bakt. Par., 1929, 11, 79, 26; A , , 1930, 1219.38 Compt. rend., 1930, 191, 497; A., 1930, 1478.39 F. Wrede and 0. Hettche, Ber., 1929, 62, [B], 2678; A., 1929, 1469;F.Wrede, Z. physiol. Chem., 1932, 210, 125; A., 1932, 1043; F. Wrede andA. Rotlihaas, ibid., 1933,215,67; 1933,219,267; A., 516, 1172.40 Bull. SOC. Chim. biol., 1933, 15, 415; A., 640.4 1 Proc. K . Alcad. Wetensch. Amsterdam, 1933, 36, 666; A , , 1206.42 Tottori Agric. Coll. Sci. Papers (10th Anniv.), 2 4 ; A., 1932, 652314 BIOCHEMISTRY.escent pigments, F. K. Georgia and C. 3'. Poe43 have shown thatpigment production necessitates the presence in the media ofmagnesium, phosphates, and sulphates. Quite small proportionsof these suffice, but the organisms are sufficiently sensitive to beutilised to detect, by their fluorescence, the presence of very smallamounts of these essential ions. The same organisms are also verysensitive to the nature of the nitrogen supply, different samples ofpeptone producing unexpectedly large variations in pigmentationA similarly intimate relationship apparently exists betweennutrient conditions and light emission by several luminescentbacteria.The work of F. F ~ h r m a n n ~ ~ shows that among mineralnutrients necessary to P. radians, both anions and cations of alkalihalides contribute separate effects to the intensity of the lightproduced. Various sugars increase luminosity not only in actualintensity but in the rate of increase to maximum intensity and thesubsequent decline. These effects differ with the nature of the sugarused and are modified t o different extents by variations in thesodium chloride concentration of the medium.The organism candevelop anaerobically, but free oxygen is necessary for luminescence.Light production and cell multiplication are apparently unrelatedprocesses. A. Mudrak,46 in an examination of several species ofluminous bacteria indicates that, while light emission occurs only inthe presence of free oxygen, the organisms can, in some instances,utilise sodium nitrate and chlorate as sources of respiratory oxygeii.Further, these bacteria can tolerate sodium and potassium chlorates(the former up to 9%) or sodium thiosulphate in considerable pro-portions, but growth is checked by sodium bromate, iodate or per-chlorate and by potassium iodate. As sources of nitrogen thenitrates of potassium and ammonium, urea, and tyrosine wereunsatisfactory. Glycine, asparagine, and aspartic acid were utilis-able, although in the case of the last-named, only the amino-nitrogenwas attacked.An instance of apparently normal development inthe absence of dissolved nitrogen suggests the possibility of the fix-ation of atmospheric nitrogen. Sodium is an essential element forthe growth of all species.S. E. Hill4' describes the effects of ammonia and of fatty acids indestroying the luminescence of B. Jischerei. This action is largelydue to actual penetration of the cell wall, and subsequent cytolysis.At a given pH in the medium the loss of luminescence, brought about43 J . Bact., 1931,22, 349; A., 1932, 198.44 Ibid., 1932, 23, 135; A., 1932, 430.45 Monatsh., 1932, 60, 69, 414; A., 1932, 652, 1066.46 Zentr.Bakt. Par., 1933, 11, 88, 353 ; A., 1334.47 J . Cell. Comp. Physiol., 1932,1, 145; A., 1932, 1290POLLARI) AND PRYDE. 315by fatty acids, increased in the order of the series, valeric -+ formicacids. I n certain ranges of pH, the cytolysis occurring on exposure toO6M-solutions of ammonium salts of the fatty acids increasedwith the molecular weight of the acid.AZgcr?.The chemistry of algae receives relatively small but persistentattention from research workers, and since little reference has beenmade to the subject in recent Reports a brief statement cf certainpoints arising during the past two or three years is included here.Our knowledge of the biochemistry of algze is of a somewhat frag-mentary character. Probably most consideration has been givento the nature of algae pigments.The elucidation of metabolicprocess is to a considerable extent restricted to the examinationof the commoner carbonaceous and nitrogenous constituents of thetissues. It is characteristic of the majority of alga? that they con-tain negligible amounts of sugars, the more persistent stage of car-bon metabolism being represented by the sugar alcohols, mannitol,sorbitol, and d u l ~ i t o l . ~ ~ , 49 According to Haas and Hill (Zoc. cit.) theformation of sugars in certain marine algae is a much slower processthan their subsequent conversion into alcohols, although the presenceof certain sugars is definitely recorded, especially from samplesobtained during the summer month~.~O, 51Two sugar derivatives, algin and laminarin (or laminaroloside)are very generally distributed in algae, the latter substance appear-ing in maximum quantities in the summer period. The laminarincontent of Laminaria appears to be closely related to sea temper-ature and the amount of sunshine.52 Laminarin yields only glu-cose on hydrolysis and is probably a condensation product of theformula (C,H,oO,), or ,.Algin is variously described as a glucosidecontaining a glycuronic acid residue 53 and as a mixed calcium-mag-nesium-aluminium salt of an acid resembling pectic acid.54 Theoxidation product, alginic acid, yields mannuronic acid on hydro-lysis and monosaccharic acid on oxidation. It occurs only to alimited extent in algae.I n an examination of the fatty constituents of seaweeds Haas48 H.Colin and P. Ricard, Compt. rend., 1930,190, 1514; A., 1930, 1072.49 P. Haas and T. G. Hill, Biochem. J., 1931, 25, 1470; A., 1932, 101;50 H. Colin and E. Gueguen, Compt. rend., 1930, 190, 884; A., 1930, 825.5 1 Z-Fucose and d-mannose isolated from sea weeds; R. H. F. Manske,52 P. Ricard, Bull. SOC. Chim. biol., 1931, 13, 417; A., 1931, 883.63 H. Colin and P. Ricard, ibid., 1930,12, 1392; A., 1931, 535.Ann. Bot., 1933, 47, 55; A., 436.J . Biol. Chem., 1930, 86, 571 ; A . , 1930, 825.J. Giral, Anal. Pis. Quim., 1929, 2, 144; A., 1930, 259316 BIOCHEMISTItY.and Hill (Zoc. cit., 1933) show that, although all species examinedcontain very similar fats, the more exposed species have, in general,a higher proportion of fat and also their fats are of lower iodinevalue.The unsaponifiable fraction of the fats tends also to becomegreater with increasing depth of average immersion of the species.In the same communication is described the distribution of cer-tain nibrogenous compounds. Again the influence of immersion isapparent. In most cases a high total-nitrogen content is associatedwith a relatively low proportion of fat. Ammonia occurs in allsubmerged species and rarely in those exposed. In the latter,amide-nitrogen predominates over amino-nitrogen, whereas insubmerged plants the order is reversed. An octapeptide of glutamicacid is recorded in Phaophycece and also in Pelvetia canali~ulata,~~and methylamine and trimethylamine in a number of species byR.Kapeller-Adler and F. Vering.56 In no case was dimethylaminedetected. The presence of nitrates in a number of green alga isshown by S. S~neson.~' Many brown and some red species do notaccumulate nitrate.The pigments of certain red alga have been the subject of muchinvestigation in recent years, and much of our present knowledge isdue to the work of R. Lemberg, who, from the chromoproteinsphycoerythrin and phycocyan, obtained the corresponding truepigments, which he named phycoerythrobilin and phycocyanobilin.He subsequently established their close relationship with the bilepigments.ss The chromoproteins contain approximately 2% of thetrue pigments, which resemble in many properties magnesiurn-free chlorophyll, and occur in the alga in proportions which varysomewhat with environmental conditions.The proteins of thetwo compounds apparently are not identical.Growth and Metabolism of Moulds.Further investigations of the r61e of zinc in the growth of moulds 59have been reported this year. A direct relationship between thezinc content of a number of fungi and their nucleolytic power isindicated by M. Mousseron and P. Fauroux.60 Where this valueexceeds 100 mg. of zinc per kilo. of dry matter, the organisms are5 5 P. Haas and T. G. Hill, Biochem. J., 1931, 25, 1472; A,, 1932, 101.56 Biochem. Z., 1931,243,292; A., 1932,204.5 7 2. physiol. Chem., 1932, 204, 81; 1933, 214, 105; A., 1932, 314; 1933,437.68 Annalen, 1928, 461, 46; 1930, 477, 195; A., 1928, 533; 1930, 488.With G.Bader, Naturwiss., 1933, 21, 206; A., 651.59 See also Ann. Repmi%, 1932, 271.Go Bull. SOC. Chim. biol., 1932, 14, 1235; A., 106POLLARD AND PRYDE. 317hsmolytic. The repeatedly observed inhibitory action of zincsalts on sporulation in Aspergillus niger is accompanied by modifi-cations in the metabolic and structural chemistry of this mould.61The addition to a sugar medium of 0.01% zinc sulphate resultedin increased utilisation of sugar and greater production of citricacid. The hemicellulose contents and the proportions of ether-and cold water-soluble matter of the mycelium were increasedand $hat of lignin was lowered. I n its stimulative effect on vegeta-tive growth in fungi zinc is associated with the " growth substance-B," probably acting as a co-catalyst of growth.62The importance of calcium as a nutrient for fungi, as distinctfrom the use of calcium carbonate as a neutralising agent in media,is further confirmed by A.Rippel and U. S t ~ e s s . ~ ~ I n A. niger andcertain other cases, the calcium concentration of the mediumexerts a marked effect on the growth of the mould only whenabnormally large proportions of magnesium are present. It thenappears to act as a regulator of intake (ion antagonism) and ofthe physiological functions. Similar effects are produced bypartial substitution of strontium (as in the case of Axotobactercited above) or tannin for calcium.The action of fungi on arsenic compounds is apparently morecommon than is generally supposed. C. Thom and K.B. Raper 64have isolated a number of organisms from soil. These producearsenical gases when grown in arsenic-containing media, andseveral strains of Penicillia grow well on such media withoutproducing gas. I n the case of P. brevicaule, reported by F.Challenger,65 trimethylarsine is formed from arsenites and fromorganic arsenical compounds, and quinquevalent arsenic com-pounds are reduced to tervalent .Organic Acids and Other Products of Fungal Metabolism.-Theproduction of formic acid by a number of species of Aspergillus andPenicillium is examined by T. Chrzaszcz and M. Zakomorny.66The complete chain of transformations from sugarwould appear to be :sugar + acetic --+ fumaric + glyoxylic --+ formic acid. Withfavourable growth conditions, formic acid accumulates beforeundergoing further decomposition to carbon dioxide and hydrogen.Under less favourable conditions formic acid may be convertedinto oxalic acid.The conversion of sodium and calcium formates6 1 N . Porges, Bot. Baz., 1932, 94, 197; A., 188.62 N. Nielson and V. Hartelius, Cornpt. rend. Trav. Lab. Carlsberg, 1932,19, No. 8; A., 1932, 661; Biochem. Z., 1933, 259, 340; 281, 70; A., 638, 751.63 Arch. Mikrobiol., 1932, 3, 492; A., 97.64 Xcience, 1932, 76, 549 ; A., 189.6 5 I n d . Chem., 1933, 9, 134; A., 638.66 Biochem. Z., 1933, 259, 156; 263, 105; A,, 536, 982318 BIOCHEMISTRY.into the corresponding oxalates and carbon dioxide is also recordedby K. Bernhauer and F. Slani~~a.~' It is not quite clear from aconsideration of the above papers whether oxalic acid is a necessaryproduct of oxidation of formic acid or whether it appears as theresult of a side reaction.Various species of Aspergillus which produce kojic acid fromsugars, inulin, etc., are examined by K.SakaguchL68 Acid condi-tions and a suitable source of nitrogen favour the productionof this acid,68, which is also stimulated by the presence of smallamounts of iron, on a glucose substrate,70 and by ethylene chloro-h ~ d r i n . ~ l A. J. Kluyver and L. H. C. Perquin 72 indicate thatkojic acid is a direct product of transformation from glucose,but when it is obtained from fructose, galactose, xylose, arabinose,mannitol, erythritol, or glycerol, a C,-carbohydrat,e is probably anecessary intermediate product.On the other hand Sakaguchi(loc. cit.), who obtained kojic acid from 3C-compounds and inone instance from ethyl alcohol, favours the conception that glucoseis first broken down to 3-carbon compounds, which by oxidativecondensation form the pyrone ring. This is in conformity withthe original view of Corbellini and Greg~rini,~~ who assumed theintermediate production of glyceraldehyde, thus,GH*OH H*QH*OH HG-CO-G-OH- 211 0 HO*H,C-C*OH + G*OH a+ HO*H,C--C-O--CHHO-CM + OKojic acid.and receives further support from the formation of kojic acid fromdihydroxyacetone by A . Oryxce reported by H. Katagiri and K.Kitahara 7* (compare bacterial production of kojic acid, p. 310).In a further series of papers H. Raistrick and his co-workersreport considerable extensions of their already very comprehensiveexamination of the metabolic products of moulds.From mediain which have grown Penicilliurn brevi-compactum and relatedspecies there have been isolated two mycophenolic acids6 7 Biochem. Z., 1933, 264, 109; A., 1082.6 8 J . Agric. Chem. SOC. Japan, 1932, 8, 264; A., 637.69 K. Katagiri and K. Kitahara, Mem. Coll. Agric. Kgoto, 1933, No. 26,70 A. di Capua, Gazzetta, 1933, 63, 296; A., 983.71 0. E. May, G. E. Ward, and H. T. Herric, Zentr. Bakt. Par., 1932, 86,72 Biochem. Z., 1933, 266, 82; A., 1332.73 Gazxetta, 1930, 60, 244; A., 1930, 959.74 Mem. Coll. Agric. Kyoto, 1933, No. 26, 1; A., 638.1 ; A., 638.IT, 129; A., 1932, 1168POLLAXD AND PRYDE. 319(C17H2006),753 77 3 : 5-dihydroxyphthalic 76 3 : 5-dihydroxy-2-carboxybenzyl methyl ketone, 3 : 5-dihydroxy-2-carboxyphenyl-acetylcarbinol, and a hydrated form of 3 : 5-dihydroxy-2-carboxy-benzoyl methyl ketone.77 Luteic acid, previously reported asproduced by P. Zuteum grown on glucose media, has now beenobtained from a number of other sugars, glycerol, and succinicand citric acids when used as sole sources of carbon for the organ-ism.78 On a glucose-sodium nitrate medium, P. griseo-fulvunaproduces 2-hydroxy-6-methylbenzoic acid, mannitol, fumaric acid,and the newly recognised product, gentisic acid (2 : 5-dihydroxy-benzoic acid). 79I n an examination of Helminthosporium gramineum, the organismresponsible for stripe disease in barley, there have been isolatedfrom the mycelium helminthosporin (4 : 5 : 8-trihydroxy-2-methyl-ant hraquinone) and hydroxyisohelminthosporin (probably 1 : S-dihydroxy-2-hydroxymethylanthraquinone).80 A related substance,cynodontin (probably 1 : 4 : 5 : 8-tetrahydroxy-2-methylanthra-quiiione), occurs in the mycelium of H .cynodontis and H . euchEmnmalBiochemistry of the Higher Plants.3IineraZ Nutrition.-The multiplicity of factors influencing therate at which nutrients pass from the external medium into plantroots continues to be emphasised Many workers have attemptedto discover direct relationships between the concentration of indi-vidual ions in the nutrient and the rate of entry into the plant.For very dilute solutions a straight-line relationship may possiblyexist, but under conditions obtaining in soils and still more in waterculture experiments the mutual effects of solute ions both within andwithout the plant root become operative. Further, the assimilativeand metabolic processes within the plant exert a powerful, if indirect,influence over the mechanism of intake by the roots.Thus S. A.Yaxinos 82 shows that with dilute solutions of individual nutrients theintake curves were of very similar form t'o the dry matter productioncurves in point of view of both gradient and time distribution. Whenall nutrients were supplied simultaneously in a necessarily more75 P. W. Clutterbuck, A. E. Oxford, H. Raistrick, and G. Smith, Biochnz.76 A. E. Oxford and H. Raistrick, ibid., p. 1907 ; A., 189.7 7 A.E. Oxford, H. Raistrick, and P. W. Clutterbuck, ibid., 1933, 2'7,7 8 J. H. Birkinshaw and H. Raistrick, ibid., p. 370; A., 752.79 H. Raistrick and P. Simonart, ibid., p. 628; A., 949.80 J. H. V. Charles, H. Raistrick, R. Robinson, and A. R. Todd, $bid.,81 H. Raistrick, R. Robinson, and A. R. Todd, ibid., p. 1170; A., 1082.82 2. Pjfanx. Diing., 1933, 28, [A], 1; B., 402.J . , 1932, 26, 1441; A., 1932, 1289.634, 654; A., 949.p. 499; A , , 752320 BIOCHEMISTRY.concentrated solution, nitrogen, phosphorus, and potassium enteredthe roots at rates which were relatively in advance of the dry matterproduction curves. The intake of magnesium and calcium laggedbehind the increasing dry matter yields, An interesting exceptionto the general nature of base intake appears in the case of sodium,which penetrated at a very much enhanced rate from the more con-centrated mixed nutrient.In experiments with a somewhatsimilar object G. Pfutzer 83 shows further confirmation of the lackof direct relationship between nutrient concentration and intakeexcept over small ranges of concentration. Working with tomatoplants in soils of different moisture content, E. M. Emmert andF. K. Ball 84 also illustrate the different extents to which the intakeof individual nutrients is affected by changing levels of water supply.Where moisture contents were low, the nitrate intake was notdepressed, but that of phosphate declined considerably. Thiseffect, combined with the depressive action of a lowered water supplyon the general metabolism of plants, resulted in a high nitrateaccumulation within the plant tissues.In an examination of thepotash nutrition of plants I). R. Hoagland and J. C. Martins5 indicatea general proportionality between the intake of potassium bycrops and the potassium concentration of the soil solution in any onesoil, i.e., under conditions in which the proportional effect of othersolutes would be generally similar. The relationship does not,however, apply to soils of different type in which the proportions ofother ions would differ very considerably and also would be affectedto different relative extents by natural changes in the water contentof the soils. That the intakes of water and of dissolved ions areindependent processes is shown by experiments of M.GrctEanin.8sMoreover the rate of water intake would appear to be approximatelyinversely proportional to the concentration of dissolved salts.Assirnilution and Utilisation of Nitrogen.-There has been noabatement of interest during the year under review in the problem ofthe relative effects of the intake of nitrogen as nitrate and as am-monia.88 A series of papers by J. W. Shive and his co-workersaffords good illustration of the fact that the growth stage of theplant is an important factor influencing the utilisation of these twoforms of nitrogen. the young plants Thus in the case of83 Landw. Jahrb., 1932, 76, 745; B., 403.84 Soil Sci., 1933, 35, 295; B., 403.86 Compt. rend., 1932, 105, 899; A,, 101.87 P. Mad, P.J. Maze, jun., and R. Axionnaz, Compt. rend. Soc. Biol.,8 8 Compare Ann. Reports, 1932, 255.89 A. C. Sessions and J. W. Shive, Soil Sci., 1933, 85, 355; B., 664.Ibid., 1933, 38, 1; B., 804.1933, 112, 852; B., 1027POLLARD AND PRYDE. 321grew equally well in nutrients containing high concentrations ofnitrate or of ammonia, but at a later stage the nitrate produced themore rapid growth. The ammonia and nitrate contents of the plantsin the respective cases varied with the concentrations in the media.Plants supplied with nutrients in which the ratio NH;: NO,’was high had relatively higher proportions of soluble organic-and total organic-nitrogen and lower amounts of inorganic-nitrogen,than wits the case when nitrates predominated in the nutrient.I nanother communication the effects of supplying approximatelyequal proportions of ammonia and nitrate are recorded. Underthese conditions also the intake of ammonia reaches a maximum inthe early stages and that of nitrate at the flowering stage of theplants. The maximum rate of absorption of ammonia is muchhigher than that of nitrate. The total nitrogen intake curve there-fore shows two maxima, the earlier being mainly dependent onammonia intake and the later on that of nitrate. Throughout thegrowth period there is no actual cessation of absorption of eitherform of nitrogen. Very similar conditions prevail in the case ofbu~kwheat,~l except that the ammonia maximum is not attaineduntil the beginning of flowering and that of nitrate after the floweringstage.Ammonia absorption is much more marked in buckwheatthan in oats, the maximum rate of absorption of ammonia exceedingthat of nitrate by about 600%.Differences both in dry matter production and in the distributionof nitrogen in cotton plants result from the use of nitrate and am-monia as nitrogen sources.Q2 The former induced the heavier cropproduction, but when concentrated media were used the totalnitrogen content of the “ ammonia-plants ” was the higher. Withdilute solutions there was little difference between the total nitrogencontents in the two series of planb, but significant differences in theaccumulation of nitrogen in roots, stems, and leaves indicated con-siderable differences in metabolic rates. H. C. M.Jacobson andT. R. Swanback03 record an alternating dominance of iiikake ofammonia and of nitrate by tobacco similar to that which is recordedabove for oats and buckwheat, from nutrients containing bothforms of nitrogen. During hot seasons the plants wilted morereadily as the proportion of ammonia in the nutrient increased.In cases in which ammonia formed the sole source of nitrogen,plants were stunted, were very susceptible to root rot infection, and90 A. L. Stahl and J. W. Shive, Soil Sci., 1933, 35, 375; B., 664.91 Idem, ibid., p. 469; B., 727.92 K. T. Holley, T. A. Pickett, and T. G. Dulin, Georgia Agric. Exp. Sta.93 Plant Physiol., 1933, 8, 340; B., 646.Bull., 1931, No. 169; B., 36.REP.-VOL. xxx. 322 BIOCHEMISTRY.contained abnormally low proportions of calcium.Ammonia,toxicity towards tobacco is shown by experiments of A. B. Beau-mont 9* to become apparent in media containing more than 6 p.p.m.of nitrogen in this form, and to be reduced by additions of sodiumnitrate; the effect iiicreases with the amount of nitrate supplied.In the early stages of growth, injury by ammonia may also bereduced by additions of calcium carbonate, but with increasing agethe effect of the ammonium ion again predominates and the drymatter yield is adversely affected. It is concluded that ammonia,toxicity is due to disturbed metabolism rather than to its physio-logically acidic action. In sand-cultured apple trees rapid ammoniaintake is associated with a lower reductase activity and also with ahigher hemicellulose content than when normal proportions ofnitrate are being as~irnilated.~~ The effect of light conditions onnitrogen assimilation is examined by L.S. L~barskaja,~~ whoseexperiments show that the reduced utilisation of nitrogen by sugarbeet seedlings in darkness is much more marked in the case of am-monia than of nitrates. Also the intake and utilisation of ammon-ium nitrate resembles more closely that of other ammonium saltsthan that of potassium nitrate.The biological process of detoxication of ammonia within the plantsystem is examined and brought into close relationship with thepH of the sap by M. Kult~scher.9~ The “ammonium plants ” ofRuhland and Wetzel are those with highly acid saps in which am-monia is neutralised directly by organic acids and stored largely inthe form of ammonium salts.In such plants deamination anddeamidation is marked. Saps having pH > 6.0 are associated with“ amide plants ” in which ammonia is largely converted into thestorage form of amide. The “mixed type” consists of plants ofintermediate ranges of pH associated with varying proportions ofammonia and amide in the stored nitrogen. No direct correlationbetween pH and the NH, : amide ratio is apparent, and it must be con-cluded that factors other than sap reaction are operative in decidingthe form of storage of nitrogen prior to elaboration into proteins.A number of recent investigations are concerned with the effectof external conditions on the nitrogen metabolism of plants and theclosely related process of carbon assimilation.In one of these,G. T. Nightingale 98 records the effects of temporary (10 day) changes94 Proc. 2nd Internat. Cong. Soil Sci., 1932, 4, 65; B., 85.s5 V. A. Tiedjens and M. A. Blake, New Jersey Agric. Exp. Sta. Bull., 1932,s6 2. Pflanz. Dung., 1933, [A], 28,340; A., 647.3 7 Plunta [Z. wisa. SWl.3, 1932, 17, 699; A., 197.OS Bot. Gaz., 1933, 95, 35; A,, 1215.No. 647; B., 981POLLARD AND PRYDE. 323of temperature on the metabolism of tomato plants. At the lowertemperature examined (13"), although nitrate absorption and trans-location took place freely, the reductase activity of the plants waslow and protein synthesis retarded. Simultaneously there was con-siderable accumulation of carbohydrates, notably starch, leaveswere deficient in chlorophyll, and stems showed anthocyanincoloration.In the absence of external supplies of nitrogen theseconditions were intensified. Increase of temperature to 21" per-mitted greater utilisation of carbohydrates and synthesis of proteineven in plants receiving no additional nitrogen. With externalsupplies of nitrate both processes were still more accelerated.Assimilation, translocation, and reduction of nitrate were muchmore rapid than at 13". At a still higher temperature (35") carbo-hydrate decomposition became extremely rapid and was associatedwith protein degradation even where external supplies of nitratewere available. Exhaustion of the plants soon occurred and wasaccelerated in those having access to nitrate supplies.The effect of varied rates of carbohydrate and nitrogen meta-bolism on fruit production in tomatoes is considerable.By con-trolling carbon assimilation through alt,ered exposure to light andsimultaneously varying the nitrogen supply, V. M. Watts 99 demon-strates that conditions producing an appropriate balance of carbo-hydrate and amino-acids within the plant system are necessary toensure optimum fruiting. Thus excessive exposure to light,with its accompanying rise in carbohydrate formation, induces anincreased dry matter production but lowers the nitrogen content ofthe plant and especially the smino-acid fraction : fruitfulnessdeclines and a weak woody growth of stems takes place. At theother extreme, in which the nitrogen supply is maintained at a highlevel, but exposure t o light is restricted, the ratio of amino-acid tocarbohydrate tends to become unduly high, and unfruitful plantsshowing a very succulent form of growth are produced.Differences in the level of nitrogen metabolism in highly manuredapple trees are manifest in differences in the lipin- and residual-nitrogen fractions in the leaves.Corresponding changes in carbo-hydrate metabolism are characterised by the nature of the reserve(insoluble) matter.lPotassium and Plant Growth.-The influence of the level of potashsupply to plants on their water economy is emphasised in severalrecent publications. Deficiency of potassium tends to lower the totaldry matter production of many plants but a t the same time to1 N.W. Stuart, New Hampehire Agric. Exp, Sta. Tech. Bull., 1932, No. 50;$0 Arkansas Agric. Exp. Sta. Bull., 1931, No. 267; B., 1931,243.A., 102324 BIOCHEMISTRY.reduce, to a greater extent, the water content of the tissues. Thisis ascribed by K. Schmalfuss2 to the effect of potassium in modi-fying the colloidal nature of protoplasmic constituents, and thusexerting a partial control over the proportions of “free” and“ bound ” water in the tissues. The practical result of this actionis shown by the increased drought resistance of plants receivingadequate supplies of potas~ium.~~Although it is generally recognised that a large proportion of thepotassium content of plants remains in a soluble condition, it wouldappear that considerable amounts have undergone elaborationinto complex compounds.Electrodialysis reveals that 35% ofthe total potash of young pea plants, and only 14% of that in olderplants, migrates to the cathode, and 16% and 18% respectivelyappear a t the anode, these proportions being independent of thenature of the membrane used. I n discussing these results S. L.Inosemzev indicates that the dialysable fraction of the potassiumcontains considerable amounts in organic combination. In re-cording the influence of the supply of potassium on the carbonassimilation of the leaves of cereals G. Gassner and G. Goetze drawattention to the fact that in curves relating potash supply withassimilation (rising with increasing potassium concentration to amaximum point and subsequently declining), the maxima aredifferent and characteristic for each cereal examined.Intake and Utilisation of Phosphorus by Plants.-In addition tothe known assimilation of inorganic phosphates by plants, manyworkers have obtained indications that a variety of organicphosphorus compounds of varying complexity may also be utilised.Recent work of J.Weissflog and H. Mengdehl adds definitionto this conception. Various types of phosphorus compounds areexamined, and, according to their effects on maize plants, classifiedinto three principle groups : (1) phytin group, which includesphytic and iiucleic acids ; (2) ortho-group, comprising phosphoricacid, hexose &phosphate, etc. ; and (3) ester group, which includeshexose monophosphates, glyceryl and sucrose phosphates.Group( 2 ) was the most effective as a source of phosphorus, as judgedby the gross weight of the crop and its phosphorus content : (1)and (3) produced similar dry matter yields, but the percentage ofphosphorus in the plants was greater with (3) than with (1). ThePhytopath Z., 1932, 5, 207; A., 545; 2. PJEanz. Dung., 1933, 28, [A],3 M. A. H. Tincker and F. V. Darbishire, Ann. Bot., 1933, 47, 27; A., 436.330 ; A., 649.0. Tornau and K. Meyer, J. Landw., 1933, 81, 175; B., 680.Ergebn. Veg. Lab.-arb. Prianischnikov, 1930, 15, 85; A., 650.Planta [Z. wiss. Biol.], 1933, 20, 391; B., 1026.7 Ibid., 1933,19, 182, 242; A., 648POLLARD AND PRYDE. 325order of efficiency of the organic substance was approximatelythat of the susceptibility to decomposition by the phosphatase ofthe maize.Plants utilising the ortho-group contained the greatestproportions of orthophosphates. Inferior utilisation of phytic andnucleic acids was indicated by accumulations of soluble organicand insoluble phosphorus in the plant roots. Among the inorganiccompounds examined, pyro- and meta-phosphates were absorbedbut were converted into ortho-phosphates before leaving the rootsystems. Hypophosphites and phosphites, although passing freelyinto the plant system, were not utilised.The nature of the phosphorus compounds in plants has beenexamined by a variety of methods. Earlier work frequently led toconfusing results, since the methods of isolation often involveddecomposition of the substances in question.To overcome thisdifficulty H. Magistris has examined the separation of organicphosphorus compounds by dialysis, and the effects of varioussalts on the process.8 The general association of seed formationwith phytin synthesis in plants is reflected in the number of recordsof investigation of the phosphorus distribution of seeds. In horsebeans 9 and in hemp seed lo the proportions of organic and inorganicphosphorus extractable with dilute acids aiid alkalis indicate thepresence of phytic, nucleic, and inorganic phosphorus in the seed.In sunflower seed more than three quarters of the total phosphorusexists as phytin.ll The phytin-lipin-nucleic- and inorganic phos-phorus of wheat grain is examined by T.S. Andrews l2 and of thewhole plant in various stages of growth by F. Knowles and J. E.VVatkin.13 The latter show that during the four months prior toharvesting the gross weight of lipin-phosphorus in the stems rosesteadily until ear emergence and subsequenkly declined to a fairlyconstant value. In the ear the lipin weight was small and re-mained a t a steady figure until shortly before harvesting. Therewas a steady transference of phytin and inorganic phosphorusfrom straw t o ear as the grain matured. A close parallelism existedbetween variations of the phytin-phosphorus and protein-nitrogenthroughout the period examined. In maize14 phytin is absentfrom all parts of the plant until after pollination, and then appears8 Biochem.Z., 1932, 253, 64, 81; A., 1932, 1181, 1203.9 E. Mnich, Bull. Acad. Polonube, 1931, [B], 123; A., 1932, 1181.10 E. Pischinger, ibid., 1932, [B], 37 ; A., 874.11 A. Goldovski and A. Bozhenko, iklasloboino Zhir. Delo, 1932, No. 7, 30;12 Ind. Eng. Chem., 1932, 24, 80; B., 1932, 574.13 J . Agric. Sci., 1932, 22, 755; A., 101.14 E. E. de Turk, J. 1%. Holbert, and B. W. Howk, J . Agric. Res., 1933,A., 1092.46, 121 ; A., 545326 BIOCHEMISTRY.only in the seed. During germination phytin disappears fromthe seed a t an early date. Little variation occurs in the percentageof lipin- and acid (0.2 yo HC1)-insoluble organic phosphorus duringgrowth of the plant. Translocation of phosphorus within theplant took place only with inorganic and acid-insoluble organicfractions.Mineral Nutrition and the Incidence of Disease in Plants.-Avery definitely increasing interest is being shown in this aspect ofplant pathology.Economic pressure has stimulated enquiry intothe quantitative as well as qualitative aspects of manurial problemsand brought in its train numerous incidental observations of thehealth condition of plants in relation to nutrition. Also the con-tinuously increasing outlay in curative measures has turned atten-tion much more emphatically to the possibility of minimising cropinjury by the often less costly method of adjusting nutrientconditions.For many years the use of potash fertilisers has been recognisedas a satisfactory method of reducing the susceptibility of plantsto certain diseases.It is becoming increasingly apparent that,whilst the influence of potassium in this respect is a very importantone, much more may depend on the relative proportions in whichvarious mineral nutrients are supplied to plants than on the specificpreventive capacity of any one of them.The effect of potassium in reducing injury by yellow rust inbarley, drought spot in oats, leaf roll in potato,15 rust in wheat,16stripe disease in potatoes,17 and wilt and rust in cotton 18 is recordedthis year. The effect of anions associated with potassium on itsaction in increasing disease resistance is shown by E. Lowig.lgThe infection of cereals by Erisiphe graminis is reduced to a greaterextent by applications of potassium silicate than by the carbonate,chloride, or sulphate, although no difference in manurial action isapparent.I n a number of instances, the proportion of calcium relative toother nutrients, and also in its relation to the reaction of the medium,appears to be concerned in the degree of infection by fungal diseasestogether with other diseases falling under the general heading of“ nutritional disorders.” The “ damping off ” of soya bean seed-lings appears to be influenced much more by calcium deficiency15 L.Kratschmer, Ernahr. PJlanxe, 1933, 29, 264; B., 981.16 W. Acker and F. Konig, ibid., p. 101; B., 518.17 Kostlin, ibid., p. 48; B., 439.18 V. H. Young, G. Janssen, and J. 0. Ware, Arkansas Agric. E’xp. Sta.19 Emiihr. PJanze, 1933,29, 162; B., 646.Bull., 1932, No.212; B., 201POLLARD AND PRYDE. 327than by the reaction of the nutrient.20 Calcium deficiency oran excessive MgO : CaO ratio is associated with severe attacksof root rot in tobacco,21 with " sand-drown " in tobacco and maize,22and with leaf wrinkle in soya beans.23 " Speck " disease in oatsis primarily the outcome of manganese deficiency, but is accentuatedby unbalanced ionic ratios in the nutrient, notably that of K : Ca.24Sulphur deficiency is the cause of, or at least a heavily contributingfactor in, the tea '' yellows " disease.25 An interesting investiga-tion of the effects of unbalanced proportions of the three principalfertilisers on the infection of tomato plants by Fusarium Zycopersiciis recorded by H. Ahmet.26 The susceptibility of the plants variedwith nutrient supplies in the decreasing order : potassium deficiency,low phosphate, excessive nitrogen, excessive phosphate, excessivepotassium, low nitrogen.An interesting resume of the effects ofunbalanced nutrition, of the absence of individual nutrients, andof other growth conditions on the distribution of plant disease isgiven by F. Labro~sse.2~ANIMAL BIOCHEMISTRY.Carbohydrate Utilisation in Muscle and Yeast.THE period under review has seen important developments inthe elucidation of the chemical processes of carbohydrate utilisationin both muscle and yeast. A new scheme of fermentation formu-lated by the lat,e G. Embden and his collaborators2* has beenaccepted and developed by 0. Meyerhof and his school, and shownto be equally applicable to the muscle and yeast processes.It will be remembered that 0.Meyerhof, K. Lohmann, and R.Meier,Z9 in an important paper dealing with carbohydrate form-ation during the recovery process in the intact muscle of the frog,showed that, apart from lactic acid, only pyruvic acid out of aconsiderable number of substances investigated, produced a recoverysynthesis of sugar with oxygen uptake. Later, E. Toeniessen20 W. A. Albrecht and H. Jenny, Bot. Gaz., 1932, 92, 263; B., 1932, 200.2 1 T. R. Swanback and H. G. M. Jacobson, Scknce, 1933,77, 169; B., 440.22 A. Gehring, Ernahr. Pjlanze, 1932, 28, 101 ; B., 1932, 569.23 E. W. Hopkins, Plant. Physiol., 1933, 8, 333; B., 644.24 H. Lumdegardh, Medd. Centralanst. Pcirscisksr Jordbruks., 1931, No.403 ;25 H. H. Storey and R. Leach, Nyassaland Dept. Agric. Bull., 1932, No. 3;26 Phytopath. Z., 1933, 6, 49; R., 840.27 Ann. Agron., 1932, 2, 774; A., 437-28 G. Embden, Deuticke, and Kraft, Klin. W O C ~ . , 1933, 12, 213.a@ Biochem. Z., 1925, 157, 469; A*, 1925, i, 7-27.B., 1027.B., 245328 BIOCHEMISTRY.and E. Brinkmann30 demonstrated that pyruvic acid was rapidlyremoved when perfused through the musculature of a rabbit.E. M. Case and R. P. Cook31 isolated pyruvic acid and methyl-glyoxal from frog and rabbit muscle and formed the opinion thatboth participated in some muscle process. The occurrence ofpyruvic acid was further investigated by Case,32 who found thatit was not an intermediate in lactic acid formation, nor was it aproduct of the oxidation of the latter acid.The possible r81e ofpyruvic acid was further emphasised by A. Hahn and W.Haarmann 33 when they found that thoroughly washed muscle(free from lactic dehydrogenase) readily produced pyruvic acidfrom fructosediphosphoric acid. Numerous well-known investiga-tions by C. Neuberg and his collaborators have established theimportance of pyruvic acid in many yeast and bacterial fermenta-tion processes.It was shown by K. Lohmann34 and by F. Lipmann and K.Lohmann35 that, when muscle tissue is minced in the presence offluoride and added glycogen or starch, only a part of the hexose-phosphoric acid ester which accumulates is the true Harden-Youngfructose-I : 6-diphosphoric acid. A considerable part, and some-times even the whole, is found to be present in the form of an esterof the same elementary composition but possessing a much greaterresistance to acid hydrolysis.In fact Lohmann called it the‘‘ unhydrolysable ” ester. The monophosphoric esters isolated byRobison, Neuberg, and Ernbden were converted into this resistantester by taking up one equivalent of phosphate from muscle extractscontaining fluoride.Early in the year under review G. Embden and his colIeagues36found that a constituent of the Lohmann “ unhydrolysable ” esterwas glyceric acid-monophosphoric acid (phosphoglyceric acid).This acid was first encountered by R. Nilsson 37 in 1930 and isolatedas the barium and strychnine salts during his studies of the enzymicbreakdown of carbohydrate by yeast.At about the same timeas Embden’s discovery another constituent of the Lohmann esterwas isolated in Meyerhof’s l a b ~ r a t o r y . ~ ~ This was Z-a-glycero-30 2. physiol. Chern., 1930, 187, 137; A., 1930, 637.81 Biochern. J., 1931, 25, 1319; A., 1931, 1184.32 Ibid., 1932, 26, 759; A,, 1932, 875.33 2. Biol., 1930, 90, 231; A., 1930, 1064; see also A. Hahn, ibid., 1933,34 Biochem. Z., 1930,222,324; A., 1930, 1210.37 Arkiv Kemi, Min., Geol., 1930, 10, [A], No. 7; A., 1930, 641.38 Nature, 1933,132, 337, 373; A., 1074; 0. Meyerhof and D. McEachorn,94, 97; A., 1194.35 Ibid., p. 389; A., 1930, 1210. 86 L O C . cit.Biochem. Z., 1933, 260,417; A., 742POLLARD AND PRYDE. 329phosphoric acid. Thus it appeared that the Lohmann ester con-sisted of equimolar proportions of phosphoglyceric acid and glyccro-phosphoric acid :( BO),PO*O*CH,*CH (OH) *CO,H + (HO),PO*O*CH,*CH( OH) *CM,* OH.Embden and his colleagues showed that phosphoglyceric acid istransformed into pyruvic acid by minced muscle.Moreover,although lactic acid was not produced by the addition of eitherpyruvic acid alone or a-glycerophosphoric acid alone to muscleextracts free from carbohydrate, it was found that the simultaneousaddition of phosphoglyceric acid and a-glycerophosphoric acid tomuscle brought about an increased formation of lactic acid. Purther-more the addition of pyruvic acid and of cc-glycerophosphoric acidtogether did lead to the production of lactic acid, the amount ofthe latter formed being equivalent to twice the pyruvic acid whichhad disappeared, thus :CH,*@O*CO,H + C€€,(OH)*CH(OH)*CH,*O~PO(OH), +2CH,*CH(OH)*CO,H + H3PQ,.Further insight into the probable course of the breakdown ofsugar in muscle is obtained from a study of the behaviour of triose-phosphoric acid.M. 0. L. Fischer and E. Baer39 have recentlysynthesised glyceraldehyde-y-phosphoric acid and the dextro-component of their racemic compound has been shown by C. V.Smythe and W. Gerischer*O to undergo fermentation by yeastwith an initial velocity a t least as great as that of glucose of thesame molar concentration. Moreover Embden has corroboratedthe formation of lactic acid on adding glyceraldehyde phosphoricacid to muscle tissue.0. Meyerhof and W. Kiessliiig 41 have shownthat in muscle extracts one half (Le., one optical component ofthe racemic compound) is transformed into phosphoglycesic andglycorophosphoric acids. When sulphite, but no fluoride, is addedto the muscle extract, pyruvic and glycerophosphoric acids areformed. Meyerhof 42 points out that glyceraldehyde-y-phosphoricacid is the first synthetic ester to be converted into lactic acid aseasily as and by the same path as the naturally occurring phos-phoric acid esters.Embden summarised his views regarding these anagrobic processesin a scheme which may be represented as shown. According tothese views the formation of lactic acid in muscle is a true oxidation-reduction process in which the mutation of hexosediphosphoric39 Ber., 1932, 65, [El, 337; A., 1932, 364.40 Biochern.Z., 1933,260, 414; A., 750.4 1 Ibid., 964, 40; A., 1080; Naturwias., 1933, 21, 223; A., 528.42 LOG. cit.L 330 BIOCHEMISTRY.acid is involved, the oxidised product being pyruvic acid and thereduced Z-a-glycerophosphoric acid.These new ideas are equally applicable to the scheme of yeastfermentation. Thus C. Neuberg and M. Kobe143 have shown thatyeast and the lactic acid bacterium (B. DeEbrucki) transform phos-phoglyceric acid into pyruvic acid. When yeast is used, someacetaldehyde is also formed. As has already been mentioned,Nilsson 44 in 1930 isolated phosphoglyceric acid from dried yeastacting on hexosediphosphoric acid in the presence of glucose,acetaldehyde, and fluoride.Meyerhof and Kiessling 45 have shownthat in yeast maceration juice containing fluoride, hexosediphos-phoric acid is converted, as in muscle extract, into Lohmann’s“ unhydrolysable ” ester, i.e., to the equimolecular mixture ofa-glycerophosphoric acid and phosphoglyceric acid. Glyceralde-hyde-y-phosphoric acid behaves in similar fashion. Phospho-glyceric acid in fresh yeast, extract is converted into acetaldehyde,carbon dioxide, and phosphoric acid. The acetaldehyde is reducedto alcohol, not by reacting with glycerophosphoric acid, but pre-sumably 46 with the triosephosphoric acid arising from the inter-action of glucose and hexosediphosphoric acid. Phosphoglycericacid is formed simultaneously, thus :(HO),PO*O*CH,*CH(OH).CHO + CH,*CHO + H,O --+(HO),PO*O*CH,*CH(OH)*CO,K + CH,*CH,*OH.Meyerhof states that, apart from the presence of carboxylase, thisreaction provides the main difference between alcoholic fermentationand lactic acid formation in muscle, the difference lying essentiallyin the fate of the pyruvic acid. The latter in muscle extract isreduced to lactic acid by interaction with glycerophosphoric acid,whilst in yeast i t is split (by carboxylase) into carbon dioxide andacetaldehyde, the latter then being reduced as already explainedto alcohol.It should be noted that acetaldehyde can be replacedby other reducible systems, for example, methylene-blue. Theseobservations can be summarised in a scheme similar to that alreadygiven for the muscle process. One may finally remark that sodiumfluoride presumably inhibits in both the muscle and the yeast processthe splitting of phosphoric acid from phosphoglyceric acid and theconsequent formation of pyruvic acid.In muscle, iodoacetic acidinhibits the interaction of glycerophosphoric acid with pyruvic48 Biochem. Z., 1933, 260, 241 ; A . , 637.44 L O C . cit.45 LOG. cit.4 6 See also F. Zuckerkandl and L. ,lilessiner-Klebermass, Biochem. Z., 1932,266, 330; A., 188P O U D AND PRYDE.1 Phosphoglyceric acid331IJ.1 a-Glycerophosphoric acid + Pyruvic acid + H,PO,4. Lactic: acid +Triosephosphoric acidRepresentution of Embden's Fermentution Scheme.1 Hexosediphosphoric acid + 1 Glucose + 2 Phosphoric acid4 Triosephosphoric acidv / \ 2 a-Glyceroyhosphoric acid + 2 Phosphoglyceric acid2 Pyruvic acid + 2 Phosphoric acid(2 Acetaldehyde + 2C0,2 Phosphoric acid1 Glucose + +I2 Triosephosphoric acid + 2 Acetaldehyde2 Ethyl alcohol + 2 Phosphoglyceric acidErnbden-Meyerhof Scheme of Yeast Fermentation332 BIOCHEMISTRY.acid and the consequent formation of lactic acid and triosephos-phoric acid.There can be no doubt that the developments just describedrepresent a very considerable advance in our knowledge of carbo-hydrate utilisation, and they provide a welcome basis for theco-ordination of many previously isolated observations.Theinitiation of the new fermentation scheme i s a fitting finale to thegreat services which Gustav Embden has rendered to biochemistry.Through his untimely death the science loses one of its most activeand inspiring leaders.Deuminisation.H. A.Krebs has continued his work described last year 47 on themechanisms of deaminisation. Me finds 48 that the kidney is a siteof this process no less important than the liver ; indeed in the kidneyof the rat deaminisation is more rapid than in the liver of the sameanimal. The ammonia formed by the kidney in this process is thesource of the urinary ammonia, and the mechanism plays an impor-tant part in regulating the acid-base equilibrium. A number ofoptical antipodes of naturally occurring amino-acids (e.g., Z-alanine,Z-valine, d-leucine, d-phenylalanine, and d-histidine) are deaminisedby rat kidney some 10-20 times as rapidly as the naturally occurringforms.Perhaps the most interesting aspect of this part of Krebs’work is his direct experimental corroboration, using liver and kidneyslices, of the Neubauer-Knoop theory of oxidative deaminisationwith formation of the corresponding a-ketonic acid. In 1922Y. Kotake 49 recorded the isolation of phenyl- and hydroxyphenyl-pyruvic acids from the urine of rabbits receiving large amounts ofphenylalanine and tyrosine respectively, and N. F. Shambaugh,H. B. Lewis, and D. Tourtellotte 50 have more recently made closelysimilar observations. But apart from the phenylamino-acids,proof of the direct conversion into the keto-acids was lacking for thebulk of the amino-acids. In the present investigations Krebsrecords the practically quantitative conversions : alanine ->pyruvic acid, a-aminobutyric acid --+ a-ketobutyric acid, phenyl-alanine + phenylpyruvic acid, valine + dimethylpyruvic acid,leucine --+ E-ketoisohexoic acid, norleucine --+ a-keto-n-hexoicacid.The keto-acids have been isolated in each of the cases quoted.For the deaminisation oxygen is necessary and the process is shownto apply also to glycine, to dicarboxylic acids, and to diamino-acids.47 Ann. Reports, 1932, 29, 244.48 Klin. Woch., 1932, 11, 1744; A., 1074; 2. physiol. Chem., 1033, 217,49 Ibid., 1922, 122, 241; A., 1922, i, 1218.60 J . Biol. Chem., 1931, 92, 499; A., 1931, 1185.191 ; A., 856POLLBRD AND PRYDE. 333In a further series of experiments, using dog, guinea-pig, andrabbit kidney slices, Krebs 51 describes the isolation of ketoglutaricacid from d-glutamic acid, and of oxalacetic acid (yielding pyruvicacid and carbon dioxide) from aspartic acid.I n these cases it wasnecessary, in order to prevent further degradation of the a-ketonicacids, to treat the kidney slices with M/lOOO-arsenious acid.H. Manderscheid 52 has investigated the process of urea formationin the livers of two of the lower vertehratea-23. esculentu andT. g r a m . There is revealed a mechanism similar to that of themammalian liver, synthesis of urea being markedly accelerated bythe addition of ornithine. It is concluded that when urea is formedin the vertebrates it is always by way of the ornithine -+ citrulline + arginine mechanism. Manderscheid has found that with theexperimental methods used in these investigations, in birds and fish(the selachimns are a possible e x c e ~ t i o n ) , ~ ~ there is no measurablesynthesis of urea from ammonia and carbon dioxide. In birds thesynthesis of uric acid is not influenced by ornithine.Citrulline has now been isolated by M.Wada 54 from the productsof tryptic digestion of caseinogen under conditions which do not leadto the conversion of arginine into citrulline. On the other handF. Horn 55 records the formation of citrulline from arginine byB. pyocyuneus but not by B. coli. He postulates an arginine-desimidase distinct from arginase and trypin.Methionine.During the past .few years there have appeared several studies ofthe most recently discovered amino-acid methi~nine.~~ That thisamino-acid probably plays an important rBle in metabolism isindicated by the fact that it is the principal sulphur-containingamino-acid of caseinogen and probably also of egg-alb~min.~'H. D.Baern~tein,~~ who assumes that proteins contain no methoxylcompounds and no methylthiol compound other than methionbe ,gives methionine-sulphur contents varying from 26% of the totalsulphur (in secalin) to 90% (in casein). It is probably justified,therefore, to regard methionine as quantitatively the most important51 2. physiol. Chem., 1933, 218, 157 ; A., 976.58 Biochem. Z., 1933, 263, 245; A., 976.53 See A. Hunter and J . A. Dauphin&, Proc. Roy. SOC., 1924, [B], 97,54 Proc. Imp. Acad. Tokyo, 1932, 8, 367 ; A., 172.s 5 2. physiol. Chem., 1933, 216, 244; A., 753.56 Ama.Reports, 1928, 25, 233 ; 1931, 28, 234.5 7 N. W. Pirie, Biochem. J., 1932, 26¶ 2041 ; A . , 305.68 J . Biol. Chem., 1932, 97, 669; A., 1932, 1149.227; A., 1925, i, 104334 BIOCHEMISTRY.sulphur-containing amino-acid in ordinary diet and in proteinsother than sclero -pro teins .In 1924 J. H. M ~ e l l e r , ~ ~ the discoverer of methionine, showedthat in man it was readily oxidised to sulphuric acid. N. W.Pirie 6O has recently extended these observations and finds that in thedog methionine is oxidised to the same extent as cystine. 8-Ethyl-cysteine and S-benzylcysteine are not appreciably oxidised, whilstX-methylcysteine is oxidised but is too toxic to permit of accuratemeasurements. R. W. Jackson and R.J. Block 61 have shown thatthe addition of methionine to a diet deficient in cystine produces amarked increase in the weight of rats, and the same observers 62state that equally effective growth responses are obtained in ratsreceiving d- or Z-methionine or Z-formylmethionine, whereas d-formyl-methionine is devoid of action. Similar conclusions regarding thereplaceability of cystine by methionine may be drawn from theresults of A. White and H. B. Lewis,s3 who find that the increasedurinary nitrogen elimination on 8, low cystine diet following theadministration of bromobenzene is prevented by the addition ofeither Z-cystine or dl-methionine. The further results of B. W. Chaseand H. B. Lewis 64 show that in rats the absorption coefficient ofdl-methionine from the gastro-intestinal tract is slightly greater thanthat of cystine.They also find that the urines of rats receivingmethionine give the cyanide-nitroprusside test for the -S-S-linking.L. W. Butz and V. du Vigneaud 65 have recorded the formationof an interesting higher homologue of cystine by the decompositionof methionine with sulphuric acid. To this homocystine theyascribed the structureH02C*CH( NH,) *CH,*CH,*S*S*CH,*CH,*CH( NH,)*CO,Hand this has been confirmed by V. du Vigneaud, H. M. Dyer, andJ. Harmon 66 by the conversion of homocystine into homocysteine,by reduction with sodium in liquid ammonia, and subsequentmethylation to yield methionine. The same workers confirm thegrowth-stimulating action of methionine on low cystine diets, butthe main interest of their results lies in the finding that dZ-homo-59 J .Biol. Chern., 1924, 58, 373; A., 1924, i, 438.6O LOC. cit.61 Science, 1931,74,414; A., 1932, 83; J . Biol. Chem., 1932, g8, 465; A.,6% Proc. SOC. Exp. Biol. &feu?., 1933, 30, 587; A., 975.63 J . Bwl. Chem., 1932,98,607; A., 89.64 Ibid., 1933, 101, 735; A., 1075.6 5 Ibid., 1932, 99, 136; A., 151.6 6 Ibid., 1933, 101, 719; A,, 1074.89POLLARD AND PRYDE. 335cystine likewise produces growth in rats in lieu of methionine orcystine. It may be that the action of homocystine is non-specificin simply supplying utilisable sulphur to the organism, but, as wasshown by B. D. Westerman and W. C. Rose,67 other disulphide acids,e. g., dithiodiglycollic, p-dithiodipropionic, and a-dihydroxy- p-di-thiodipropionic, are unable to support growth on cystine-deficientdiets in spite of their oxidation in the body.The results ofduvigneaud and his colleagues are therefore in harmony with the viewthat demethylation may occur in the intermediary metabolism ofmethionine with subsequent formation of homocysteine.Vitamin A .The impetus which from time to time organic chemistry gains frombiology has been demonstrated in several fields in recent years, butnowhere more strikingly than in the attempts to elucidate thestructures of vitamins A and C and cognate problems. In previousyears the relationship of vitamin A to natural polyene pigments ofthe carotene and related groups has been dealt with in this sectionof these Reports,68 but the interest has now widened considerablyand purely chemical problems of outstanding interest have arisen.Although these problems, and also those centring round the natureof vitamin C, may justly be regarded as biochemical in their incep-tion, they are perhaps now more appropriately dealt with in theorganic chemistry sections of these Reports. Thus the importantand voluminous structural work of R.Kuhn and P. Karrer and theirassociates on the polyene group is described elsewhere in thisvolume.69 The reader is referred to this account for details of therecent structural work on vitamin A and on carotene and relatedpolyene pigments. It will be seen that the C,, formula tentativelyadvanced for the vitamin by Karrer in 1931 'O still remains apossibility, although the C,,H,,O structure is now preferred.71According to the latter, vitamin A is derived from exactly one halfof the carotene molecule.The three a-, p-, and y-carotenes so fardescribed all possess growth-promoting action and it now appearsthat this action is dependent not on the number of double bondspresent in the parent polyene hydrocarbon, but on the presence of67 J . Biol. Chem., 1927, 75, 533; 1928, 79,413, 423; A., 1928, 87; 1928,1396.68 Ann. Reports, 1929, 26, 245; 1930, 27, 276; 1931, 28, 219; 1932, 29,248.8s P. 145.70 Ann. Reports, 1931, 28, 221.7 1 p. Karrer, 0. Walker, K. Schopp, and R. Morf, Nature, 1933, 132, 26;A., 805; see also F. H. Carr and W. Jewell, ibid., 131, 92; A,, 323336 BIOCHEMISTRY.at least one unmodified p-ionone ring.72 T.Moore 73 has publisheddata showing that P-carotene and a vitamin A concentrate are utilisedequally efficiently by the rat when administered at levels approach-ing the minimum dose, and it is concluded that P-carotene possessesa biological activity equal to that of about the same weight of“ pure ” vitamin A. It would be interesting, as Moore suggests, tocompare the activities of a- and y-carotene with that of @-caroteneat levels approaching the minimum dose. It is to be noted that,owing to the failure to secure crystalline preparations or derivatives ofvitamin A, absolute criteria of its purity are still lacking. The pre-paration of 3’. H. Carr and W. Jewel1 74 when administered to ratsin daily doses of 0.006 mg.gave slightly better growth than 0.001mg. (or 1 unit) per day of International Standard carotene. Thiswould seem to be the purest preparation of vitamin A so farobtained.Vitamin C.Last year an account was given of the progress made in the studyof vitamin C and the steps which had led to a tentative identificationof the vitamin with ascorbic a ~ i d . 7 ~ Later work has all tended tocodirm this identification, but the outstanding achievement in thisfield is the elucidation of the structure of ascorbic acid by E. L.E r s t and his collaborator^,^^ and the synthesis of the d- and E-forms(the latter is the natural form) of this acid and of structurally analog-ous substances by W. N. Haworth, E. L. Hirst, and their colla-borators 77 in the Birmingham laboratories.A detailed account ofthese brilliantly successful investigations i s given elsewhere in theseR e p ~ r t ~ s . ~ ~ It is sufficient to state here that the structurer-----o--~CH2( OH)*CH( OH)*CH--C==~--C:Odeduced by Hirst and his co-workers from degradation and otherexperiments has been fully substantiated by the later syntheses.Although it seems extremely probable that ascorbic acid is vitaminOH OH72 R. Kuhn and H. Brochann, Natumoiss., 1933, 21, 44; A., 278; Ber.,73 Biochem. J., 1933, 27, 898; A . , 871.74 L O C . cit.7 5 Ann. Reports, 1932, 29, 252.78 R. W. Herbert, E. L. Hirst, E. G. V. Percival, It. J. W. Reynolds, andF. Smith, J., 1933, 1270; A., 1143.77 R. G. Ault, I).K. Baird, H. C. Carrington, W. N. Haworth, R. Herbert,E. L. Hirst, E. G. V. Porcival, F. Smith, and M. Stacey, ibid., p. 1149; A.,1933,66, [B], 407; A., 431.275.7 8 P. 167POLLARD AND PRYDE. 337C, and numerous publications have appeared asserting or stronglysuggesting that this is it cannot yet be said that this is a cer-tainty. It is clear, however, that the biological investigation of thesynthetic material will place the matter beyond a doubt. At thetime of writing data are not available.J. L. Svirbely and A. Szent-Gyorgyi 80 find that the isopropylidenederivative of ascorbic acid 81 is a moderate antiscorbutic, whilst theacid recovered from it is fully active. E. L. IIirst and S. S. Zilva 82have examined the vitamin C activity of various preparations ofascorbic acid, and although considerable variations in activity werenoted,S3 it is found that active preparations can be regenerated frominactive, or much less active, oxidised preparations.A similarclaim is made by V. Demole.84 The general conclusion of Hirst andZilva is that it is much more probable that ascorbic acid is activeper se, than that the vitamin is associated with ascorbic acid and,like it, is reversibly oxidised and regenerated quantitatively. It isinteresting to note that ascorbic acid immediately after oxidationwith iodine shows little loss of antiscorbutic activity. This observ-ation corroborates earlier statements of S. S. Zilva 85 and of J.Tillmans and his collaborators 86 as regards the vitamin present indecitrated lemon juice, and further confirmation is furnished byS.W. J~hnson.~' The explanation now favoured by Hirst andZilva is that originally advanced by Tillmans, namely, that thevitamin itself, whilst retaining most of its activity, may be reversiblyoxidised. It will be evident that the labile structure of ascorbic acidwell fits it for such a r81e.Vitamin B,.The identity of the various crystalline preparations 88 with whichvitamin B, activity is associated is still uncertain. B. C. P. Jansenand his collaborators 89 now state that the analytical data agree best79 See T. Mi'. Birch, L. J. Harris, and 8. N. Ray, Nature, 1933, 131, 273;A., 433; W. J . Dann, ibid., p. 274; A., 433; S. S. Zilva, ibid., p. 363; A.,433; A. L.Bacharach, ibid., p. 364; A., 433; L. J. Harris and S. N. Ray,Biochm. J., 1933,27, 680; A., 646.80 Biochem. J., 1933, 27, 279; A., 541.81 Ann. Reports, 1933, 29, 252.82 Biochem. J., 1933, 27, 1271; A., 1091.83 Compare L. J. Harris and S. N. Ray, Zoc. cit.84 2. physiol. Chem., 1933, 217, 83; A., 756.85 Biochem. J., 1927, 21, 689; A., 1927, 702.86 Ann. Reports, 1932, 29, 252.87 Biochem. J . , 1933, 27, 1287; A., 1090.88 Ann. Repwts, 1932, 29, 250.89 B. C. P. Jansen, J. P. Wibaut, Y. J. Hubers, and P. W. Wiardi, Rec.trav. chim., 1933, 52, 366; A., 645338 BIOCHEMISTRY.with the formula C12H,,02N,S,2HC1 for their air-dried vitaminhydrochloride.During the past few years a series of investigations of the physio-logical action of vitamin B, has been in progress in the Oxfordlaboratories.These are of considerable interest in view of the newfermentation schemes of Embden and Meyerhof. There has fre-quently been discussed in the literature a supposed associationbetween the antineuritic vitamin and certain aspects of carbohydratemetabolism, but i t is only recently that the work of R. A. Peters andhis co-workers has established a clear correlation. I n 1929 H. W.Kinnersley and R. A. Peters 90 found that the brains of polyneuriticpigeons have a high lactic acid content, and Peters with N. Gavri-lescu 91 showed a little later that the brain tissue from such pigeonshas in vitro a sub-normal oxygen uptake. The latter workers thenfound,92 on adding a concentrate of vitamin B, to the polyneuriticbrain tissue in witro, that the oxygen uptake was increased. Theeffect was obtained with the optic lobes and with lower parts of thebrain and later with the cerebral and higher parts.So far no othertissue is known to show this effect.N. Gavrilescu, A. P. Meiklejohn, R. Passmore, and R. A. Peters 93established the specificity of the effect in the presence of addedglucose or lactic acid. Experiments by A. P. Meiklejohn, R. Pass-more, and R. A. Peters 94 on birds recovering from polyneuritis aftertreatment with vitamin B, concentrate, showed that there was animprovement in the oxidative behaviour towards lactate of mincedbrain from such birds, corresponding to the disappearance of thenervous symptoms, and with this improvement the effect of addedvitamin B, concentrate in vitro diminished.It seemed, therefore,that vitamin B,, or a substance present in B, concentrates, wascapable of repairing the same defect both in the living bird and in theisolated brain.The behaviour with regard to pyruvic acid, a question of obviousinterest, differs from that shown towards lactic acid. Although thenormal pigeon's brain when minced gives a large oxygen uptake,and the avitaminous brain a low oxygen uptake, both in the presenceof pyruvic acid, the low uptake is not usually increased in this caseby vitamin B, concentrates in vitro, although results are somewhatvariable. The effect of vitamin B, seemed, therefore, to be related90 Biochem. J., 1929,23,1126; 1930,24,711; A., 1929, 1496; 1930, 963.9 1 Ibid., 1931, 25, 1397; A., 1931, 1338.92 Ibid., p.2150; A , , 1932, 200.93 Proc. Roy. SOC., 1932, [ B ] , 110, 431; A . , 1932, 644; Biochem.. J . , 1932,9d Proc. Roy. ij'oc., 1932, [B], 111, 391; A , , 1933, 1176.26, 1872; A., 195POLLARD AND PRYDE. 339to lactic acid formation, with the possibility that pyruvic acid playedsome secondary r81e. E. Boyland 95 has shown that the vitamin isnot a co-enzyme for lactic acid oxidation, and A. P. Meiklejohn 96has obtained no evidence of an increased disappearance of lactic acidcorresponding to the increased oxygen uptake.R. A. Peters and H. M. Sinclair 97 have continued these studies andfind that after previous washing of the avitaminous brain tissue itsoxygen uptake under the influence of vitamin B, is reduced.Thein vitro action of B, upon the avitaminous brain is abolished bycyanide and fluoride, whilst pyrophosphate interacts with B, andlactate to produce large rises in oxygen uptake over periods of from2 to 3 hours. On the other hand, hexosediphosphate and Robison’smonophosphate increase only the initial rate in avitaminous (as innormal) brains, but the effect is not sustained and there is no specificinteraction with the vitamin. Of particular interest is the observ-ation that a-glycerophosphate (but not the @-isomeride) increasesthe respiration largely, but the increase is not related directly to thespecific vitamin action. Freshly minced avitaminous brain givesno pyruvate reaction, but this appears strongly after shaking withlactate-Ringer solution a t pH 7.3.It is suggested that pyruvate isformed from phosphoglycerate, which would account for the observedready disappearance of the reaction in the presence of a-glycero-phosphate. Although, therefore, there is no evidence a t present tosuggest that the vitamin action is concerned with a-glycerophosphate,or directly with pyruvate, it is inferred that a-glycerophosphate isprobably one of the missing tissue substrates and that, since vitaminB,, lactic acid, and pyrophosphate appear to form a coupled oxidationsystem, the vitamin lack must affect more than one phase of cellularmetabolism.Carcinogenesis by Pure Hydrocarbons.In continuation of their work reported last year 98 J. W. Cookand his associates 99 have now isolated from a medium soft pitchan actively carcinogenic hydrocarbon in a state of purity.This is1 : 2-benzpyrene. It was obtained along with the inactive 4 : 5-benzpyrene, 1 : 2-benzanthracene, and perylene. The structuresof both benzpyrenes, and the physiological activity of the 1 : 2-isomeride, have been confirmed by synthesis: This new hydrocarbonis the most active carcinogenic substance now known and its iso-lation and synthesis constitute a remarkable advance in this field.95 Biochem. J . , 1933, 2’4, 786; A., 872. O6 Ibid., p. 1310; A., 1090.O 8 Ann. Reports, 1932, 28, 246.99 J. W. Cook, C. L. Hewett, and I. Hieger, Nature, 1932, 130, 926; A . ,Ibid., 1933, No. 6.86; J . , 1933, 395; A., 601340 BIOCHEMISTRY.Like the synthetic carcinogenic hydrocarbons previously described,1 : 2-benzpyrene contains the 1 : 2-benzanthracene nucleus.Itwill be remembered that 1 : 2-benzanthracene is itself inactive, orvery nearly so, but that it yields substances, sometimes of a veryhigh order of carcinogenic activity, by the substitution of alkylgroups at position 6 or of new rings in the 5 : 6-position. J. W.’Cook1 makes the interesting statement that a mouse tumourproduced by 1 : 2 : 5 : 6-dibenzanthracene has now reached the 67thtransplanted generation, and in rats the 40th generation has beenattained. Spindle-celled tumours, with metastases in the heartand other organs, have also been obtained by the injection of1 : 2 : 5 : 6-dibenzanthracene into fowls.Synthetic Gstrogenic Substances.It will be observed that the tricyclic phenanthrene ring systemis common to all the carcinogenic hydrocarbons so far described.This system is now known to be present in a considerable numberof naturally occurring substances of great physiological interest.Amongst these are the bile acids and sterols, vitamin D (calciferol),the ovarian hormones (cestriol and cestrone),2 the cardiac-stimulat-ing glucosides (strophanthin and digitoxin): and certain alkaloidssuch as morphine and codeine of the opium group, and the corydalisalkaloids and colchicine.It is for tohis reason that the results ofJ. W. Cook, E. C. Dodds, and C. L. Hewett are of great interest.These workers found that the synthetic substance l-keto-1 : 2 : 3 : 4-tetrahydrophenmthrene could induce estrous when injected intocastrated animals, and Cook and Dodds found that similar effectswere obtained with 1 : 2 : 5 : 6-dibenz-9 : 10-di-n-butylanthraquinoland with the carcinogenic substances 5.: 6-cyclopenteno-1 : 2-benzanthracene and 1 : 2-benzpyrene. More detailed results havebeen presented by J. W. Cook, E. C. Dodds, C. L. Hewett, and W.Lawson,G who investigated a series of diols derived from 1 : 2 : 5 : 6-1 Contribution to “ Discussion on Experimental Production of MalignantTurnours,” PYOC. Roy. SOC., 1933, [B], 113, 275.This vol., p. 216.3 W. A. Jacobs and E. E. Fleck, J . Biol. Chem., 1932, 97, 57; A . , 1932,$48; see also W. A. Jacobs and N. M. Bigelow, ibid., 1033, 99, 521 ; A,, 278.4 Nature, 1933, 131, 56; A., 323.6 Communicated at the Meeting of the Royal Society, Nov.16tJh, 1933.Jbid., p. 205; A., 323341 POLLARD AND PRYDE.dibenzanthracene. These have the general formula shown. Ofthese the dimethyl, di-n-amyl, and di-n-hexyl compounds amH 2 5 & 2 1 : 2 : 3 : 4 - l-Keto-/v\ tetrahydro- I 11 f phenanthrene\A/inactive, whilst the intermediate diethyl, di-n-propyl, and di-n-butyl compounds are all highly active, the propyl derivative show-ing the maximum activity. I n addition to the compounds alreadymentioned, neoergosterol, calciferol, and ergosterol also exhibitsome estrogenic action, diminishing in the order cited. Thehydrocarbons are the least active of the estrogenic substances sofar investigated and it appears that the presence of oxygen-con-taining groups greatly increases the potency.I n all cases the activesubstances contain the phenanthrene ring system. S. Aschheimand W. Hohlweg,’ who record the presence of cestrogenic substlancesin extracts of bituminous materials, have no doubt encounteredthe same, or similar, hydrocarbons or derivatives of these. Thereis so far no evidence that carcinogenic substances are formed fromcestrin in the animal body, but, none the less, B. P. Wiesner andA. Haddow8 find that rats treated with naturally occurring ma-trogenic hormone show, in contrast to non-treated rats, a markedincrease in the rate of growth of implants of Jensen sarcoma.The OZstrin Group.I n a praiseworthy attempt to systematise nomenclature in themtrin group it has been suggested that the follicular hormonehydrate (trihydroxyestrin, theelol, emmenin) first characterisedby Marrian should be designated “ estriol,” and the follicularhormone (ketohydroxycestrin, theelin) of Butenandt similarlydesignated “ estrone.”It is satisfactory to record results which place beyond doubtthe relationship of members of the oestrin to the cholane group, forwhich assumed relationship much circumstantial evidence ad-mittedly existed.A. Butenandt, H. A. Weidlich, and H. Thomp-son 10 have converted cestriol (hormone hydrate) by fusion with7 Deut. rned. Woch.. 1933, 59, 12: A.. 870.8 Nature, 1933, 132, 97 ; A., 852.9 N. K. Adam, J. F. Dztniolli, E. C. Dodds, H. King, G. F. Marrim, A. S.Parkes, and 0. Rosenheim, ibia., p. 205.10 Ber., 1933, 66, [R], GO1 ; A.., 540342 BIOCHEMISTRY.potassium hydroxide into a phenoldicarboxylic acid ( C1,H,,O,)formed by fission of the five-membered ring between the twosecondary hydroxyl groups. Dehydrogenation of this dicarboxylicacid with selenium yielded hydroxy-1 : 2-dimethylphenanthrene,which was converted by distillation with zinc dust into 1 : 2-di-methylphenant hrene. The last -mentioned compound was alsoobtained by similar means from ztiobilianic acid of the cholane series.OH(Estriol.1 ; 2-Dimethylphenanthrene.A. Girard, G. Sandulesco, A. Fridenson, and J. J. Rutgers,llcontinuing their investigations of the sex hormones of pregnantmare’s urine,l2 have characterised equilenin l3 as C1,H1,02. Thishas a higher acidity than oestrone and apparently differs from itin having two aromatic rings in place of one. It would thereforehave the constitution shown and it represents a natural dehydrogen-ation product of oestrone. It is of interest to note that the reductionQH3 CHEquilenin. 3-Methyl- 1 : 2-cyclopentenophenenthrene.of the keto- and hydroxyl groups and the aromatisation of ring I11would yield a methylcyclopentenophenanthrene. Such a compound,C18H16, has been obtained from sterols and bile acids by 0. Diels,W. Gadke, and P. Kording l4 by dehydrogenation with selenium.0. Rosenheim and H. King l5 suggested that this hydrocarbon wasl1 Compt. rend., 1932, 195, 981; A., 98.la Ann. Reports, 1932,29, 242.l4 Annalen, 1927, 459, 1; A., 1928, 169.l6 J . SOC. Chem. Ind., 1933, 52, 299.l3 See A,, 1932, 433POLLARD AND PRYDE. 3433-methyl-1 : 2-cyclopentenophenanthrene and this has now beenconfirmed by its synthesis by G. A. R. Kon.l6 It is apparent,therefore, that a migration of the methyl group must occur duringselenium dehydrogenation, and the earlier observation of Dielsand Gadke 1 7 that more drastic dehydrogenation of cholesterolyielded the fully aromatic tetracyclic hydrocarbon chrysene,Cl8HI2’ must depend upon a further migration of carbon leading toring enlargement. A. Butenandt l8 has recorded the formation ofa hydrocarbon C18H14 by zinc dust distillation of cestrone. Thismay be impure chrysene.19 The hydrocarbon C17H14, 1 : 2-cyclo-pentenophenanthrene, has recently been synthesised by G. A. R.Kon 2o and independently by J. W. Cook and C. L. Hewett.,l Thisis Diels’s original hydrocarbon minus its methyl group. Thusremarkably interesting developments in the chemistry of thepolycyclic hydrocarbons have rapidly followed the biological prob-lems discussed in this and the preceding two sections of this Report.Flavins or Lyochromes ,The important newly-discovered group of flavin dyes or lyo-chromes is described elsewhere in this volume.22 The flavins arewater-soluble, nitrogenous substances which have been isolatedfrom human and canine liver and kidney (P. Ellinger and W..K0schara),~3 and from egg-white and milk whey (R. Kuhn and co-w o r k e r ~ ) . ~ ~ A flavin is also found associated with a protein in thenew oxidation enzyme isolated from yeast juice (0. Warburg andW. Christian),% and the co-enzyme (cytoflav) of I. Banga and A.Szent-Gyorgyi24 is stated by K. Laki25 to be identical with theflavin component of the Warburg oxidation enzyme. The pre-parations obtained by Kuhn and his co-workers possess an intensevitamin B, activity, and a preparation of lactoflavin (from whey)is claimed to have the highest B, activity so far recorded. TheAavins would seem to show every promise of a bright ‘future.A. G. POLLARD.J. PRYDE.16 J . SOC. Chern. I n d . , 1933, 52, 950.1 7 Ber., 1927, 60, 140; A., 1927, 241.1 8 Nature, 1932, 130, 238; A., 1932, 971.19 See J . SOC. Chm. Ind., 1933, 52, 268, 287.2o J . , 1933, 1081; A., 1153.22 P. 159.24 Biochem. Z., 1932, 246, 203; 247, 216; A., 1932, 537, 775.a5 Ibid., 1933, 266, 202 ; A., 1318.21 Ibid., p. 1098; A., 1299.23 This vol., p. 159

ISSN:0365-6217    

DOI:10.1039/AR9333000305

出版商:RSC    

年代:1933

数据来源: RSC

摘要:

SUB-ATOMIC PHENOMENA AND RADIOACTIVITY.THE eventful time in nuclear physics, characterised last year bythe discovery of the neutron and the hydrogen atom of mass 2,has continued. A particle, apparently the same in mass andcharge as the electron but of opposite sign, has been observed forthe fist time, and is now easily produced in the laboratory. The“ diplon ” or “ deuton,” the nucleus of the hydrogen atom ofmass 2, has proved itself a powerful projectile against the nucleiof light elements and is as promising a weapon as the proton. Theartificial disintegration of light nuclei has now reached a stagewhere the proton, the diplon, the neutron, and the a-particle eachhave their merits as projectiles. After their temporary capture,any of these four, or, in addition, the positive electron, may emergein the ensuing break-up.The number of isotopes found duringthe period under review is smaller than usual, but only because theelements still to be examined are very rare or difficult to study.Accurate determinations of atomic masses and isotopic abundanceshave been made. Interest in radioactivity, as formerly understood,has declined ; the work done has been mainly among the rare-earthelements. The penetrating radiation which had latterly come tobe described as “cosmic” has continued to be the subject ofwide study and interpretation. There is no dispute about itspenetration. The tendency, noted last year,2 to replace‘‘ radiation ” a t least in part by ‘‘ particle ” continues; “ COSI~~~C,”also, may conceivably be replaced by (‘ terrestrial.”Radioactivity of Xamrium.G.von Hevesy and M. Pahl have continued their work onsamarium. They find that 1 g. of the element emits 75 a-particlesper second; the half-period is in consequence 1.2 x 10l2 years.The particles have an initial velocity of 1.05 x lo9 cm. and a rangeapproximately of 1.1 cm. at 15”. This success has stimulatedinvestigations of the radioactivity of neighbouring rare-earthelements. M. Curie and S. Takvorian4 have fractionated amixture of neodymium and samarium oxides. In the fraction ina Bid., p. 313. 1 Ann. Reports, 1932, 29, 299.Nature, 1933,131, 434; A., 442; Z. Phy:j.ik, 1933, 83, 43; A., 762.Convpt. rend., 1933, 196, 923; A., 442RU SSELIi . 345which element 61 was to be expected, x-particles more penetratingthan those from samarium were observed.These are possiblydue to traces of element 61 a t a concentration which has provedto be too small to be detected spectroscopically.and lanthanum 33 4 are definitely inactive, although the lattersometimes shows an activity which has been traced to actiniumand its products.3.4 W. Yeh 6 confirmed the complete absence ofactivity in lanthanum but found with a Geiger-Muller counter avery feeble a-particle emission from neodymium. G . von Mevesyand M. Pah13 raised the question whether J. Joly's " hibernium "is not samarium. The latter found in the pleochroic haloes inArchEan mica two very small rings which corresponded with rangesof a-particles in air a t 15" of 1-14 and 2.1 em. (Rings of similarradii were later found also by S.Iimori and J. Yoshimura.)* Theymust be due to traces of unknown radioactive elements and arethe only rings that cannot be ascribed to known radio-elements.The element giving the smaller range and named " hibernium "may be samarium on account of the similarity of range. J. H. J.Poole would agree with this if range were the only criterion ; hecalculated, however, that if the speck of material in the mica issamarium the number of a-particles it has expelled since its im-prisonment is, in the happiest circumstances, only one-fifth of whatis the minimum to give the rings J. Joly observed; to identify thetwo it would have to be shown that the " hibernium " mica wasspecially sensitive to marking by a-particles. No name has beengiven to the emitter of the particles of range 2.1 em., nor has anyidentification of it with a supposed active element been made.It may be element 87 l o (which, on general grounds, is likely t oexpel a-particles) or element 61 or neodymium.6 Further work onthe radioactivity of these elements is necessary to test this view.If " hibernium " prove to be samarium, it can be calculated thatthe mica has acted as a detector of a radio-element which has ex-pelled an average of only two a-particles per thousand years-surelya record in sensitivity.Isotopes and Mass-spectra.During the year, knowledge of the isotopic composition of eightelements has been extended. The results are summarised inTable I.5 W.F. Libby and W. M. Latimer, J . Amer. Chem. SOC., 1933, 55, 433;6 Compt. rend., 1933, 197, 142; A., 882.7 Proc. Roy. Soc., 1922, [A], 102, 682; Ann. Reports, 1928, 25, 316.* Ibid., p. 315.9 Nature, 1933, 131, 654; A., 661.GadoliniumA., 204.lo Ann. Reports, 1932,29, 300346 SUB -ATOMIC PHENOMENA AND RADIOACTIVITY.Element, Atomicnumber.Zinc ......... 30Cadmium ... 48Tellurium ... 52Neodymium 60Samarium ... 62Europium ... 63Gadolinium 64Terbium ...... 65TABLE I.Minimumnumber ofisotopes.58855251Masses (nearest integer) of isotopes inorder of abundance, except whereparentheses are given.64, 66, 68, 67, 70114, 112, 110, 111, 113, 116, (108, 118)130, 128, 126, 125, 124, 122, 123, (127)142, 144, 146, (143, 145)152, 154, (147, 148, 149)151, 153(155, 156, 157, 158, 160)159The work on zinc and tellurium was done by K.T. Bainbridge.He found that the Zn65 and Zn69 ions measured by F. W. Aston l2were quite certainly hydrides of Zn64 and Zn68; the number ofisotopes must therefore be reduced to five. F. W. Aston’s order ofrelative abundance of these was confirmed. For tellurium, F. W.Aston’s analysis 13 was confirmed and extended ; Tela, Te123,Te122 with faint indications of Te12‘ were found. Te124 forms anisobaric triplet with the heaviest isotope of tin and the lightest ofxenon. The atomic weights of zinc and tellurium calculated fromthese mass-spectrographic data, 65.32 & 0-02 and 127.58 & 0.15respectively, are in excellent agreement with the Internationalvalues. In an investigation of the isotope effect in the spectrumof cadmium hydride, E.Svensson l4 found components belongingto Cd118 and Cd108 in much smaller abundance than those alreadyknown; their abundances were not noticeably different. After along and troublesome series of investigations, F. W. Aston l5 hasnow made a provisional analysis of some of the rare-earth elements.The results give at present only a rough estimate of relative abund-ance; the packing fractions have yet to be determined. Nd143and Nd145 have been found in addition to the more abundantNd142, Nd144, and Nd14s. Samarium gave a strong pair Sa152 andSa154 and a weaker triplet SalP7, Sa14*, Sa149. Europium consists ofEu151 and Eu153 in roughly equal abundance, in agreement withthe atomic weight 152.0.Gadolinium has five isotopes of un-determined abundance. Terbium is Tb159 only ; there is certainlyno Tb16l in the abundance which the atomic weight 159.2 suggests.The isotopic composition of seventy of the ninety-two elementshas now been investigated. Of the twenty-two which remain,numbers 43, 61, 84, 85, 86, 87, and 89 are very rare, numbers 88,90, and 91 are less rare but radioactive, and six, numbers 66-71,are rare earths. It is intended to investigate the rare earths,11 Physical Rev., 1932, [ii], 39, 847, 1021; A., 1932, 1099.12 Ann. Reports, 1928, 25, 306.14 Nature, 1933, 131, 28; A., 108.l3 IbicE., 1926,23, 280; 1932, 29, 303.l6 Ibid., 132, 930RUSSELL.347masurium, thorium, and protoactinium in the near future. Proto-actinium is the element whose atomic weight is most anxiouslyawaited. The determination by chemical methods is still lacking.If the value should be 231, as simple theory that pointwould be settled; if it should be 233, as has also been suggested,17material particles other than the a-particle must be expelled inthe actinium series, and these might be found with the techniquerecently developed for detecting protons, diplons, and neutrons.Twenty-two of the seventy elements investigated are simple,i.e., have one mass only. I n the range of mass-numbers 1-210,about one-fifth of them are still unappropriated by any element;thirty-one (twenty-four of even mass) are isobares of which three(two of even mass) are shared by a third element.These figuresdo not include the isobares 23 and 39, because the existence ofNe23 and CP9, although they still receive support on optical evi-dence,l* is doubtful on mass-spectrographic evidence ; theyinclude 203 which is a possibility for lead as well as a certainty forthallium and mercury.Exact knowledge of the masses of the isotopes of the light elementsis becoming increasingly important in experimental work onartificial disintegration and for theories of nuclear structure. InTable I1 are given the isotopic masses determined (or redeterminedwith increased accuracy) since the last list was given. (0l6 =16.0000, He4 = 4.00216.)TABLE 11.Atom. Mass. Atom. Mass. Atom. Mass. Atom. Mass.H1 1.007775 0 1 7 17.0029 77.937 Ba138 137-916K2 or D 2.01363 0ls 18.0065 SesO 79-941 TalS1 180.927Li6 6.0145 NeZ0 19.9967 Nbs3 92.926 W18* 184.00Li7 7.0146 Ne22 21.9947 Moss 97.945 RelS7 186.981Bes 9.0155 SiZ8 27.9818 MoloO 99,945 189.981J311 11.0107 Cr52 51.948 TelZe 125.937 Oslgz 191.981C13 13.0039 Ni5* 57.942 TelZ8 127.936 T1203 203-036N15 15.007 Zn64 63.937 Cs133 132.933 TIzo5 205.037H2, now called D, was determined originally by K. T.Bainbridge 2Oas 2.01351 and by J. D. Hardy, E. F. Barker, and D. M. Dennison 2Oas 2.01367; more recently21 the first has redetermined it as2.01363 5 0.00004 with reference to He4 = 4.00216 and 2.01363 &0-00008 with reference to 0l6 = 16.0000. The determinations of16 Ann. Reports, 1928, 25, 314.18 H.Kallman and W. Lamrev, 2. Physik, 1933, 80, 237; A., 333; F. W.Aston, “ Mass Spectra and Isotopes,” London, 1933; M. Ashley and F. A.Jenkins, Physical Rev., 1931, [ii], 37, 712; 1932, 42, 438; A., 204, 1223;Ann. Reports, 1932, 29, 303.19 Ibid., 1928, 25, 304.21 ph.ysica1 Rev., 1933, [ii], 43, 103; 44, 57; A., 203.l7 Ibid., 1930, 27, 316.ao Ibid., 1932, 29, 302348 SUB-ATOMIC PHENOMENA AND RADIOACTIVITY.H1 as 1.007775 and of Lie, Li7, BeQ, Bl1, Ne2Q, and Ne22 were alsodone by him.22 A. McKellar 23 has done Lie and Li7. A. S. Kingand R. T. BirgeS did C13, R. T. Birge25 N15 and 017, and H. D.Babcock and R. T. BirgeZ6 O1*. The remainder have been doneby F. W. A ~ t o n . ~ ~The abundances of the rarer isotopes of the light elements havegenerally been under-estimated in preliminary work ; most of theredeterminations of the abundance ratios of common to rarerisotope have given smaller values. The change in the ratio H : Dfrom 35,000 to 5,000 has already been discussed.28 K.T. Bain-bridge 29 concluded that 11.3 is the most probable value for Li7 : Li6.L. S. Ornstein and J. A. Vreeswijk, j ~ n . , ~ O found 4-43 from bandspectrum work for Bll : BIO ; earlier values ranged from 3.63 to4.86. The ratio C12 : C13 formerly given 31 as 650 and 400 is nowfound to be as small as 106 by F. A. Jenkins and L. S. Orhstein32from band spectrum work on C12C13 and C12C12. (The calculatedatomic weight of carbon on this ratio is 12-01, a value greater thanthe International value but the same as the latest determination,12-011, of M.Woodhead and R. Wh~tlaw-Gray.~~) For N14 : W5,R. T. Birge and I>. H. Menze134 found 320, H. M. Urey and G. M.Murphy 35 347 ; these are about half the original estimates. Therehas been no change l8, 34 in the most important of these ratios,that of 0l6 : O1*. For gallium the ratio Ga69 : Ga71 calculatedfrom optical evidence36 is 1.5, a result which leads to an atomicweight in excellent accord with the experimental value.In the main, atomic weights 37 calculated from mass-spectro-graphic data are now in good accord with those determined directly.Where there have been differences the values got by the formermethod have generally prevailed, and that despite the three-fold2a Physical Rev., 1933, [ii], 43, 424 ; 44, 56 ; ,4 ., 442.23 Ibid., 44, 155; A., 994.24 Astrophys.J., 1930, 72, 19; A., 1931, 15.25 Physical Rev., 1931, [ii], 37, S41; A., 204.26 Ibid., 233 ; A., 203.27 Ann. Reports, 1930, 27, 305; 1932, 29, 301.topes,” London, 1933.28 This vol., p. 30.2s J . Pranklin Inst., 1931, 212, 317; A., 1931, 1201.3o 2. Physik, 1933, 80, 57; A., 204.31 Ann. Reports, 1930, 2’7, 306.32 Proc. K. Akad. Wetensch. Amsterdam, 1932, 35, 1212 ; A., 333.33 J . , 1933, 846; A., 894.34 Physical Rev., 1931, [ii], 37, 1669; A., 204.38 Ibid., 38, 575.3G J. S. Campbell, Nature, 1933, 131, 204; A., 334.37 Ber., 1933, 66, [A], 21; A., 203.“ Mass Spectra and IsoRUSSELL. 349disadvantage that its accuracy is not greater than 1 in lo5, thatthe conversion factor from the scale OI6 = 16.0000 to the scale0 = 16.0000 introduces an error as large, and that there are inherentuncertainties in the degree of precision in the measurement ofabundance ratios.The principal battleground between physicaland chemical methods of determining atomic weights during theperiod has been the Katanga deposits of pitchblende, the work onwhich i s discussed in the section on Atomic Weights.38 The dis-crepancies found by such careful and experienced workers showclearly how unwise it is to deduce the abundances of different leadisotopes from chemical determinations of atomic weights, and thisprobably applies generally. Recent work on the atomic weightof potassium 39 supports this point.Many results have been obtained by F.Allison and the Alabamaschool of workers with their magneto-optic method of deter-mining isotopes, but it is not yet time to discuss them fully here.The method itself consists in observing the Faraday magneto-optical rotation produced by solutions of metallic salts, and measur-ing the lag between the application of the field and the appearanceof the optical effect. Even with very dilute solutions it is claimedthat the presence of a particular isotope makes itself evident bythe occurrence of a particular lag. The existence of these lagshas been confirmed by W. M. Latimer and H. A. Young,41 whohad no difficulty in finding a new isotope H3 along with D and H.Others, however, have thrown doubt even on the very existenceof the time-lags; and there is no doubt of the wrongness of theirinterpretation.In the same laboratory G. N. Lewis and P. H.Spedding41 had found no optical evidence of H3 in a hydrogen-diplogen mixture; if it exists a t all, it is present a t a concentrationbelow 1 in G x lo6. Bismuth, whose single isotope, Bi209, is inaccord with work on radioactivity, on chemical atomic weights,and on mass-spectrography, is reported *O as having, in decreasingorder of abundances, the isotopes 211, 210, 209, 212, 215, 214, 213,216, 207, 205, 206, 208, 219 and 217, and thallium (atomic weight204.4; T P 5 and T1203 in the ratio of 12 : 5) the masses 207, 205,211, 203, 201, 209, 215, 213. As a matter of arithmetic theseresults lead to absurd values of atomic weights.Less pretentiousbut hardly more plausible lists have been given for cobalt, lead,elements 85 and 87, thorium, uranium, and other metals. Their38 See this vol., p. 84. 39 See this vol., p. 85.40 Physical Rev., 1931, [ii], 37, 1178; 1932, 40, 1015; 1933, 43, 38, 43, 47,48, 49; A., 204, 241, 1223; J . Amer. Chem. SOC., 1933, 55, 3207, 4370; A.,994; 1934,6.4 1 Physical Rev., 1933, [ii], 44, 690; 43, 964; A., 759360 SUB-ATOMIC PHENOMENA AND RADIOACTIVITY.existence is the best evidence of the untrustworthiness of thisinterpretation of the time-lags.Artijicial Disintegration by Xwif t Protons, Diplons, Neutrons,and u-Particles.Four projectiles essentially different are now available for reveal-ing the structure of the nuclei of the lighter elements by the effectsthey produce on bombardment. J.D. Cockcroft and E. T. S.Walton42 have continued their work43 on the bombardment oflithium, boron, and other nuclei with protons accelerated a t highvoltages. Lithium on disintegration yields a group of particlesof range less than 2 cm. and a homogeneous group of range 8.4 cm.The process corresponding with the latter is believed to beH + Li7 -+ 2He4. From boron one a-particle per 2 x lo6incident protons at 500,000 volts was obtained. The energiesfound support the view that a proton is captured by the Bll nucleus,which then splits into three a-particles : H + Bl1+ 3He4.E. 0. Lawrence, M. S. Livingston, M. C. Henderson, and M. G.White44 confirmed both the lithium and the boron results andfound some evidence in addition of the disintegration of aluminiumby swift protons.M. L. E. Oliphant and (Lord) Rutherford45have extended this work. Lithium and boron in very thin filmswere bombarded by protons accelerated a t quite low voltages(20,000-200,000 volts). The former gave off a-particles whenstruck by 30,000-volt protons; the latter did not do so until theprotons were at 60,000 volts. Beryllium and fluorine were similarlydisintegrated but to a markedly smaller extent. No disintegrationwas observed with targets of nitrogen, oxygen, sodium, aluminium,iron, gold, thallium, lead, bismuth, thorium, or uranium withprotons accelerated up to 200,000 volts ; earlier observations 48 onaluminium, uranium, and other elements are ascribed to contamin-ation by boron from the glass of the apparatus.The disintegrationof Bll is confirmed as consisting in the absorption of the proton,followed by a split into three u-particles. The most probable modeof disintegration is a symmetrical escape of the particles withequal velocities. The energy which becomes available in thischange (9 x lo6 volts) is similar to that calculated from the massesof the particles.The diplon has now become an important projectile. E. 0.Lawrence, D. H. Sloan, and M. S. Livingston46 had developed an42 Nature, 1933, 131, 23; A., 111.44 P h y s k l Rev., 1932, [ii], 42, 160; 1933, 43, 98, 304, 369; A., 1932, 654;45 Proc. Roy. SOC., 1933, [ A ] , 141, 259; A., 883.4 6 Physical Rev., 1931, [ii], 38, 2021; 1932, 40, 19; A., 1932, 106.654.43 Ann. Reports, 1932, 29, 307.1933, 204,1225RUSSELL. 351apparatus for obtaining high-speed ions corresponding with morethan a million volts in energy. With G. N. Lewis and M. C .Henderson 47 they found that diplons were much more effective thanprotons of equal energy in causing transformations in a numberof elements. From lithium bombarded by diplons, a-particles wereejected with a greater velocity than any yet observed from radio-active elements. The diplon is captured by the Li6 nucleus andthe resultant breaks into two cc-particles, which escape in nearlyopposite directions : D + Li6 -+ 2He4. They bombarded beryl-lium and nitrogen (in ammonium nitrate) and obtained or-particles,but the processes thought to have occurred have not yet beenformulated.There were similar rare disintegrations with mag-nesium and aluminium. They noticed that protons of 18-cm.range were produced when the molecular diplon ions struck anytarget whatever; they suggest that the diplon may break up intoa proton and a neutron : D -+ H + n, either in the bombardednucleus or in the strong field in its neighbourhood. If this occurs,however, and the conservation of energy holds, the mass of theneutron should be less than J. Chadwick’s estimate 48 : 1.0006instead of 1.0067. Expressed in another way, the differencebetween the masses of D, 2.0136, and of n + H, 2.0145, correspondswith a binding energy less than lo6 volts. If this were all, thediplon would very probably be broken by collision with swift=-particles ; preliminary experiments on this effect, however, donot support this s~ggestion.~~ They also found that neutronswere produced by the disintegration of Be9 either by capturing adiplon, D + Be9 -+ n + B1O, or by merely disintegrating theberyllium nucleus.The latter was the more likely, as the prob-ability of the disintegration of Be9 was found to be independentof the energy of the diplon. This process furnishes an excellentsupply of neutrons for purposes of investigation. M. L. E. Oliphant,B. B. Kinsey, and (Lord) Rutherford50 with their technique45have confirmed and extended some of these results. The a-particlesejected from lithium by diplons fall into two groups : one of 13.0cm. range and 11.5 x lo6 volts energy and another which containsparticles of energies from 1.7 to 8.3 x lo6 volts. The formerprobably arises from the transformation D + Li6 + 2He4, thelatter from D + Li7 --+ 2He4 + n.The maximum energy ofthe neutron may be as high as 15 x lo6 volts. They examined alsothe short-range particles emitted when protons strike lithium ; 42these may include two groups of ranges 6.5 mm. and 11.5 mm.48 Ann. Reporta, 1932,29,306. 47 Phyaical Rev., 1933, [ii], 44,55,56,781,782.49 (Lord) Rutherford, Nature, 1933, 132, 965.60 PTOC. Roy. Soc., 1933, [A], 141, 722; A., 1100352 SUB- ATOMIC PHENOMENA AND RADIOACTIVITY.P. I. Dee and E. T. S. W a l t ~ n , ~ ~ using camera and expansionchamber, have confirmed the processes, H + Li7 + 2He4 andD + Li6+ 2He4, with a, different technique and with muchhigher potentials (up to 400,000 volts), and discussed the processes.The proton emission from lithium on bombardment with diplonshas been extended t o carbon and iron by J.D. Cockcroft andE. T. S. Walton ; 42, 49 their diplons were accelerated to about 500,000volts; they did not observe proton emission from copper or gold.The neutron emission from lithium bombarded by diplons hasbeen confimed by H. R. Crane, C. C. Lauritsen, and A. S01tan.~~They found also that beryllium gave neutrons readily when thediplons were accelerated up t o 900,000 volts; the process isformulated as D + Be9 --+ n + BIO. The kinetic energy of theissuing neutrons was about 9 x lo6 volts; the yield was severalhundreds of times greater than that obtained by the methodwith which they were disco~ered.~~Relative to the diplon the neutron has been neglected as a pro-jectile.w. D. Harkins, D. M. Gans, and H. w. Newson 54 foundevidence in Wilson photographs of the disintegration of nitrogenby neutrons from beryllium struck by rays from a mesothoriumpreparation. The neutron was captured ; the resulting nucleusthen split according to n + N14 -+ N15 + He4 + Bll. Pluorineand neon were also disintegrated; the former process is possiblyn + 4 He4 + NI6, the latter n + Ne20 4 He4 + O17.These results confirm and extend those of N. Feather48 with asimilar technique. He 55 has examined the tracks of neutrons inoxygen, and in oxygen-hydrogen and acetylene-helium mixtures.Most of the neutrons (from a polonium beryllium source) have anupper energy limit of about 4.5 x 106 volts. Disintegrations wereobserved in oxygen, confirming the original work : n + 0 1 6 --+He4 + CIS ; the energy relations required in addition the productionof a y-ray of high energy.The disintegration of carbon by neutronsis very rare, if existent.The conditions of emission of neutrons by the bombardment oflight atoms with a-particles from polonium have been determinedby' (Mme.) I. Curie and F. J01iot.~~ Beryllium and boron yield51 Proc. Roy. SOC., 1933, [A], 141, 733; A., 1100; Nature, 1933, 132, 818;A., 1934,6.62 Physical Rev., 1933, [ii], 44, 692, 783; Compt. rend., 1933, 197, 639, 913;A., 1225; 1934,6.53 Ann. Reports, 1932, 29, 305.64 Physical Rev., 1933, [ii], 43, 203, 236; 44, 529, 945; A., 1225.55 Proc.Roy. SOC., 1933, [A], 142, 689; A., 1934, 6.5 6 Compt. rend., 1933, 196, 397; A., 335; J. Phys. Radium, 1933, [vii], 4,21, 278; A., 335, 883RUSSELL. 353neutrons most readily (cordinning the original work of J. Chad-~ i c k ) , ~ 8 lithium comes next, then fluorine, sodium, and aluminium.With beryllium and boron a minimum energy of the a-particlewas required to excite the target; as the energy increased, theexcitation first rose to a maximum and then dropped to a minimum.The energies at maximum excitation were 26 x lo6 volts forberyllium and 2.2 x lo6 volts for boron. With lithium y-rayswere emitted when the a-particle was at 3 x lo6 volts, and neutronswhen it was at 5 x lo6 volts.It appears 57 that a-particles ofcomparatively low velocities penetrate the nucleus by a resonanceprocess and that swift a-particles enter over the top of its potentialbarrier. No penetrating radiation was found 56 when hydrogen,carbon, nitrogen, oxygen, phosphorus, and calcium were bombardedwith a-particles. G. who had earlier 59 found protonemission from a-particle bombardment, where others had failed toobserve it, found small neutron emission in the bombardment bya-particles from polonium of carbon, nitrogen, magnesium, alu-minium, sulphur, iron, nickel, copper, zinc, molybdenum, palladium,arsenic, tin, tungsten, platinum, gold, and lead ; lithium, beryllium,and boron were found to give strong emission.He thus confirmsthe work of (Mme.) I. Curie and F. Joliot on aluminium (butapparently not on fluorine and sodium) and differs from them onnitrogen and oxygen. P. Auger and G. M. Herzen,60 using theWilson method, have also confirmed the emission of neutrons fromaluminium; this is very small compared with that from berylliumunder similar conditions. If nitrogen and oxygen be confirmedas emitting neutrons, this property must be assigned to N15 and017 and Ol8, not t o the common atoms, for in all known nuclei theatomic mass is never less than twice the atomic number, and thiscondition would be violated by the nuclei formed from N14, 0 1 6(and He4) if these atoms emitted a neutron under a-particle born-bardment. For some elements y-rays accompany neutron emission ;with others protons accompany neutrons.The disintegration ofberyllium, for example, may lead to a swift neutron or to a slowerneutron accompanied by a y-ray of high energy. From fluorineor aluminium both protons and neutrons may be emitted; asthese elements are isotopically simple, the emission processes arealternative.(Frl.) L. Meitner and K. Philipp 61 have studied the connexion67 J. Chadwick, Proc. Roy. SOC., 1933, [A], 142, 1; A., 1224.5 8 Naturwiss., 1933, 21, 332; A., 659.59 Ann. Reports, 1926, 23, 285.60 Cornpt. rend., 1933, 186, 643; A., 335.61 Naturwias., 1932, 20, 929; A., 110.REP.-VOL. XXX. 354 SUB-ATOMIC PHENOMENA AND RADIOACTIVITY.between neutrons and the accompanying y-rays by the Wilsonmethod. They think it probable that neutrons and y-rays areemitted simultaneously. They have emphasised the fact that somenuclear processes can be carried out in both directions; e.g., He4 +Bll s n + N14.When the neutron was discovered it was not regarded as a funda-mental particle because it could be represented as n = H + e.The existence of the positive electron, however, suggests that nor H is fundamental since now €I itself can be represented asrt, + e+.(Mme.) I. Curie and F. Joliot 62 have found that positiveclectroiis may result from the bombardment of aluminium andboron (though not of lithium, carbon, or silver) by a-particles, aswell as protons and neutrons. If these emissions are simultaneous,there is a possibility that a proton has broken up into a neutronand a positive electron, Le., H = n -+ e t .On this view the trans-formation of BIO would be He4 + BIO = C13 + n1 + eT or = C13 +H simultaneously. Taking account of the energies involved, thisleads to a mass of the neutron 1.011, which is greater than J. Chad-wick's estimate; there are other difficulties also in getting massesand energies to balance. This question, which is part of the com-plex question of the structure of light nuclei, has been discussed byJ. Chadwick,57 W. Hei~enberg,~~ and many others, and has beenextended to the diplon, which may be a union of proton and neutrona s suggested by the above experiments describing its break up.47This possibility has been strengthened by different evidence.D. E. LeaO4 studied the scattering of neutrons by hydrogen andother gases and found a y-ray effect in hydrogen only which mightbe interpreted as resulting from the combination of neutron andproton to form diplon in some of the collisions.The differentschemes suggested 65 may be represented as H = n -b e+,n = H + e , n = H + d ; D = 2 H + e , D = 2 n + e + , D = H + n(H = proton, D = diplon, e = electron, eL = positive electron,n = neut;fon, d = Dirac's electron of negative energy). It isobvious that no decision on these points can be formed in the62 Compt. rend., 1933,196, 1885; 197, 237; A., 762, 883.63 2. PhysiE, 1932, 77, 1 ; 78, 156; A., 1932, 894, 1074.O 5 W. D. Harkins, J . Amer. Chem. SOC., 1933,55,855; A., 334; D. Meksyn,Nature, 1933,131, 366; W.Elsasser, ibid., p. 764; A , , 658; N. Thon, ibid.,p. 878; A , , 762; G. W. Todd, ibid., 132, 65; A., 883; T. Sexl, ibid., p. 174";A., 995; R. M. Langer, Science, 1932,76, 294; A., 1932, 1187; E. N. Gapon,%. Phpik, 1933, 84, 509, 520; A., 995; R. Fiirth, ibid., 85, 294; A., 1100;J. J. Placinteanu, ibid., 84, 370; A., 995; Compt. rend., 1933, 196, 1474;197, 549; A., 680, 995; 13. R. von Traubenberg, Naturwiss., 1932, 20, 934;z4., 111; J. Perrin, Cornpt. rend., 1933, 197, 628; A., 1224.Nature, 1934, 133, 24RUSSELL. 355present tentative position both of experimental work and of theory.New particles may still be discovered (e.g., a neutron 66 of mass 2)whose existence may profoundly modify present speculative theory ;the fact that difficulties await solution is a powerful incentive toobtain great accuracy in the determination of the masses andenergies involved in the transformation of light atoms by bombard-ment with the various projectiles.In Table I11 are summarised the principal disintegration pro-cesses which occur with light elements.These include (for theTABLE 111.Process. Reference. Process.H+Li7 ---f2He4 42, 43, 44, He4+B11 -+n+Nl445, 50, 51 He4+N14 +H+Ol7H+BeD ---)He4+? 45 He4+FlD + H+NeZ2H+Bl1 -3He4 42, 44, 45 He4+FlD +n+ 1H+FID +He4+? 45 He4+Ne +H+NaD+Li6 ---+2He4 47, 50, 51 He4+NaZ3 +H+MgZ6D+Li7 +n+2He4 50, 52 He4+NaZ3 -+ n+ ?D+BeD ---)n+B1° 47,52 He4+Mg +H+AlD+BeD +He4+? 47 He4+A127 + H+Si30D+N14 +He4+? 47 He4+Ala7 -+n+ ?n+N14 +He4+Bl1 52, 59, 65 He4+Ala7 +e++ ?n+016 + ~ e 4 + ~ 1 3 52 He4+Si +H+Pn+Fla ---+He4+? 54 He4fP31 --+- H+S34n+Ne20 -+He4+O1' 54 He4+S +H+ClHe4+L17 --+TI+B~~ 56, 68 He4+C1 +H+AHe4+BeD +n+C12 48, 66, 68 He4+A +H+KHe4+B1° +H+C13 67 H&+K +H+CaHe4+B10 +n+e++C13 62Reference.48, 56, 68676766, 57676766, 57676756, 68, 6062676767676767sake of comparison and completeness) old results 67 for protonemission from light elements bombarded by a-particles.Wheremasses are unknown or obvious they are omitted. In each schemethe symbol for the projectile precedes that of the target; thesymbol of the more certainly recognised disintegration productfollows the arrow.The Positive Electron.The positive electron was discovered independently by C.D.Anderson 68 and by P. M. S. Blackett and G. P. S. Occhialini 69 inan investigation of the penetrating or cosmic radiation by the cloudrnethod.'O Amongst the tracks of the particles of very greatenergy included in or produced by this radiation were found somewhich differed mainly from the tracks of electrons in being curved66 H. Walke, Nature, 1933, 132, 242; A., 995.13' (Lord) Rutherford, J. Chadwick, and C. D. Ellis, "Radiations fromRadioactive Substances," Cambridge, 1930.68 Science, 1932, 76, 238.60 Proc. Roy. SOC., 1933, [A], 139, 699; A., 441.'O Ann. Reports, 1932, 29, 314356 SUB-ATOMIC PHENOMENA AND RADIOACTIVITY.by a magnetic field in the opposite direction. Anderson 71 found15 out of 1300 cosmic-ray tracks whose curvature and ionisationsuggested a positive particle much smaller than a, proton, in fact,of mags not more than twenty times that of an electron.Later heconcluded that positive and negative electrons were within 10%of each other in charge and 20y0 in mass. Blackett and Occhialinifound substantially the same. An interesting feature of thenumerous photographs taken was the existence of “ showers ” ormultiple pairs of particles of opposite sign due apparently to somenuclear disintegration process stimulated by particles of highenergy in the cosmic radiation. The discovery of terrestrialsources of positive electrons which followed quickly on the otherhas led to their investigation under more controlled experimentalconditions. Positive electrons are produced by the absorption bymatter (e.g., lead) of the penetrating y-radiation 72 from radio-elements (e.g., from thorium-C”) or of the penetrating radiation 73(in part neutrons, in part y-radiation) produced by the impact ofa-particles on beryllium.They are also produced 62 in what may bea nuclear transformation by the impact of a-particles on aluminiumand boron. When quanta produce positive electrons on theirabsorption in matter, preliminary work 74 shows that the numberincreases rapidly with the energy of the quanta and with the atomicnumber of the absorber. Prom this it has been deduced thatthe effective area for the production of a positive electron i s alittle larger than the area of cross-section of the nucleus. Themaximum energy of the positive electron produced by a givenradiation seems to be the same for all absorbers, and this, also,makes it unlikely that the production of positive electrons is a,nuclear process.It is obvious that if quanta alone produce positive electrons,and do so outside the nucleus, the discussion is simplified.(Mme.)I. Curie and F. Joliot 73 have shown that it is the y-radiation in themixed radiation produced by a-particles on beryllium which pro-duces the positive electrons; on the other hand, in their nuclear-71 Physical Rev., 1933, [ii], 43, 491; 44, 406; A . , 441, 1100.72 C. D. Anderson, Science, 1933,77,432 ; C. D. Anderson and S. E. Nedder-meyer, Physical Rev., 1933, [ii], 43, 1034; (Mme.) I. Curie and F. Joliot,Compt. rend., 1933, 196, 1581; A., 658; (Frl.) L.Meitner and K. Philipp,Nuturwiss., 1933, 21,468; A., 762; C. Y. Chao and T. T. Kung, Nature, 1933,132, 709 ; (Lord) Rutherford, ibid., p. 709 ; A., 1934, 5.73 Compt. rend., 1933, 196, 1105; A., 649; see also 3. Chadwick, P. M. S.Blackett, and G . P. S. Occhialini, Nature, 1933, 131, 473 ; A., 441 ; (Frl.) I,.Meitner and K. Philipp, Naturu-bs., 1933, 21, 286; A., 560.74 Nature, 1933, 132, 917; M. Grinberg, Compt. rend., 1933, 197, 318; A.,883RUSSELL. 357transformation experiments 62 they suggest that positive electronsoriginate in the nucleus. P. M. S. B l a ~ k e t t , ~ ~ however, has inter-preted them otherwise; the positive electron may be producedoutside the nucleus by the internal conversion of a y-ray emittedby it.If the positive electron is produced outside the nucleus,it follows that, as there is no place in atomic theory for its permanentexistence, a positive electron cannot arise without the simultaneousproduction of a (negative) electron. This conclusion is in accordboth with experimental observation of the pairs of tracks and withdeterminations of the maximum energies of the positive and negativeelectrons in their tracks. The occurrence of the reverse process,namely, the production of a single quantity of radiation by apositive and a negative electron, is a necessary consequence of theprocess just stated. The theory in this tentative form is akin inone way to the “ neutral pair” theory of the y-ray of (Sir) W. H.Bragg,75 which was abandoned later for the wave-theory.Heimagined an electron as capable of attracting t o itself sufficientpositive electricity to neutralise its own charge and of doing thiswithout appreciable addition of mass. (This is the transformationp-particle -+ y-ray.) The reverse process was believed to occurwhen the electron put down its positive charge again. Neitherchange was thought to take place except when an entity traversedan atom. Blackett and Occhialini interpret their work on P. A. M.Dirac’s theory 76 of “ holes.” Dirac deduced from the wave-equation obtained from the relativity quantum theory of an electronthat states are possible in which electrons should possess negativekinetic energies. He dealt satisfactorily with the paradoxes whichthis view entails.He assumed that there are so many electrons inthe world that all the states of negative energy except perhaps afew are occupied, and supposed that the infinite number of electronspresent in any volume will remain undetectable if they are uniformlydistributed ; only an unoccupied state or ‘( hole ” would correspondwith an observed particle. These should behave in an externalfield like particles with the same mass and spin as an electron butwith a positive charge. (These were a t the time identified withprotons, but the difficulty of explaining why protons differed sowidely in mass from electrons was recognised.) The experimentaldiscovery of the positive electron has thus greatly strengthenedDirac’s theory and extended the field of phenomena over whichit may be applied.J. R. Oppenheimer and M. S. Plesset 77 havecalculated from it the probability of production of pairs of electronsof opposite charge when penetrating y-rays are absorbed by matter.75 Ann. Reports, 1909, 6, 242.7 1 Physical Rev., 1933, [ii], 44, 53, 948.‘13 Ibid., 1030, 27, 325358 SUB-ATOMIC PHENOMENA AND -RADIOACTIVITY.The calculated are in good agreement with the experimental results.They regard the absorption of y-rays by atoms resulting in theproduction of pairs of oppositely charged electrons as a photoelectricabsorption by electrons with negative kinetic energy-" virtual "electrons-near the nucleus. G. Beck 78 has calculated the numberof these which are effective for absorption; it is proportional tothe square of the atomic number and is small; lead has about one.He has indicated also that the place of production is outside thenucleus but within the I<-ring. Many experiments have been doneon this subject and it has been widely discussed.79 Dirac's theoryhas not yet explained the production of the multiple pairs whichform the showers.The relation of positive electrons to the cosmic radiation is stillimperfectly known.Certain facts, however, have been establishedby Anderson,so Rlackett and Oc~hialini,~~ P. Kunze,Bl andT. H. Johnson : g2 half the fast particles which produce cosmicray ionisation at sea-level are positive electrons, half are electrons ;the same ratio occurs in the " showers " ; the majority of theparticles incident on the earth's atmosphere are positively charged,probably posibivc electrons, not protons.The Penetrating Radiation.Investigations of widely-diff erent character-cloud-chamberexperiments,74 absorption of the penetrating rays with the Bothe-Kolliorster double counter, measurenients of the relative intensityof the radiation a t different parts of the earth,83 and determinationof its intensity at different heights in the atmosphere and depthsin water "-all now support t'he view that the penetrating radi-ation a t high altitudes consists of electrified particles producedin the atmosphere or entering it from remote space. (At lower.altitudes a mixture of particles and photons is likely to be presentin the radiation, the latter being produced by the former by pro-cesses similar to the production of X-rays by an electron.) The2.Physik, 1933, 83, 498; A . , 884.79 W. H. Furry and J. F. Carlson, Physical Rev., 1933, 44, 237; (Mme.)I. Curis and F. Joliot, J . Plzys. Radium, 1933, [vii], 4, 494; A., 1224; D.Skobeltzyn, Nature, 1934, 133, 23.Physical Rev., 1932, [ii], 41, 405.2. Physik, 1933, 80, 559 ; 83, 1 ; A., 763.Physical Rev., 1933, [ii], 43, 1059.83 J. Clay, Proc. K. Akad. Wetensch. Amsterdam, 1932, 35, 1282; A., 335;A. H. Compton, Physical Rev., 1933, [ii], 43, 28; Nature, 1933, 131, 713;Science, 1933, TI, 480 ; A., 660.84 E. Regener, ibid., 1932, 130, 364; 1933, 132, 696; Physikal. Z., 1933,34, 306 ; A., 552RUSSELL. 359supposed invariance 85 of the distribution of the radiation withlatitude, implying that the earth's magnetic field does not affectthe radiation, was the one strong piece of evidence against theparticle theory.Actually, however, there is variation.83 Attemperate and polar latitudes (from 50" N. and S. to the poles) theradiation is nearly constant and more intense than in tropicallatitudes; the differences are about 14, 22, and 33% a t sea-level,2000, and 4000 metres respectively. The extensive experimentswhich have been carried out a t high altitudes by sounding andpassenger balloons, and on high mountains, support the particletheory also. The data show a rapid increase in intensity of theradiation with altitude ; it continues nearly exponentially to aheight of 15 km. and above thatl approaches a limiting value;there is apparently no detectable decrease in ionisation as the topof the atmosphere is approached. Quite different results are tobe expected if the radiation consist appreciably of photons. Theionisation is constant in time to within 2% at any one place. Ashas been said, the majority of the particles high in the earth'satmosphere are positively charged and are probably positiveelectrons. At sea-levelhalf the particles are electrons, half positive electrons ; theirenergies vary from a few million t o nearly 1O1O volts. They areaccompanied by photons which are probably the precursors of thesingle pair and multiple pairs of oppositely charged electronsobserved in the cloud-chamber apparatus.If it comesfrom outside and is mainly positive electrons, it can be calculated 86that the total mass of the radiation in the universe is staggeringlyenormous-about O.lyo of the mass of all the stars and nebulaOn the other hand, the only terrestrial source that has beensuggested 87 is thunderstorms, and there the main difficulties havebeen the constancy of the intensity with time and its very highenergy. The last is evidence that it does not arise from a nuclearprocess. The constancy with time rules out the sun, and pre-sumably therefore the stars, as a source. There is, further, nothingin the cold of empty space likely to produce a radiation which hasfailed to appear in the greater cold of the laboratory. At thepresent time the evidence for the radiation points to an unknownterrestrial source. A. S. RUSSELL.Their energy is of the order of 1O1O volts.The origin of the radiation remains quite obscure.8 5 Ann. Reports, 1932, 29, 315.87 C. T. R. Wilson, PTOC. Can& Phil. SOC., 1926, 2!2, 534.G. LemeEtre, Xature, 1931, 128, 704

ISSN:0365-6217    

DOI:10.1039/AR9333000344

出版商:RSC    

年代:1933

数据来源: RSC

摘要:

CRYSTALLOGRAPHY.CRYSTALLOGRAPHIC research in the last two years has no longerbeen primarily concerned with structure determination. The maintypes of structure have already been determined, and the morecomplicated structures are found ip general to be variations onsimpler types or combinations between them. The prevailinginterest has shifted to the relations between crystal structure andphysical properties, and to relations and transformations betweencrystal types. Crystal physics more firmly grounded on quantummechanics is widening its scope and becoming increasingly quan-titative and precise, whilst crystal chemistry is appearing moreand more as the basis on which the chemistry-of the future willbe written. The rational chemistry of metals and ores is a creationof structural crystallographic methods ; the chemistry of celluloseand other complex polymerides such as the proteins has throughthem made a critical advance.The appearance of the first volume of “The Crystalline State,”a general survey by W.L. Bragg, has at last provided anauthoritative, attractive, and simple presentation of the con-tribution of structure analysis to the problems of physics, chemistry,biology, and technics. This volume is not intended for specialistsin the field, but should be in the hands of every chemist who wishesto do original work on modern lines.Ewald’s new edition of “ Die Erforschung des Aufbaues derMaterie mit Rontgenstralilen ” furnishes a more theoretical approachto the problem of crystal analysis : it is particularly valuable forits account of X-ray optics.The Properties of Cuprous Oxide.-The commercial use of rectifiersand photocells based on the remarkable properties of semi-conduct-ing crystalline films of cuprous oxide (cuprite) has drawn muchattention to the electrical and optical behaviour of this substance.An excellent review of this field until the end of 1932, with copiousreferences, is given by L.0. Grondah1,l which leads us to concludethat the theoretical side of the subject is not as far advanced as thepractical.Cuprous oxide, as a large single crystal, has a slight electricalconductivity which appears to depend on the concentration of freeRev. Mod. Physics, 1933, 5, 141ROBINSON AND WOOSTER. 361oxygen which it contains.2 When the crystal is illuminated, a,photoconductivity appears (inner photoelectric e f f e ~ t ) , ~ photo-electrons are ejected from the surface (outer photoelectric effect)and a potential difference may be set up between different parts ofthe crystal, as if electrons had been carried in the direction of thelight rays (Dember effect).Photoconductivity is, of course, ausual phenomenon with non-conducting crystals (for a generalreview of the subject see F. C. Nix 5, ; the outer photoelectric effectis well known in metals, but little work seems to have been done onit for less good conductors.6 The measurement of the photocon-ductivity is not easy with cuprite, since the normal conductivitymay mask the effect ; recourse has been made to the use of periodicillumination, allowing the periodic current to be separately ampli-fied; and to measurements at low temperatures,s at which thenormal conductivity is small.The technique of R. Schultze withvery thin films would also seem to be applicable.The theoretical pictures of the passage of electrons through feebly -conducting substances have been developed by R. H. Powler,loA. H. Wilson,ll R. W. Gurney l2 and others on the basis of quantumstatistics; and these pictures seem to be accepted as adequate 13for the explanation of the inner photoelectric effect with insulatingcrystals. But Wilson’s theory is dismissed by (Mme.) A. Joff6 andA. F. Joff6 l4 atEl unable to explain for cuprite either the potentialdistribution in the illuminated crystal or the variation of thispotential with intensity of light.It has in fact been questionedwhether, apart from the presence of impurities, semiconductors likecuprite have any real existence as such.There seems to be some difference of opinion as to what spectralH. Dunwald and C. Wagner, Z . physikal. Chem., 1933, [B], 29, 212; A.,887; E. Engelhard, Ann. Physik, 1933, [v], 17, 501; A., 1000; G. MKnoh,2. Phyaik, 1932, 78, 728; A., 8; M. Leblanc, H. Sachse, and H. Schopel,Ann. Physik, 1933, 17, 334; A., 764.Pfund, Physical Rev., 1916, 7 , 289; but cf. F. Waibel and W. Schottky,Physikal. Z . , 1932, 33, 683; A., 1932, 898.4 H. Dembor, ibid., 1931, 32, 554, 856; 1932, 33, 207; A., 1931, 999;1932, 8.5 Rev. Mod, Physics, 1932, 4, 723.ti References in Schultze’s papor (rof.9) to work on phosphorus, aniline7 Schonwald, Ann. Physik, 1932,15, 395.8 I. Kikoin and M. Noskov, Physikal. Z. Soviet Union, 1933, 4, 531.9 Physikal. Z., 1933, 34, 381.10 R. H. Fowler, Proc. Roy. Soc., 1933, [ A ] , 141, 56; A., 887.11 A. H. Wilson, Nature, 1932, 130, 913; A., 114.12 R. W. Gurney, Proc. Roy. Soc., 1933, [A!, 141, 209; A., 887.13 H. Teichmann, ibid., 139, 105; A., 209.1 4 Z. Physik, 1933, 82, 754; A., 764.colours, sulphur, shellac, paraffin, water, silver halides.M 362 CRI'STALLOORAPHY.range is responsible for the photoconductivity of cuprite : thusF,. Englehard postulates the excitation of an infra-red absorptionband, while D. Nasledov and L. Nemenov l5 even ascribe the increaseof conductivity on illumination of ordinary cuprite to a heatingeffect.Schonwald,' on the other hand, finds definite photoelectricmaxima at wave-lengths of 0.8, 0.63, and 0.2 8. The simultaneousirradiation of a cuprite photocell with different wave-lengths gavenon-additive effects,16 while R. Deaglio 1' obtained a marked fatiguefor the Dernber effect after driving a current continuously throughhis crystal for six days. The indefiniteness of these data must surelybe ascribed to the significant variation of electrical resistance, andeven of optical absorption, with oxygen content, previously men-tioned; it is not the pure crystal lattice which has been the subjectof the majority of these experiments. A definite alteration of theproperties of the crystal seems to have taken place after severalexperiments.18The Recti$er.-When the cuprous oxide is in the form of a thinlayer on a basis of metallic copper, additional phenomena maketheir appearance.The rectifier discs of commerce are formed byoxidising pure copper sheet a t about 1000" until a layer of oxidesome 0.1 mrn. thick is produced. Different methods of cooling thediscs lead eventually to different characteristics.l* l9 The super-ficial high-resistance layer of cupric oxide is removed mechanicallyor by solution, and contact with the free surface of the cuprous oxideis made, e.g., by rubbing it with carbon and pressing against it athin lead disc. Such a rectifying unit has strongly asymmetricalproperties; thus it may show an electrical resistance per sq.cm.with the copper positive, of 10,000 ohms, while with copper negativethe value may be only a few ohms. Currents of the order of 0.1amp. per sq. cm .-more under special conditions-and permissiblevoltages per disc of 10-30 volts allow the construction of cheap andreliable rectifiers of very wide application.The theoretical explanation of the rectification is in a very con-fused state. Though, of course, contact rectification betweencrystals and point contacts (e.g., the familiar wireless " crystal ',)has long been known, such rectification has always been very localand precarious. The rectification of the copper-cuprous oxide discis definitely located at the interface of the two substances (as wehave seen, cuprous oxide by itself has no marked asymmetricl5 D. Nasledov and L.Nemenov, 2. Physik, 1933, 81, 584; A., 554.16 C. Lapicque, Compt. rend., 1933, 196, 1301 ; A., 554.18 W. Leo, Ann. Physik, 1932, 15, 129; A . , 8.19 W. Bulian, Physikal. Z., 1933, 34, 745.17 Ihid., p. 1303; A . , 554; 2. Physik, 1933, 83, 179ROBINSON AND WOOSTER. 363properties), and it is the constancy of behaviour of this interface,formed, of course, from the mother copper, and protected from allcontamination, which makes the commercial applications of therectifier possible. Grondahl has well summarised the theoreticalposition at the end of 1932, and since then there do not seem to havebeen any constructive suggestions for the clearing up of the difficul-ties. The following is a very compressed summary of his discussion.Attempts have been made to ascribe the rectification to thermo-electric effects ; to a microscopic " point and plane " structure ; toan electrolytic effect due to the presence of a more highly conductingsubstance between the crystals of cuprous oxide; in terms of workfunctions of the two substances; as due toethe existence of a high-resistance thin film, deficient in electrons, between oxide andcopper ; as due to a cold electron emission into a thin space, largelyseparating oxide and copper; and in terms of the diffusion ofelectrons from copper into oxide, producing a " hump " of negativecharge whose shape will vary with the direction of the appliedpotential between oxide and copper. Of these explanations, someseem to have no valid arguments in their favour, while none is ableto explain completely the experimental facts.The Photoelectric Cell.-Comparatively recently it was found thatthese rectifier discs, if illuminated, generated a potential differencebetween the electrodes.20 The earlier experiments were made witharrangements where the light struck only the edge of the discs;later, more efficient cells were made in which the light reached theoxide layer through a semi-transparent coating, e.g., of sputteredgold, which also served as the electrode, thus allowing use to bemade of the whole sensitive surface.(Similar cells are now beingmade commercially in which selenium, rather than cuprous oxide,is the sensitive element.) Considerable electrical outputs are obtain-able from such cells; thus E. D.Wilson 21 obtains one watt persq. m. of active surface. The technique of the use and amplificationof these photocurrents is very different from that appropriate to theusual vacuum or gas-filled photoelectric cell, owing t o the muchlower internal resistance of the source.The properties of these cells have been widely investigated, thoughattempts to explain their action have not been completely successful-as one might expect, in view of the probably intimate relationbetween the rectifier and the photoelectric action. The emission ofelectrons is supposed to take place at the boundary between oxideand copper, where a " blocking layer " or " Sperrschicht " has been' 0 Grondahl and Geiger, Trans.Amer. Inst. Elect. Eng., 1927, 46, 367;Lange, Physikal. Z., 1930, 31, 139, 964; A., 1931, 9.21 Electronics, 1932, 5* 312364 CRYSTALLOGRAPEY.located by W. Schottky22 and his followers. Such cells are called(( back-wall ” cells (“ Hinterwand ”). But in certain circumstancesthe photoeffect must be located at the “ front wall I’ (Vorderwand) ofthe cell, L e . , the boundary between the oxide and the upper (trans-parent) electrode ; and in these cases the direction of flow of currentthrough the cell is reversed. The assumption of a blocking-layerpossessing unilateral transparency to electrons, and the transferenceof this layer from one side to the other of the oxide layer, accordingto the observed direction of flow of the current, certainly seems tobe rather arbitrary.Dember thinks the blocking-layer (which isessentially regarded as a characteristic of the oxidemetal junction)to be an unnecessary assumption, in view of the photoelectricpotential difference which he obtains in pure cuprous oxide crystals,as we have described; and the criticisms of E. Perucca and R.Deaglio 23 support this view. Bulian,ls in a very interesting paper,comes to the same conclusion; he shows that the same photocellmay exhibit a reversal of E.M.F. as its temperature is altered, or ifthe thickness, i.e., the transparency, of the upper layer is changed.The significance seems to lie in the optical properties of the oxidelayer ; light for which the oxide is transparent (say 1>04 p) exhibitsthe back-wall effect, while light to which it is more opaque gives thefront-wall effect.The thickness of oxide, nature of the upperelectrode, and colour of the light will decide for any particular cellwhich of these tendencies has the upper hand. Very similar are theconclusions of H. H. Poale and W. R. G. AtkinsYz4 who explored thesensitive surface of a photocell with a small patch of light of wave-length about 0.61 v; they found that the direction of current flowvaried as the patch was moved from one point to another, andascribed this effect to the varying opacity of the film.Therml Expansion of Crystal Lattices.-Before the advent ofX-rays, (‘ coefficient of expansion ” referred only to the behaviourof the massive substance.The question has been reoently raisedwhether this coefficient, measured by macroscopic methods, oughtnecessarily to agree with the expansion deduced from the spacingof the lattice, measured by X-rays. If there is a difference, what isits significance? For a crystal to give a sharp X-ray reflexion, itslattice must be perfect over many thousands of elementary periods ;but it is unlikely that this perfection is strictly maintained over thewhole of any crystal of experimental size without occasional dis-locations. Elsewhere in this Report reference is made to the massF. Waibel and W. Schottky, Naturwiss., 1932, 20, 297; A., 1932, 580;Physikal. Z., 1932, 33, 583; A., 1932, 898; W. Schottky, i b d . , 1931, 32,833.23 2. Physik, 1931, 72, 102.Nature, 1933, 131, 133; A., 209ROBZNSON AND WOOSTER. 365of evidence concerning the " secondary structure " of crystals;arguing on the basis of this assumed structure, F. Zwicky25 pre-dicted that a difference between these two expansions should exist.A difference has, in fact, been reported by A. Goetz and R. C.Hergenrother 26 with single bismuth crystals. These workers claimto have shown that along the trigonal axis the X-ray expansioncoefficient increased with temperature up to the melting point, butthe coefficient measured optically (i.e., by using interference fringesto measure the small displacements) remained constant till near themelting point, after which it decreased rapidly. They also foundthat the X-ray coefficients do obey Gruneisen's law,27 whereas theoptically determined ones did not.(Gruneisen's law requires anapproximate proportionality between expansion coefficient andspecific heat.) The discontinuities in the properties of bismuthwhich other workers have found a t 75" and near the melting pointwere not detectable on the X-ray expansion coefficient curve. Asan explanation it was suggested that an amorphous componentcoexists with the lattice, which increases relatively to the latticewith increasing temperature. On the other hand, A. H. Jay28finds no difference between the coefficients measured optically andby X-rays for silver and for quartz over a temperature range of18" to 730".*The X-ray expansions of several hydroxides, those of magnesium,calcium, and aluminium, all layer structures, have been measured by(Miss) H.D. Megaw,as who finds a general anisotropy, the expansionsperpendicular to the layer over a range 0-100" being greater thanthose parallel to the layer. I n general, the measurement of latticeexpansions in various directions should lead to results of significance.Thus J. B. Austin and R. H. H. Pcarce 30 showed for crystals ofsodium nitrate that the gradual transition occurring between 150"and 278" is marked along the c-axis, but not detectable dong thea-axis. This transition they interpret as a separation of nitrate25 Proc. Nut. Acad. Sci., 1929, 15, 253, 816; Helu. Phys. Acta, 1930, 3,26 Physical Rev., 1932, 40, 643; A., 1932, 796.E. Griineisen, Ann. Physik, 1915, 39, 257 ; E.Gruneison and E. Goens,Physikal. %., 1923, 24, 506 ; Z. Physik, 1924, 29, 141 ; G. Rorelius and C. H.Johansson, Am,. Physik, 1934, 75, 23; A., 1925, ii, 27.629 ; A., 1930, 139.28 Proc. Roy. SOC., 1933, 142, 237; A., 1237.29 Ibid., 142, 198; A., 1237.30 J . Amer. Chem. SOC., 1933, 55, 661; A., 342.* A. H. Jay (Proc. Roy. Xoc., 1934,14& 465) has repoatod the measurementsof Gootz and Hergenrother, but disagrees with their conclusions. He findsno difference between the two coefficients of expansion for bismuth, andascribes the contrary result to experimental error366 CRYSTALLOGRAPHY.layers corresponding to a change of thk vibrational energy of thenitrate ion to the rotational state. R. Becker and E. Orowan 31find that the expansion of a single zinc crystal is not smooth, butexhibits several jumps, whose steepness increases rapidly with riseof temperature.A. Miiller 32 investigated the expansion coefficientsof a series of long-chain compounds from liquid-air to melting-pointtemperatures, and found marked differences between the coefficientsmeasured along the three crystal axes.Melting-point Phenorn,em.-The measurement of thermal expan-sions leads naturally to the drastic changes in the lattice which occura t the melting point, There is considerable evidence, of interest onthe one hand to the study of liquid structure in general and on theother to the problems of nuclear formation and crystal growth, thatimmediately above the melting point a great deal of molecularorientation persists.This is shown by the experiments of A.Goetz33 on the growth of single bismuth crystals, also by W. L.Webster,34 who investigated the persistence of nuclei above themelting point for several metals, including bismuth. The nucleishowed time-temperature hysteresis effects and exhibited greatdifferences in stability. It has also been found that crystals oferythritol, held for several minutes 1" or 2" above the melting point,will recrystallise into single crystals with approximately the sameorientation as that of the original cry~tal.~S The stability ofcrystalline nuclei in superheated melts has been considered fromthe kinetic standpoint by R. Bloch, T. Brings, and W. K ~ h n , 3 ~ andit is shown that they may be expected t o melt more slowly, thoughthermodynamically they are a t least as unstable as large crystals atthe same temperature.Recent evidence as to the precise temperature range over whichmelting-point or transition phenomena occur is rather scanty,W.L. Webster3' showed that for bismuth the temperatures ofemission of latent heat, of acquisition of rigidity, and of anomalousdiamagnetic susceptibility all coincided within the experimentalerror of 0-3". A. Miiller 32 found remarkable changes in the latticesof the long-chain hydrocarbons " within a few tent,hs of a degree."The melting point of substances in a state of fine subdivision isknown to be lower 38 (often as much as 10") than that of the normal31 2. Physilc, 1932, 79, 566.32 Proc. Roy. SOC., 1930, [ A ] , 127, 417; 1932, [ A ] , 138, 514; A., 116.33 Physical Rev., 1930, 35, 193.34 Proc.Roy. SOC., 1933, [ A ] , 140, 653; A., 767.35 B. W. Robinson (unpublished results).36 2,physikal. Chem., 1931, [B], 12, 415; A., 1931, 898.37 Proc. Roy. Soc., 1931, 133, 162; A., 1931, 1221.38 Pawlow (1910)ROBINSON AND WOOSTER. 367substance; but this effect is usually ascribed to some form ofcontamination .39We may mention the work of P. A. Thiessen and E. Ehrlich,4Owho observed the structure of the crystalline sodium salts of thelong-chain fatty acids. The melting points of these acids are belowthose of their salts; and in each case as the salt was taken throughthe temperature at which the corresponding acid melted, it wasfound that the cooling curve showed a discontinuity and the expan-sion coefficient a sudden change.Lattice Theory.-A very interesting paper by H.S. Taylor,H. Eyring, and A. Sherman41 attempts to trace the growth of a crystallattice from the vapour phase, i e . , single atoms, applying theperturbation methods of quantum mechanics to the calculation ofthe energies of various possible configurations. The calculationsare applied chiefly to sodium, and the binding energies, as afunctionof distance apart in various configurations, are evaluated for fromtwo to eight sodium atoms. Prom the curves, the most probableprocess of growth of the lattice may be followed; a square havingonce been formed, the fifth atom will be added a t another corner ofthe cube, the sixth will go to the body-centre position (where it hasa considerable stabilising influence), and remaining atoms willcomplete the cube corners.Although these are the most probableconfigurations, definite amounts of energy are required for theirconsummation ; the process of building up the lattice is difficultrather than easy, and we may expect finely divided particles andsurfaces a t which crystallisation is taking place to be abnormallyactive-as, of course, we know to be the case. The elementarycube which this calculation indicates has a unit edge of 3.4 A. asagainst 4.3 A. for massive sodium.M. Born, J. E. Mayer, and L. Helmholtz 42, 43 have improved thegeneral lattice theory by three variations of the energy content.The repulsion potential is expressed as an exponential term wherebefore it was given as a power term.Van der Waals's cohesionforce is also introduced into the calculation. This theory has beenapplied to the empirical data for the alkali halides. A very simplerelation between lattice energies and ionic radii and valency hasbeen discovered by A. F. KapustinskiM and applied by him and39 N. Schoorl, Proc. K . Akad. Wetensch. Amsterdam., 1932, 35, 378; 2.40 Ibid., 1932, [B], 19, 299; A., 116.4 1 J . Chern. Physics, 1933, 1, 68; A . , 213.42 M. Born and J. E. Mayer, 2. Physik, 1932, 75, 1 ; A., 1932, 564.43 J. E. Mayer and L. Helmholtz, {bid., p. 19; A., 1932, 564.44 2. physikal. Chem., 1933, 82, 257; A., 1001.physikal. Chern., 1932, [ A ] , 160, 158; A . , 1932, 799B.Veselovski45 to about 100 substances. If the lattice energy beplotted against l/(rK + rA) where rK, T, are the radii of kation andanion respectively, then for these substances the points lie close toa straight line, so that the maximum divergence of the lattice energyfrom that corresponding to the straight line is about 10%. Thevalidity of this formula makes approximate calculations very simple.Compressibility of Crystals.-The most direct way of investigatingthe bonds between the atoms of a crystal lattice is, of course, to alterthe distance between them by externally applied forces; but tomake any perceptible effect, these forces must be very great. Themasterly technique of P. W. Bridgman 46 has rendered possibleaccurate measurements at pressures as high as 12,000 kg./sq.cm.;a full description of the methods and precautions necessary for suchwork is given in his book.If we attempt to form the equation €or the potential energy interms of position of an atom in a lattice-considering, e.g., an ioniclattice of the alkali halide type-then the attractive term is usuallyascribed t o the electrostatic (Coulomb) forces of the oppositelycharged ions, so that its form is known. E. Madelung4' hascalculated the potential energy due to this attractive force forvarious types of lattice. We are, however, in ignorance of the formof the repulsive term necessary to give stability to the lattice; allthat we can say i s that it vanishes a t great distances and is verylarge at small ones.The measured lattice spacing, of course,represents the position of equilibrium between these two forces,It is usual to express this repulsive term as a convergent infiniteseries and to determine the coefficients of the separate terms directlyby comparison with the measured compressibilities (for if thepotential function is known the theoretical compressibility can bewritten down immediately). This has been done, e.g., by J. C.Slater,48 whose work also confirms our assumption above that theattractive forces are purely electrostatic in nature. But the scopeof the measurements, even over the largest possible range, is notsufficient t o determine more than the first two coefficients in theexpansion, so that extrapolation of the potential function beyondthe experimental range becomes very hazardous. Several attempts 43have been made to suggest forms of function which on theoreticalgrounds would permit of this extrapolation with more confidence.Dielectric Properties of Crystals.-The theory of the dielectric45 2.phyeikal. Chern., 1933, 22, 261 ; A., 1001.46 " The Physics of High Pressures," and subsequent papers, especially47 Physikal. Z., 1918, 19, 529; Sherman, Phil. Mag., 1932, 14, 745.4 8 Physical Rev., 1924, 23, 488.in Proc. Amer. Acad. Arts SciROBINBON AND WOOSTER. 369properties of crystals and a discussion of the work on ice was sum-marised by P. Debye.49 More recently J. Errera 6og 61s 62 and hisschool have investigated a number of substances, in many cases withspecial reference to the relation of the dielectric constant to thecrystallographic direction.The work falls into two parts accordingas the crystals do or do not contain orientable permanent dipoles.The alkali and alkaline-earth halides illustrate the rule that theportion of the dielectric constant which may be attributed to relativedisplacement of the ions of opposite electric signs decreases withincreasing size of anion for a given kation and also decreases withincreasing size of kation for a given anion. The nitrates of lead,calcium, strontium, and barium follow the same rule, the case of lead,which has the same ionic radius as barium, showing the increase ofionic polarisation with increase of atomic number. The fluoridesof iron, cobalt, and nickel have ionic polarisations which increase inthis order, corresponding to the decrease in ionic size.The remark-able case of rutile and anatase, for which the dielectric constants are114.0 and 48.0 respectively, has not received adequate explanation.The orthorhombic sulphates of strontium, barium, and lead showan unexpectedly high value of the dielectric constant along theb-axis. As for the nitrates, the ionic polarisation for the lead salt ismuch greater than for the barium and the strontium salt. Therhombohedra1 carbonates show the greatest ionic polarisation alongthe trigonal axis, whilst for the rhombic carbonates the differencesalong the crystallographic axes are much less marked. The platino-cyanides of calcium, strontium, and barium show in each case thegreatest dielectric constant perpendicular t o the principal cleavage.Although they contain water of crystallisation, they do not shornthe dispersion with frequency characteristic of ice.The magnesiumsalt, containing seven instead of five molecules of water of crystallis-ation, does show this anomalous dispersion for the direction per-pendicular to the good cleavage (001). Thus two molecules of waterof crystallisation are differentiated from the others and are orientablewhen the field is applied in a particular direction.The investigation of the crystals containing permanent dipoleswhich may be oriented in the electric field has given results whichsuggest important applications in problems of crystal structure. Asmentioned above, salts containing water of crystallisation sometimesshow a dispersion of the dielectric constant both with frequency 4ndwith temperature similar to that observed in ice.It occurs in49 “ Polare Molekeln,” Loipzig, 1929.50 Physilcal. 2. Soviet Union, 1933, 3, 443.5 1 Physikal. Z., 1933, 34, 368 ; A., 663.52 J. Errore and H. Bra~~eur, Compt. rend., 1833, 197, 480; A., 1000370 CRYSTALLOGRAPHY.CuS04,5H20 and Na2S,0,,6H20 and appears to be associated withloosely bound water of crystallisation. Thus in gypsum and inmost alums 52 there is no such anomalous dispersion. Rochelle salthas been extensively studied in connexion with its piezo-electricand dielectric characters. Errera has extended the frequency rangeof the investigation and found the same great increase in the di-electric constant a t a particular frequency and over a restrictedtemperature interval of - 10" to -+ 25".The maximum dielectricconstant recorded is 1300 for a field strength of 200 volts/cm. Theelectrical properties of mixed crystals from sodium potassiumtartrate 5 3 3 j4 have also been studied. The application of X-raydiffraction to the study of the displacement of the ions in Rochellesalt 55 in an electric field marks an important advance. Anotherphenomenon50 which may be inportant in the study of liquidcrystals is the anomalous rapid rise in the dielectric constant ofacetic acid and of p-azoxyanisole just a t the moment of crystallis-ation. It occurs only for low frequencies. The tentative explana-tion put forward by the discoverer is that just on solidifying theliquid condenses into a large number of polar groups which in afield of low frequency can orientate themselves.Electrical Byeakdown in Crystals.-During the past three years asystematic study of the phenomena accompanying electrical break-down in crystals has been made.To cause the discharge to passthrough the crystal it is necessary to overcome the tendency of thedischarge to pass round the edges of the specimen. This has beendone by (Miss) L. Inge and A. Walther,56 who used a point electroderesting on the crystal which in turn rested on a large plate. Undernormal conditions of discharge, of duration greater than 10-7 sec.,the crystal is destroyed, but by using special electrical circuits theseauthors were able to apply the potential for so short a time that thedischarge did not strike right through the crystal.A. von Hippel 57also succeeded in avoiding spurious effects by placing the crystalbetween two slightly rounded electrodes, filling the space betweencrystal and electrodes with specially prepared bees-wax, and mount -ing this apparatus in nitrogen a t a pressure of about 100 atmospheres.The results obtained by the two methods are not always the same.For instance, for rock salt, Inge and Walther find that the dischargepasses from the pointed electrode along [lll] directions in thecrystal, whereas Hippel finds that first [110) and then [ill] are the53 M. A, EremBev, P. P. Kobeko, B. V. Kurtschatov, and I.V. Kurtschatov,J. Exp. Theor. Physics Russia, 1932, 2, 102.54 B. V, Kurtschatov and M. A. EremBov, Physikal. 2. Soviet Union, 1932,1, 140; A . , 1932, 560.5 5 H. Staub, Physikal. Z., 1933, 34, 292; A., 447.5 6 2. Physik, 1930,64,830 ; A., 1930,1504 ; Arch. Elektrotechnik, 1929,22,410.57 2. Physik, 1931,67,707 ; 68,309 ; 1932,75,145 ; A., 1931,546 ; 1932,565ROBINSON AND WOOSTER. 37 1favoured directions. In the latter experiments the dischargealways branches from the anode, whereas in Inge and Walther’swork the branches grow from the point in each case, but much betterwhen it is positive than when it is negative. The difference betweenthese results is presumably due to the fact that in Hippel’s apparatusthe discharge begins in a fairly uniform field and only after the trackhas penetrated a certain distance into the crystal does the gradientat its end become very steep.In Inge and Walther’s method thegradient a t the beginning of the discharge, being that around thepoint, is very steep. For barytes, the favoured directions for thedischarge are approximately [loo], [loll and [ O l l ] ; in calcite, twoof the [ 1111 directions (which correspond to [ 11 11 in sodium chloride) ;in fluorite [loo] and 111111 occur according as the plates are cutparallel to (1 11) or (100) ; in beryl the discharge tracks run parallelto the hexad axis-a direction in which there are “ holes ” in thestructure. Hippel has investigated the variation of the strength ofthe breakdown field in the alkali halide series and finds, after makingallowance for small differences for different crystallographic direc-tions, values which range from >2.0 x lo6 volts per cm.for lithiumfluoride to 0.50 x lo6 volts per cm. for rubidium iodide. Parallelwith this decrease in the breakdown field goes a change in theappearance of the tracks. For lithium fluoride they are sharp butbecome more diffuse as the ionic radii increase. Experiments onmagnesium oxide indicate that the exchange of a doubly for a singlycharged ion in the same structural type leads t o a great increasein dielectric strength. Considerable difference of opinion existedbetween the investigators as to the mode of origin of the tracks,whether by acceleration of positive ions or of electrons.The totalevidence appears to favour Hippel’s view that the phenomenon isprimarily due to electrons, accelerated towards the anode, acquiringsufficient energy to ionise by collision. The track of the electronsbecomes conducting, and as it advances towards the cathode thepotential gradient at its advancing end increases. Plastic deform-ation increases the dielectric strength, which would not be expectedif ions were the conductors, since the cracks introduced by thedeformation would provide easier paths for the ions than exist in theundeformed crystal. Other difficulties in accounting for the factsby an ionic discharge are that the time of the discharge is shorterthan would be expected in this case, and that the breakdownpotential is independent of the previous ion current.Hippelsupposes that a slow-moving electron makes collisions with theatoms and loses energy to them in thermal agitation. As thevelocity of the electron is increased, it will lose more energy perimpact, corresponding to the greeter number of acoustic, thermal,and infra-red characteristic frequencies to which it may lose energy373 CRYSTALLOGRAPHY.If it has energy greater than the characteristic frequency of thelattice, i.e., the “ reststrahlen ” frequency, i t will begin to accumulateenergy and bring about the discharge. This is borne out by theconstancy (within 12%) for all the alkali halides of the ratio of theenergy acquired by an electron in falling in the breakdown fieldthrough a length equal to the side of the unit cell to the quantumassociated with the “ reststrahlen ” frequency.The correlationof the direction of the breakdown paths with the crystal structureis not yet clear, but it has been suggested with a certain amount ofexperimental justification that the discharge occurs in directions inwhich there are rows of neighbouring ions having the same sign ofcharge.The Elastic and Plustic Properties of Crystals.-Elasticity. Severalnew methods of determining elastic moduli have been intro-d ~ c e d . 6 ~ 6 ~ I n one of thesess H-shaped rnasscs were attached tothe ends of the rod under test, which was supported by a longthin bifilar suspension. The outer limbs of the upper H werebar-magnets each of which was surrounded by coils of many turnsin which currents were induced by the vibration of the magnets.These induced currents were amplified and the frequency wascompared with a known frequency impressed on the same circuit.The bar was set into oscillation by a blow with a small hammerimparted to the end of one of the magnets.This mode of producingoscillation gave rise to both torsional and flexural vibrations, butthese could be distinguished by observation of the sum and differenceof the currents induced in the two coils. The elastic moduli werecalculated from the vibration frequency and the dimensions of therod. Another apparatus 59 for determining elastic moduli con-sisted of two identical gravity pendulums suspended from the topof the test plate which was clamped at the bottom. When onependulum was started, the slight movement of its support causedthe other pendulum to begin swinging.This process continued untilthe first had stopped swinging, its energy having been transferredto the second. The time between two successive stoppages of thesame pendulum, together with the relevant dimensions of thoapparatus, were sufficient data for calculating Young’s modulus forthe crystal plate. By using torsional instead of gravity pendulums,the rigidity modulus was found. The method devised originallyby Quirnby 62 has been adapted to measurements of elastic moduli68 E. Goens, Ann. Physik, 1930,4, 733 ; A., 1030, 764.80 E. Goens, Ann. Physik, 1933, 16, 793.61 R. M. Davies, Phil. Mag., 1933, 16, 97; A., 893.62 S . L.Quimby, Physical Rev., 1926, 25, 568.63 J. Zacharias, ibid., 1933, 44, 116; A., 893.P. le Rolland and P. Sorin, Cornpt. rend., 1933, 196, 536ROBINSON AND WOOSTER. 373by Z a ~ h a r i a s . ~ ~ A composite bar consisting of quartz cemented tothe end of the bar under test was suspended by threads so as tovibrate freely. The quartz had been cut so that it could be set intolongitudinal vibration on applying an alternating potential to itsside faces. To find the unknown modulus for the crystal, it wasnecessary to know the frequency for which resonance occurredbetween the vibrations of the composite rod and the electricalcircuit, the dimensions of the quartz and crystal, and the elasticconstants of the quartz. Recent determinations of elastic constantsinclude those of 65 quartz,66, 67 gold-silver,68 alumin-ium,69 nickel,63 and Rochelle salt.61 The dynamicalmethods of determining the elastic moduli depend on the theory oftorsional and flexural vibrations of a monocrystalline rod.Thishas been worked out by E. Goens,'l using the fundamental work ofVoigt. Although primarily concerned with piezo-electric characters,the book by P. Vigoureux 72 gives an excellent pfesentation of thetheory and data of the elastic properties of quartz.Plasticity. Much has been written on the plastic deformationof metals, but it would not be profitable to attempt a review of thissubject here. Among non-metallic substances, roclr-salt has beenmost studied. Whilst there is good reason for investigating thissubstance on account of its simple crystal structure and the developedstate of its lattice theory, yet generalisations are dangerous until thecorresponding data are available for more substances.The experi-ments on the increase of the tensile strength of rock-salt on wetting,described originally by JoffB, have called forth many investig-ation~.'~-*~ Some of the results oE these have verified the increase64 I?. W. Bridgman, Proc. Amer. Acad. ArtsSci., 1932,67,29; A., 1932,681.6 5 E. Goens and E. Schmid, Naturwiss., 1931, 19, 376; A., 1931, 673.6 6 V. Freedericksz and G. Michailov, 2. Physik, 1932,76,328 ; A., 1932, 799.67 R. de Mandrot, Helv. Phy8. Acta, 1932,5, 362.6 8 H. Rohl, Ann. Physik, 1933, 16, 887 ; A., 561.69 E. Goens, ibid., 17,233; A., 768.70 K.Kimura, Sci. Rep. Tdhoku Imp. Univ., 1933, 22, 553.71 Ann. Physik, 1932, 15, 455.72 " Quartz Resonators and Oscillators," London, 1931.73 A. Smekal, Physikal. Z., 1931, 32, 187; A., 1931, 853.74 U. Heine, 2. Physik, 1931, 68, 591; A., 1931, 791.7 5 E. Rexer, ibid., 72, 613; A . , 1932, 13.76 G. F. Sperling, ibid., 1932, 74, 476 ; A., 1932, 462.77 N. Seljakov, ibid., 1932, 76, 535; A., 1932, 904.78 L. Piatti, ibid., 77, 401; Nuovo Cim., 1932, 9, 102; A., 342.59 Idem, ibid., p. 180.80 N. Seljakov, 2. Krist., 1932, 83, 426.81 IC. H. Dommerich, 2. Physik, 1933, 80, 242; A., 342.82 L. Piatti, NUOVO Cim., 1933, 10, 43.83 N. N. Davidenkov and M. V. Klassen-Nekludova, Physikal. 2. SovietUnion, 1933, 4, 25374 CRYSTALLOGRAPHY.of tensile strength caused by ~ e t t i n g .~ * - ' ~ A rotation of someportions of the crystal relative to others through angles up to 30"is observed in the wet salt but not in the d r ~ . ~ 6 The distortion ofthe lattice makes it monoclinic near the surfaces along which glidingoccurs.73, 76 The crystals show a considerable increase in strengthafter previous plastic deformation, but this author finds that theelastic limit of previously undeformed and uniformly dissolvedcleavage rods of rock-salt agrees within the error limit with that ofdry natural crystals.81 Measurement of the hardness of rock-saltby the damping method with the Kusnetzov pendulum 83 shows nodifference in hardness between the dry and the previously wettedrock-salt.Further, the elastic limit of the rock-salt in water,determined in polarised light, is the same as that of the dry salt.Investigations on dry rock-salt have been summarised by A.Smeka1.84 He concludes that the purest crystal prepared under thebest conditions still has inhomogeneities which are responsible forthe difference between the characters, in particular the breakingstrength, of the actual and the theoretically ideal crystals. Onapplying stress to the crystal, glide planes develop from theseinhomogeneities as soon as the stress exceeds the breaking or shear-ing stress for the ideal crystal. The crystal is no longer single butpolycrystalline, and further deformation merely alters the numberand mutual orientation of the individual grains.At a temperaturesomewhat below the melting point and for extensive deformationthe polycrystal reaches a breaking strength close to that of thetheoretical crystal, a fact which is explained as being due to the greatdisarrangement of the grains which prevents any further gliding, sothat the real strength of the crystal is observed. The justificationfor this theory is based on extensive experiments. The strength ofthe pure crystal (99.995% pure) has been found at temperaturesfrom 1.2" K. up to the melting point.85-88 The strength is constantup to - 190°, decreases to + 40", and then increases until near themelting point. The effect of known amounts of impurities has alsobeen 893 The complications arising from theplasticity have been almost overcome by working at low tempera-tures.The influence of impurities may be seen from the followingexample.s6 Rock-salt containing 0.045% of strontium chloride has,at room temperature, a tensile strength of 1000 g./mm.2 as compareda4 Phyqikal. Z., 1933, 34, 633; A . , 1005.W. Burgsmuller, 2. Physik, 1933, 80, 299; A., 342.86 Idem, ibid., 83, 317.*' W. Burgsmuller and K. Steiner, ibid., p. 321.8 8 W. Theile, ibid., 1932, 75, 763; A., 1932, 683.A. Edner, ibid., 1932, 73, 623; A., 1932, 219.9O H. Schonfeld, ibid., 1932, 75, 442ROBINSON AND WOOSTER. 375with 213 g./mm.2 for the pure substance. The increase of strengthcaused by the impurity decreases as the temperature is lowered,until a t - 110" it is zero. On still further lowering of temperature,the impurity weakens the crystal until a t - 190" and - 252" thesame amount of impurity reduces the strength from 560 g./mm.2 to135 g./mm.2 A.Smekal8* concludes that the influence of impuritiescould certainly explain the difference between the ideal breakingstress (20,000--40,000 kg./cm.2) and that of the best realisablecrystal (60 kg./cm.2). If an artificially coloured rock-salt crystal bedeformed as far as possible, its absorption maximum moves towardsthe red (465 to 475 P ) . ~ The corresponding change in excitationenergy is 0.05 volt, and if this be interpreted as being the strainenergy of the lattice a t the coloured inhomogeneity, the correspond-ing stress is 16,000 kg./cm.2. This value is close to that of thetheoretical rock-salt crystal and the observation shows that thestress a t the points where gliding occurs reaches nearly to the idealbreakdown stress.As the crystal is progressively deformed, irre-versible double refraction91992 first appears a t a stress of about6 kg./cme2. At the same stress, the photoconductivity begins toincrease 93 and the ionic conductivity first begins t o vary.S3 Thevery beginning of the process of deformation is thereforc connectedwith an increase in the number of inhomogeneities and with thebreak-up of the single crystal. It is therefore improbable that idealgliding ever occurs, and that even in metals, e.g., zinc, where thegliding appears so simple, there is from the beginning of the deform-ation an irreversible increase in the number of faults in the lattice.The plastic properties of rock-salt a t temperatures up to 600" havebeen investigated with special reference to annealing and increaseof strength with plastic deformation.88, 94-g8 Since very smallamounts of impurities greatly affect the mcchanical properties ofrock-salt, a method has been devised of observing them by meansof an ultramicroscope. The number of ultramicroscopic particlescan be correlated with the mechanical properties and ultra-violetabsorption.99, 1 The energy imparted to the crystal as a result of9 1 I.W, Obreiniov and L. W. Schubnikov, 2. Plzysik, 1927, 41, 907.92 W. Schutze, ibid., 1932, 76, 135; A . , 1932, 799.93 H. J. Schroder, ibid., p. 608; A . , 1932, 898.94 K. Przibram, Sitzungsber.Akad. Wiss. Wien, 1932, 141, 63.g 5 Idem, ibid., p . 645.96 E. Rexer, 2. Physik, 1932, '95, 777.97 F. Rinne and W. Hofmann, 2. Krist., 1932, 83, 56.98 N. A. Brilliantov and I. W. Obreimov, Physikal. 2. Soviet Union, 1933, 3,99 R. Matthai, 2. Physik, 1931, 68, 85; A., 1931, 563.83.1 E. Rexer, Physikal. Z . , 1931, 32, 215376 CRY 13TAW;OGRAFHYmdeformation may, in the case of metals, be of a kind which cannotbe detected by X-rays, i.e., is not accounted for by the change inthe lattice dimensions.2 In one case it was shown that about O-lyoof the extension energy was retained by the strained lattice, whereas15% of the energy was left in the crystal under some other form ofstrain. Measurements of t'he force and work necessary to producecleavage in several crystals under various conditions have beenTwo investigations on quartz 798 have thrown light onits previously obscure plastic properties.The production ofmechanical twins and their demonstration by etching is both strikingand convincing. The breaking stress under compression is fifteentimes smaller at the transition temperature, 573", than at roomtemperature. The plastic properties of fluorspar calcite,1° potass-ium halides,g2 mica 11 and the nature of the glide planes in a catholicselection of substances 12, l3 have also been studied.Secondary Structure.-It has been generally assumed that anideal lattice of the rock-salt type is stable. The calculations of thetensile strength based on this assumption lead, as was mentionedabove, to a value several hundred times greater than that observedin the best crystals.This led F. Zwicky l4 to put forward the theoryof '' secondary structure which is, briefly, that superimposed onthe well-established lattice, known from X-ray diffraction, i s anotherlattice on a scale such that its unit cell would have a side about1 p long. The planes of this lattice are supposed to contain moreions per unit area than parallel planes in the rest of the crystal.According to this picture, the crystal would, however pure and how-ever carefully prepared, grow with & " block " structure. Theweakness of the natural crystals is, according to this view, a neces-sary consequence of the nature of the secondary structure. Thishypothesis has been used to explain many of the periodic or pseudo-periodic phenomena occurring in crystals and the existence of theseperiodicities, on a scale of about 1 p, has been taken by Zwicky andV.Caglioti and G. Sachs, 2. Yhyoik, 1932, 74, 647; A., 1932, 452.H. Tertsch, 2. Krist., 1930, 74, 476; A., 1931, 422.Idem, ibid., 1931, 78, 6 3 ; A., 1931, 1360.Idem, ibid., 1932,81, 264; A., 1932, 452.Idem, ibid., p. 275; A., 1932, 452.A. Schubnikov and I(. Zinserling, ibid., 1932, 83, 243.* K. Zinserling and A. Schubnikov, ibid., 1933, 85, 454.S E. Rexer, ibid., 1931, 78, 351.10 0. Mugge, ibid., 1932, 82, 59.11 E. Orowan, 2. Physik, 1933, 82, 235.12 0. Mugge, 2. Kriat., 1930, 75, 32; A , , 1931, 415.18 Idem, aid., 1931, 76, 359; A., 1931, 550.l4 Helv.Phys. Acta, 1930, 3, 269; 3, 466; 4, 49; 1933, 6, 210; PhysicalRev., 1932, 40, 63;- A., 1932, 564ROBINSON AND WOOSTER. 377others as support for the theory. Smekal has concluded that nosuch hypothesis is necessary to explain the weakness of the naturalrock-salt crystal, the natural inhomogeneities both mechanical andchemical being the prime cause of the diminution of strength. Ifthe hypotheses of secondary structure is unnecessary it remains toaccount for the large-scale periodicities. A short account of thesefollows.The diffraction effects observed in crystals of potassium chloratewere thoroughly investigated by R. W. W00d,15 and he concurredwith Stokes in interpreting them by assuming periodic twinning.As many as 1000 planes, almost equidistant from one another, mustoccur in a single crystal in order to explain the narrowness of thespectral band reflected at a particular angle by the crystal.Thenatural (111) faces of perfect crystals of bismuth show fine strizintersecting at 60". A. Goetz l6 has attempted to show that thedistances between the striz are all simple multiples of about 1 p bysuperposing a triangular network of lines on the observed patternand adjusting the size of the triangles until each observed line coin-cided with a line of the net. The side of the smallest triangle sofound is given as 1.4 -+ 0.2 p. The correspondence between net andstria? is not convincing, particularly as neighbouring lines are notalways parallel and their distance apart is only a few times theirwidth.Goetz has etched the (111) faces of bismuth crystals verycarefully, and found that triangular pits are developed. He assertsthat the sides of these triangles coincide with the lines of the netalready fitted to the crystal in the previous observation. Thepublished photographs do not give much support to this. Anotherline of research has been pursued by M. Straumanis.l7, 18 Crystalsof cadmium and zinc were prepared by sublimation under carefullycontrolled conditions. The edges of the hexagonal plates so formedwere bounded by heavily striated faces. A vertical section throughthe crystal showed a step-like formatioil at the edge, arid he foundthat the heights of the steps were approximately whole multiplesof 0.8 & 0.1 p.Zinc crystals also formed by sublimation weredeformed so that gliding took place along the basal plane. Thecrystal after deformation appears, when viewed from the side, like apack of cards pushed over to one side. The steps, however, are notbounded by vertical faces but by curved sloping faces, showing thatgliding has occurred also on planes in between those usually denotedas slip-planes. The latter are 0.8 & 0-1 p apart, a distance equal toI5 " Physical Optics," 1911, p. 160.16 Proc. Nat. Acad. Sci., 1930, 16, 99; A., 1930, 670.17 2. physikal. Chem., 1931, [I?], 13, 316; A., 1931, 1115.18 2. Krist., 1932, 83, 29; A., 1932, 1080378 CRYSTALLOGRAPHY.the height of the steps obtained when the crystal is grown bysublimation. It is difficult to judge froni the photographs withinwhat limits the separation of the glide planes is a whole multiple of0.8 p, but it is clear that the separation is often several times thisdistance.Thus, although the phenomenon of gliding may beassociated with a lattice the planes of which are 0.8 p apart, they arenot all equivalent-a fact which distinguishes this type of latticefrom that determined by X-ray analysis. On a scale about 100 timesgreater than that involved in the above experiments, periodicprecipitation of cadmium between single crystals of zinc wasobserved by Straumanis.19 Photographs show that the precipita-tions occur at intervals which are constant to within about 10%and vary according to the rate of crystallisation and the concentra-tion of cadmium from about + to mm.Investigations have alsobeen made of the etching of various faces of zinc crystals by hydro-chloric acid, and in many cases the etching has taken place alongnearly parallel and equidistant lines. It is important to observethat the lines are sometimes not parallel, though the deviationfrom parallelism is small. I n a paper zo on the lattice-likeinclusions of magnesium silicate in calcite from North Burgess,Ontario, are some photographs which show inclusions arrangedparallel and, over a limited region, equidistant. The separation isof the order of 0.1 mm. When larger regions are considered itwould be necessary, if a fundamental repeat distance were sought,to choose one not much larger than the width of the inclusions them-selves.The inclusions were presumably deposited whilst the crystalwas solid along the intersections of twin planes. The regularity ofthe spacing of the needles is thus an indication of the regular spacingof the twin planes.It has been suggested that the discrepancy between the wave-lengths of X-rays determined from crystals and from artificiallyruled gratings is due to the secondary structure.21 The planes ofthe secondary structure are supposed to be of greater density thanthe rest, and this would therefore cause a small error in the deter-mination of wave-length. However, there is disagreement betweenvarious authors 22-24 as to whether there is any discrepancy betweenthe wave-lengths obtained by the two methods. Some physicalproperties of bismuth crystals are different according as they are19 2.anorg. Chem., 1929, 180, 1.20 R. P. D. Gradham, Min. Mag., 1915, 18, 252.21 F. Zwicky, Proc. Nut. Acad. Sci., 1930, 16, 211 ; A , , 1930, 660.22 E. Backlin, Phpical Rev., 1932, 40, 112.as J. A. Bearden, ibid., p. 471.24 Yuching Tu, ibid., p. 662BERNAL AND CROWFOOT. 379grown in or out of a magnetic 26 No difference could be foundbetween the spacing of the (111) and (111) planes of crystals grownin, and of those grown out of, a magnetic field, indicating that thechanges in physical properties were associated with the way in whichthe blocks of perfect crystal were assembled rather than with theproperties of the blocks themselves.27 Mention has already beenmade of the reported difference between the thermal expansiondetermined optically and by X-rays for bismuth.This result isinterpreted as indicating that the portions of the crystal between theplanes of the secondary structure expand uniformly up to themelting point, but that the planes of the secondary structure beginto break down at a temperature 35" lower.In attempting to form an opinion on the existence of the secondarystructure from these periodic phenomena, it is important to bear inmind our ignorance of the exact nature and origin of the inhomo-geneities which are so important in rock-salt. The other substancescited were probably not so pure as the rock-salt used by Smekal, anda rhythmic separation of the impurities might well give rise to theobserved phenomena.A special theory is not required to explainthe separation of cadmium from zinc on a macroscopic scale, and itseems unnecessary to introduce one when the scale is reduced athousand- f old.B. W. R.W. A. W.STRUCTURAL CRYSTALLOGRAPHY.The methods of structure analysis have not been radically alteredin the past two years, but they have acquired additional certaintyand precision. The Weissenberg goniometer has come more andmore into use, and, together with other moving-film goniometers,2has made the indexing of reflexions and determination of spacegroup a purely mechanical operation, so that mistakes on thisscore have practically disappeared. At the same time the use ofintegrating photometers 3 has made photographic intensity deter-mination as accurate as ionisation spectrometer measurement,thus immensely extending the range of crystals available for com-2 5 A.Goetz and M. F. Sasler, PhysicaE Rev., 1930, 36, 1752; A., 1931, 292.26 A. Goetz and A. B. Focke, ibid., 1931, 37, 1044; A., 1931, 792.27 A. Goetz and R. C. Hergenrother, ibid., 1932, 40, 137; A., 1932, 681.1 2. Physik, 1924, 23, 229; J. Bohm, ibid., 1926, 39, 557.a H. Braekken, 2. Krist., 1932, 81, 309; A., 1932, 449; E. Schiebold, ibid.,1933, 86, 370; E. Sauter, ibid., 1933, 84,461; 85, 156; A., 451, 480; B. W.Robinson, J . Sci. Instr., 1933, 10, 165; A., 689.Idem, ibid., p. 233; A., 1026380 CRYSTALLOGRAPHY.plete analysis, especially for organic crystals, where the work ofthe Davy Faraday Laboratory (see below, p. 414) has been sooutstanding.Fourier analysis, though still very laborious, hasbeen undertaken in many cases, and it has become the recognisedend of a structure determination to give, not only the atomicpositions, interatomic distances, and angles, but some idea of theelectronic density at different points in the cell. At the sametime, methods of handling small crystals,4 amorphous solids, fibres,and liquids have so improved that it may now be fairly claimedthat there is no material, however minute in quantity or unpromis-ing in appearance, that will not yield significant information onexamination with X-ray or electron methods.Quite apart from structure analysis, X-ray methods can be usedfor accurate quantitative analysis by measurement of intensitiesof diffraction : this has been used by M.E. Nahmias to study thechanges occurring in the production of porcelain from clay.6 Ofmore general application have been the methods based on theprecise measurements of spacing. Lattice dimensions can now bemeasured to one part in 100,000, and without undue difficulty toone in 10,000, particularly by a simple method, adaptable to anypowder camera, developed by A. J. Bradley and A. H. Jay.'Dimensions thus obtained are highly characteristic for a pure sub-stance. J. ,A. BeardenY8 for instance, found for three samples ofcalcite from different localities spacings of 3-02805, 3.02807, and3.02808 A. at 20". However, traces of substances in solid solutionaffect the spacings in a regular manner, and the variation can beused to determine the amount of the phase constituents.Byquenching a t different temperatures, or still better by the use of ahigh-temperature X-ray camera: phase boundaries can thus befollowed with an accuracy which, for the vertical lines of theequilibrium diagram, surpasses that of any other method.l* Pre-cision spacing measurements have also been used for measurement0. Kratky and K. Eckling, Z . physikal. Chem., 1932, [R], 19, 278; A . ,2. Krist., 1932, 83, 329; &4., 44; ibid., 1933, 86, 319.6 Ibid., 1933, 85, 355; A., 892; A. J. Bradley and A. L. Roussin, Tmm.C'eramic SOC., 1932, 31, 422.Proc. Phylsical soC.7 1932, 44, 563; 1933, 45, 507; A., 1932, 1078; 1933,891; (Miss) H.D. Megaw, Phil. Mag., 1932, [vii], 12, 130; A., 1932, 800; cf.G. Sachs and J. Weerts, 2. Plzysik, 1930, 60, 481 ; 64, 344.139.* Physical Rev., 1931, [ii], 38, 2089; A., 1932, 217.A. H. Jay, 2. Krist., 1933, 86, 106; A., 1026; Proc. Physical Soc., 193345, 635.lo E. A. Owen and L. Pickup, Proc. Roy. SOC., 1932, [A], 137,397; A,, 1932,990; 1933, [A], 139,526; A., 464; A. Westgren, Assoc. Int. Essai Mat., 1932,1, 484; A., 1007BERNAL AND CROWFOOT. 381of true thermal expansion of crystals in different directions,ll as theycan be used where interferometer methods are out of the question.Extension of X-ray methods to low temperatures have also beennotably improved, especially through the work of M. Ruhemann.12CRYSTAL CHEMISTRY.Daltonian and Berthollide Compounds.Possibly the most interesting development of crystal chemistryis that centring round the regular and irregular structure of com-pounds and solid solutions.From this field, evidence has beengradually accumulating which goes far towards changing our ideason the laws of chemical composition. These laws, in particularDalton’s laws of constant and multiple proportions, were foundedon the study of simple molecular compounds, hydrogen, oxygen,nitrogen, water, carbon dioxide, etc., and binary ionic compounds,such as sodium chloride and calcium carbonate. Here these laws holdin full rigidity, but there was something in Berthollet’s view thatthey should not be extended to the whole of chemistry. In fact,it is only in molecular chemistry-which includes organic chemistry-that they are strictly true.In ionic chemistry they break down forthe simplest ternary systems, as mixed oxide and mixed halidesystems show, and for the chemistry of ores and metals even theconception of valency breaks down completely, and attempts toapply it only lead to unwieldy or incorrect formulze.The validity of constant composition and valency rules depends,we see now, either on dehite directed electron-sharing bonds, oron the necessary equality of positive and negative ionic charges.The first is true only of molecules or molecule ions (NO,)-,(CH,COO)-,the second only leads to simple laws if all the ions of each sign areof the same kind. Apart from these cases, regularity may arisefrom the conditions of geometrical packing in crystals.Here it isnot so much a question of bonds or charges, but of the size andshape of the units concerned, atoms, ions, molecule ions, or molecules.If the sizes of the various components are very different and thestructure well fitted together, compounds result apparently of thetrue Daltonian type. Such compounds are NaK, the alums, or themolecular compounds of organic chemistry. They differ, however,from true chemical compounds by having no existence apart fromthe crystalline state.If, however, the different units from which the crystal is built11 (Miss) H. D. Megaw, PTOC. Roy. SOC., 1933, [A], 142, 198; A. H. Jay,12 2. Physik, 1932, 76, 368; J. A. Santos and J. West, J. Sci. Instr., 1933,ibid., p.237.10, 219 ; A., 925382 CRYSTaLLOGRAPHY.only differ slightly in size and shape, they may come to occupyindifferently the same positions. The simplest case of this kindis that of normal solid solution, as of silver and gold or the mixedcrystal NaC1, KBr, and NaBr which may be written (NaK)(ClBr).That size, and not charge or chemical character, is the pre-ponderating factor, is shown by the work of E. Posnjak andT. F. W. Barth13 in the case of lithium ferrite LiFeO,, which hasthe rock-salt structure with Li+ radius 0-78 8. or Fe3+ radius 0.67 A.occupying indifferently the cation positions. A more complicatedcase is presented by the spinels discussed in the last report (1931,28, 299). Barth’s theory of interchangeability of ions, which hecalls “ variate atomic equipoints,” has been established l4 for thespinels MgFe,O,, MgGa,O,, MgIn204, TiMg,O,, TiFe,O,, and SnZn,O,,but it does not hold for the aluminates of manganese, iron, cobalt,nickel, and zinc, and probably not for the typical spinel MgAl,O,.This conclusion has been confirmed by F.Machatschki : l5 the previousdifferences had arisen because the observers were working on differentcompounds. The ion equivalence is not confined to cations. E.Kordes l0 has prepared spinels of the type (Mg,Li)Al,(O,F),. Thepresence of this extreme form of ion equivalence is, of course, con-nected with the high crystallising temperature-in the region oflOOO”-of these crystals, That this, however, is not a necessarycondition is shown by the appearance of the same phenomenonin molecular crystals.S. B. Hendricksl7 has shown that thecrystals of p-chlorobromobenzene and an equimolar mixture ofp-dichloro- and dibromo-benzene are indistinguishable by X-raymethods, the molecule of the former possessing a statistical centreof symmetry showing that the intermolecular forces do not in thiscase distinguish between chlorine and bromine. The clearest casesof the conditions for interchangeable atoms are furnished by metals.Nearly all metals mix in the liquid state, and in the solid not onlydo similar metals, such as tungsten and molybdenum or antimonyand bismuth, form mixed crystals in all compositions, but theface-centred metals y-Mn, y:Fe, P-Co, Ni, Cu, Rh, Pd, Ag, Ir, Pt,Au, all form solid solutions in each other, and in most cases theseare unlimited.(Silver is somewhat exceptional, see p. 387.) Insuch a mixed crystal the atomic positions are rigidly fixed-in thiscase to the corners and face centres of a cube-but any kind ofatom can occupy any place. In all cases this is the state of affairsl3 Physical Rev., 1931, [ii], 38, 2234; A., 1932, 217.14 T. F. W. Barth and E. Posnjak, Z. Krist., 1932, 82, 325; A., 1932, 830.15 Ibid., p. 348; A., 1932, 830; S. Holgorsson, 2. anorg. Chem., 1932, 204,378, 382; A., 1932, 481, 485.16 Fortschr. Min., Krist. Petr., 1932, 17, 432; A., 1020.17 2. Krist., 1932, 84, 8 5 ; A., 216BERNAL AND CROWFOOT. 383just below the melting point, but in some a further regularityappears on slow cooling.For instance, in a mixture of the com-position of Cu,Au between 390" and the melting point a t 960" allthe atoms occupy indifferently the points of a face-centred lattice,but below 390" the gold atoms occupy the cube corners while thecopper occupy the cube face centres. Such structures are veryfrequent among metals and have been called " uberstrukturen,"not very happily translated as '' superstructures." Alloys withsuperstructures differ sharply from solid solutions in their mechanical,magnetic, and electric properties,ls which resemble those of puremetals : but this is merely an effect of regularity, and they shouldnot be considered as compounds unless we carefully distinguishbetween chemical compounds and crystal compounds. Our know-ledge of them has been greatly extended by the-detailed quantitativestudy of the iron-aluminium system by A.J. Bradley and A. H.Jay.19 Pure iron and FeAl both have atoms arranged in cubicbody-centred lattices : simply in iron, and with the centres occupiedby A1 atoms in FeAl (czesium chloride structure). The intermediatealloys quenched from higher temperatures showed up to 25 atomicyo of aluminium an irregular solid solution of a,luminium in iron,between 25 and 50% aluminium an irregular solid solution of ironin FeA1, i.e., with all the aluminium atoms not replaced by ironstill in the cube-centre positions. I n annealed specimens a furtherregular superstructure appeared in the composition 25 yo aluminium.This FeAl structure is also body centred but with a doubled cell of16 atoms, and with iron atoms regularly disposed in half the centresof the eight small cubes making up the unit cell.For compositionsbetween 18 and 40% aluminium this Structure persists, but X-rayintensities show that replacement of iron or aluminium atomstakes place in a complex but statistically orderly way, tending toapproximate to the simple Fe-A1 or FeAl-Fe solid solutions.It is clear that in such cases we have to do with a phenomenonanalogous to the melting of a crystal, referring, not to changebetween the regularity of atomic positions in a crystal and theirirregularity in a liquid, but to the regular and irregular arrange-ments of atoms inside a given crystal structure. Such regulararrangement must have a slightly lower potential energy, andmelts sharply with a correspondingly small latent heat of mixingof the atomic species.The melting point is depressed by additionof either constituent, as is that of a true compound. Undercooling,18 U. Dehlinger and R. Glocker, Ann. PhysiE, 1933, [v], 16, 100; A., 338;R. Glocker, i b d . , 1932, [v], 14, 40; A., 1932, 907.19 Iron and Steel Inst., May 1932; A., 1932, 567; Proc. Roy. Soc., 1932,[A], 136,210; A., 1932, 685; K. Schiifer, Naturwiss., 1933,21, 207; A., 455384 CRYSTALLOGIRAPHY.however, is easy, because the transformation requires interchangeof atomic places; thus slow diffusion hinders the formation ofsuperstructures. At low temperatures the irregular solid solutionis in the same state, as far as position of specific atoms, as a glass;it is not in equilibrium but cannot acquire it.The quantitativethermodynamic study of such systems has been carried out byW. L. Bragg,20 who has been able to show why the more compli-cated forms of superstructure such as Fe-Al cannot exist. In thisconnexion, it is worth recalling the recent studies in diffusion inthe crystalline state, particularly by G. von Hevesy and hisbut also by a great number of other authors.22 There is no spacehere for an adequate discussion, but it may be pointed out thatthe diffusion rate D is given by Qa2INh. e-QIRT, where N is Losch-midt’s number, 73 Planck’s constant, a is the shortest distancebetween exchangeable atoms, and Q is the characteristic heat ofatom interchange of the substance; Q varies from SO kg.-cal.permolecule for tungsten-molybdenum to 2.2 kg.-cal. for silver in silveriodide, and is itself a function of the characteristic temperature,or more simply, of the melting point of the substance = 3b2RTs,where b does not vary much from 2.In ionic compounds the anion diffusion is not much affected bythe nature of the cation, but the converse does not hold. Thevelocity of diffusion of silver is least in the fluoride, increases to itmaximum for the iodide and sulphide, and then falls to low valuesfor the metallic AgSb, AgSn, AgAu. The effect of crystal distortionor h e grain size raises the diffusion constant in some cases by asmuch as 40 times, but does not alter the interchange energy.The types of non-Daltonian compound that have been discussedcan all be brought into line by abandoning an absolute effectivedistinction between atoms or ions of different elementary species,and treating them as substitutional mixed crystals.They can beformulated in a Daltonian manner by such expressions as(NaJX1-5)(ClyBrl--y) or (Fel-,Al,),(Al,,Fe,), where x-and y are nolonger whole numbers, and where no particular value of x and ygives a compound sharply different from neighbouring values.There are, however, two other forms of crystalline phase whichstrain the definitions and laws still further, The first of these is2o Private communication.21 2. Elektrochem., 1933 [B], 490; A., 1000; Natumuiss., 1933, 21, 357;A., 669.z2 E. Owen and L.Pickup, Nature, 1932,130,201 ; A., 1932,989 ; C. Matano,Mem. 0011. Sci. Kyoto, 1932, [A], 15, 167, 351; A., 1932, 907; 1933, 218; N.Ageev and M. Zamotorin, Ann. Inst. Polytech. Leningrad, Sect. Math. Phys.Sci., 1928, 31, 15; A., 1932, 16; W. Jost, 2. physikal. Chem., 1933, [B], 21,168; A., 561BBRNAL AND OROWE’OOT. 385the interstitial compounds discussed in previous reports. Themetals of the transition groups can, without changing their struc-ture, admit small atoms such as hydrogen, boron, carbon, or nitrogeninto the interstices between them, and these can exist either statis-tically distributed or in definite positions. These may be calledadditional solid solutions or additional crystal compounds respect-ively, and may be formulated [FeIN, or [PclIH,,.Another non-Daltonian type is that called by Hiigg subtraction solid solution orcompound, and has been studied particularly in the systems Fe-S 23and Fe-Se,24 and also in Pe-0 studied by E. R. Jette and F. Foote.25The structures found are for FeO the simple sodium chloride struc-ture, for FeS the NiAs structure, for FeSe the lead oxide structureat high and NiAs at low temperatures. In each case, however, arange of solid solution on the iron-poor side exists, and in fact thestoicheiometric compound FeO or E’eS cannot be formed a t all, inspite of the assurance of elementary chemical text books! Thelattice shrinks as the iron content is reduced, which is consistentwith the theory that it consists of an ideal lattice FeO, FeS, or FeSe,with an increasing number of iron atoms removed, leaving spaceswhich are partially filled by adjustments of the other atoms.Suchsubtraction solid solutions may be formulated (Fe”Fe”’) l--zO, etc.They may be expected in metallic compounds, where, owing togreat difference of atomic type, substitution cannot take place, andin ionic compounds where the cation valency can vary. Subtrac-tion compounds may exist as well as subtraction solid solutions :the y structure of Cu,Zn,, CugA14, etc., can be derived by droppingtwo positions from a three-fold body-centred structure of 54 atoms.Another example is the C-sesquioxide type of Mn,O,. This is asubtraction compound derived from a two-fold fluorite structure,in which one out of every four anions is dropped out, with conse-quent slight displacements of all the ions.The necessity of considering these non-Daltonian or Berthollidecompounds (a term first introduced by Kurnakow) 26 in relationto the ordinary conceptions and notations of chemistry is becomingincreasingly urgent, as most metallic and sub-metallic (sulphide)systems are essentially of this type, as well as many ionic andmolecular compounds.A. 0hma11,~’ who has discussed this ques-tion from a structural and thermodynamic point of view, proposesz3 Nature, 1933, 131, 167 ; A., 342.24 R. Juza, and W. Biltz, 2. anorg. Chem., 1932, 205, 273; A., 1932, 563;G. Hiigg and A. I,. Kindstrom, 2. physikal. Chem., 1933, [B], 22, 453; A.,1111.25 J . Chern. Physics, 1933, 1, 29; A., 214.26 8.anorg. Chern., 1914, 88, 109.27 %. phyaikal. Chem., 1033, [ A ] , 165, 65.RE:P.-VOL. xxx. 386 CRYSTALLOGRAPHY.to distinguish them by the use of non-integral suffixes instead ofthe integral ones of Daltonian compounds : for instance, for &-brasshe would write c Z n ~ : ~ C u ~ : ~ ~ expressing merely the ranges of com-position. A more informative symbolism would refer to the numberof interchangeable places in a crystal cell, extra places being indicatedby +, empty places by -, but it is time the whole matter wastaken up by an authoritative committee.Transformation and Segregation in the Solid State.Closely related to the question of mixed crystals and solid solu-tion is that of the actual mechanism of separation of a homo-geneous crystal phase into two or more others.Recent work,particularly that of G. Sachs,28 R. F. Meh1,29 D. W. SmithFO M. L.FullerFl and H. HanemannF2 has shown that when an ore metalphase separates on annealing into two, the separation takes placealong certain Lines or planes of the original crystal, giving rise tothe Widmanstatten structure first observed in meteoric iron. I ngeneral, it is possible to show crystallographic reasons for themutual arrabgement on the interface between the new phases, thegeneral principle being to minimise the amount of atomic inter-change necessary. This phenomenon is by no means confined tometals : it lies a t the back of most of the pseudomorphism, parallelgrowths, mimetic and polysynthetic twinning that occurs in naturalminerals.The process has been experimentally studied in manyoxide and hydroxide systems. W. Bussem and F. Koberich33have shown that when hexagonal brucite, Mg( OH),, dehydrates tocubic periclase, MgO, the latter is oriented with its trigonal axisparallel to the original hexagonal axis. As, however, the corre-sponding planes are not quite identical, the periclase grows in astrained double refracting form which does not become isotropictill heated to 1150". Similar observations have been made byM. Deflandrew and H. Schwiersch36 on the transformation ofdiaspore AlO(0H) into corundum AI,O,. These reactions areessentially similar to the polymerisation of organic molecules in thesolid to be discussed later (p. 425).28 2. Metallk., 1932, 24, 241; A., 1932, 1196.29 R. F.Mehl and C. S. Barrett, Amer. I n s t . Min. Met. Eng. (Inst. Met.Div.), 1931, 93, 78; R. F. Mehl and 0. T. Muzhe, ibid., p. 123; R. F. Mehl,C. S. Barrett, and F. N. Rhines, ibid., 1932, 99, 203; A., 1932, 685.s1 M. L. Fuller and J. L. Rodda, ibid. (Preprint), 1933; A., 454.32 H. Hanemann and 0. Schroder, 2. Metcclllc., 1931,23,271; A., 1932,220.33 2. physikal. Chem., 1932, [B], 1'9, 310; a., 1932, 811.34 Bull. Soc. frarq. Mh., 1932, 55, 140; A., 1107.35 Ohem. Erde, 1933, 8, 252; A., 1020.C. H. Mathewson and D. W. Smith, ibid., 1932,99, 264BERNAL AND OROWFOOT. 387Crystal Topology.The reduction of the multiple observed crystal types to somecommon order based on topological relations between atoms andgroups is becoming an urgent question, because crystal chemistryis largely unintelligible owing to its not being systematised. P.Niggli and his school 36 continue their work in this field, but theirtreatment is too formal as yet to have acquired many followers.However, the work of F.Laves 37 on the classification of the silicatesindicates that the method has possibilities,CRYSTAL STRUCTURE.The number of papers on crystal structure in the last two yearsis still greater than ever before, and. it has become quite impossiblein the scale of these reports even to mention all the new structuresdescribed. The principle of selection has been to introduce onlysuch investigations as break new ground, or those of wider chemicalinterest. The main lines of classification remain unchanged-Metallic, Sub -metallic, Ionic, and Molecular.Metallic.There has been great activity in the field of fundamental theoryof metals, particularly of their electrical and magnetic properties,but most of this falls outside the scope of a chemical report.Thework of U. DehlingerF8 however, is of peculiar systematic interest.He points out that the properties of metals depend on the onehand on the number of free electrons, and on the other, on theinteraction between the remaining ionic cores. This interaction,which is practically nil for the alkali metals, reaches a maximumin the transition and coinage metals, where it is the determiningfactor for metallic properties. In particular, he links the coloursof copper, silver, and gold to the instability (multivalence) ofcopper and gold, compared to the peculiarly stable alkali-like ionof silver. I n metals of this type, miscibility in liquid and soliddepends far more on the specific nature of the core than on atomicradius.Silver, for instance, though it has almost exactly the sameatomic radius as gold, is far less miscible in the liquid or solid36 2. Krist., 1932, 83, 111; A., 1932, 986; ibid., 1933, 86, 121; A., 1001;P. Niggli and W. Nowacki, {bid., p. 65; A., 1001; W. Nowacki, ibid.,1932, 82, 355; A., 1932, 796; idem, ibid., 83, 97; A,, 1932, 986; H. Heesch,ibid., 1933, 84, 399; A., 450; R. Reinicke, ibid., 1932,82, 394, 419; A., 1932,796.3 7 Ibid., p. 1 ; A., 1932, 682 ; H. Heesch and 3’.Laves, ibid., 1933, 85, 443.as 2. Elektrochem., 1932,38, 148; A., 1932, 452; Metallwirt., 1932,11,223;A., 1932, 1193; 2. physikal. Chern., 1933, [B], 22, 45; A., 881; Naturwiss.,1933, 21, 607 ; A., 1005388 CRYSTALLOGRAPBY.state with iron, cobalt, copper, etc. Electrons in metals can fulfilthree functions. They can be free, belonging to the lattice as awhole ; if the number of such electrons differs from the numberof atoms a p-, y-, or &-structure results; or the electrons can belongpredominantly to atom pairs, as in gallium, tin, bismuth, and B-groupmetals in general; or they may be bound to the cores, as in mer-cury, thallium, and lead, which function in the metallic statemainly as Hg, Tl', and Pb" (cf. this vol., p. 120). I n general,atoms in compounds will not have the same charge as in the element.Fe" in iron becomes Fe in iron-zinc alloys, as shown by the alloytype and magnetic measurements.There can be resonance betweendifferent possible states, as in the body-centred compound[Cu++Pd+,-]=[Cu+Pd tt3-] , which thus has the required elec-tron : atom ratio 3 : 2. We are clearly at the beginning of atheoretical metal chemistry of great complexity and interest. Itsdifficulty can be judged by the fact that the Hume-Rothery rules,which are more and more recognised as the empirical basis of metalchemistry, have not as yet been satisfactorily explained, thoughmany attempts have been made, notably by U. Dehlinger,3* H.Yerlitz,39 and Laves.40 W. Hume-Rothery himself ha's shown,by a study of the alloys of copper with B-group metals, that theposition of the liquidus and solidus curve is determined by theelectron and not by the atom content of the solid solution.As already mentioned, high precision methods have been muchapplied recently, particularly to pure metals.Separate referenceswill not be given, as the measurements up to the middle of 1933have been collected by M. Ne~burger.~~ Our knowledge of a t leastone form of structure of the metallic elements is now nearly com-plete; we lack knowledge only of scandium, masurium, ten of therare-earth metals , radium, actinium , and protoactiniuin.Transition Metals and Their Interstitial Compounds.-Furtherwork on structures of transition metals at different temperatureshas shown that the allotropy of the type of iron is the rule ratherthan the exception. Besides the regular, face-centred cubic, body-centred cubic, and hexagonal close-packed, which seem to be theiiormel forms at high temperatures, several much more complicatedstructures have been found, often formed by electrolytic deposition.These stable low-temperature forms are clearly difficult to get fromthe melt, owing to slow diffusion, and many transition metals areprobably only known in their simple metastable forms.The39 Acta Comm. Univ. Tartu., 1932, No. 24, 3; A., 118.40 Nach. Ges. Wiss. Gcittingen, Mat,h.-phys. KI., 1932, 519; A . , 1932, 1232.4 1 Private communication.42 Z. KrisL., 1933, 86, 5'35HRRNAL AND CROWB001'. 389particular structures, thoqh complicated, arc of specid 8ta,bility,i i s shown by their appearance in more than one metal, as the 58atom per cell structure common to y-Cr 43 and u-Mn, the 20 atomper cell structure of 8-Mn iound also in Cu,Si, and the 8 atom percell u-W structure found also in Cr3Si.45 The last structure isone of some interest.A cubic cell of side 5.03 A. for tungstenand 4.55 8. for Cr3Si contains two atoms at the corners and bodycentre, and six others in pairs on each face with co-ordinatesof the type Oy+y$, O,+,i. This gives a pseudo-icosahedral 12-foldco-ordination of the second type of atom (chromium) round thefirst (silicon), which is actually closer than that of normal cub-octahedral close packing.The mechanism of allotropy has been studied in detail by U.Dehlinger 46 in the case of cobalt.The transformation from cubichexagonal close packing takes place reversibly in single crystalg .A similar transformation from cubic to tetragonal has been observedby L. Graf 47 to occur for the crystal compound AuCu. Much workhas been done on the solid solution and superstructures among thetransition metals : references only can be given.48Great advances have been made in the study of interstitialcompounds. K. Becker49 has given an interesting summary ofour knowledge of their physical properties. Most of them arc,exceedingly stable at high temperatures-vanadium carbide melts43 K. Sasaki and G. Sekito, Trans. Amer. Electrochem. SOC., 1931, 59, 437.44 M. C . Neuburger, Z . Krist., 1933, 85, 232; A., 665; H.Hartmann. 14'.45 See ref. 57.46 U. Dehlinger, E. Osswald, and H. Bumm, Z. Metallk., 1933, 25, 62; A . ,4 7 L. Graf, ibid., 1932, 24, 248; A , , 1932, 1196.Ebert, and 0. Breitschneider, ibid., p. 232.665.F. M. Jaeger and E. Rosenbohm, Proc. K. Akad. Wetensch. Amsterdam,1931, 34, 808; A., 1932, 14; W. Stenzel and J. Weerts, Siebert Festschr.,1031, 285; A., 1932, 221 ; S. Kaya and A. Kussinann, 8. Physik, 1931, 72,293; A , , 1932, 15; P. Wiest, ibid., 1932, '74, 325; A., 1932, 330; 1933, 81,121; A., 450; H. J. Seeman, ibid., 84, 557; A., 1000; G. D. Preston, PhiZ.Nug., 1932, [vii], 13, 419; A., 1932, 221 ; (Miss) H. D. Megaw, ibid., 14, 130;A., 1932, 800; R. W. Drier and H. L. Walker, ibid., 1933, [vii], 16, 294; A .,895; H. J. Seeman, 2. Metallk., 1932, 24, 299; A . , 344; 0. E. Zvjagintsevand B. K. Brunovski, 2. Krist., 1932, 83, 172; A., 1932, 1107; A. Claassenand W. G. Burgers, ibid., 1933, 86, 100; A., 1003; M. Lo Blanc and G.Wehner, Ann. Physik, 1932, [v], 14, 481; A., 1932, 989; J. 0. Linde, ibid.,15, 249; A., 115; 81. Le Blanc and W. Erler, ibid., 1933, [v], 16, 321; A.,455; H. Rohl, ibid., 18, 155; A., 1110; (Miss) M. L. V. Gayler, Iron and SteelInst., 1933, Sept. ; A., 1008; W. Koster and W. Schmidt, Arch. Eisenhiittenw.,1933-34, 7 , 121; A., 1007; E. Schiedt, Z . anorg. Chern., 1933, 212, 415; A . ,773.4O Physikal. Z., 1933, 34, 186; A., 343390 CRYSTALLOGRAPHY.above 4000"--aiid are also extremely hard, approaching if notexceeding diamond.More work50 has been done on the char-acteristic transformations of steel, which are typical examples ofWidmanstatten structure (see above, p. 386). A. Westgren andhis school have examined the more complex carbide, boride, andsilicide systems. I n the carbides, including high-speed steel carb-ides, the phases studied in order of increasing carbon content werethe following :-(1) Cubic Fe3W3C,51 96 metal atoms per cell. This is reallya double interstitial structure, being based on a tungstencarbide framework held together by iron atoms. Psomorphonswith this were Fe,W2C, Co3W3C, Ni3W3C, and Fe3M03C.This complexstructure is built from cubes and cuboctahedra of metal atomsat an average distance of 2.5 8. apart. Isomorphous with itare Ml123c6, Fe,lW2C6: Fe,,Mo2C,, and Cr21W.&&.(3) Orthorhombic Ye&.This is the familiar cementitestructure. Westgren 53 has shown that the structures previouslyput forward for this (Ann. Reports, 1931, 28,294) are incorrect,in that the carbon atoms should be placed as in tungstencarbide and other carbides inside a trigonal prism, height andside 2.66 A. Isomorphous with it are Mn3C and (MnFe),C(spiegeleisen) .This somewhat unstable compoundseems to be of Hagg's 12b, 6 type (see ibid., p. 293).This has not been furtherstudied and may have a more complicated formula.(2) Cubic CrZ3C6,Ii2 92 metal atoms per cell.(4) Hexagonal Ni3C.5*(5) Trigonal Cr,C3 and Mn,C,.(6) Hexagonal V2C and Mo,C, etc., of Hiigg's 12b, 6 type.(7) Rhombic Cr3C2.s5 This is the most interesting of thenew structures. The cell is 11-5 x 5.5 x 2.8 8., space groupDit - Pbnrn.All atoms lie in symmetry planes. The carbonatoms are placed at the centres of a distorted trigonal prismof chromium atoms, and form long zig-zag chains parallel tothe diad screw axis c, the G-C distance being 1.64 A. Herewe have no longer a true interstitial compound, but onebetween chromium and dehydrogenated paraffin chains.50 K. Honda and Z. Nishiyama, Sci. Rep. Tdholcu Imp. Univ., 1932,21,399;A., 12; H. J. Wiester, 2. Metallk., 1932, 24, 276; A., 119; H. Hanemann,Arch. Eisenhiittenw., 1931-32, 5, 621, 625; A., 1932, 796, 797.5 1 Jernk. Ann., 1933, 1-12, 14.52 Ibid., pp. 501, 1215, 1217.53 Ibid., 1932, 457.54 2. physikal.Chem., 1933, [B], 20, 361; A., 558.5 5 Svensk Kern. Tidskr., 1933, 45, 141; A., 1003BERNAL AND CROWFOOT. 391(8) Hexagonal WC, MoC, of Hiigg's Sb, 6 type. Here againeach carbon atom is at the centre of a trigonal prism of tungstenatoms.The structures of the borides of iron, cobalt, and nickel have beenstudied by T. Bjurstrom : 56 the most interesting is that of FeB,in which, as in Cr3C2, the non-metal atoms form a zig-zag chain witha B-B distance of 1-78 A. The average radius of the boron atomin these compounds is 0.93 A. The silicides of chromium, man-ganese, cobalt, and nickel have been studied by B. B~ren.~' Themost interesting is that of Cr3Si, already mentioned. It is nowwell established that the hydrogen dissolved in palladium existsin the form of protons, as the heat of solution exceeds by about20 kg.-cals.the heat of dissociation of molecular hydrogen. Thenature of the solutions and their conductivities have been studiedby F. Ifiiiger and G. Gehm,58 A. Coehn and K. S ~ e r l i n g , ~ ~ andG. RosenhalLG0 The conductivity is partially ionic, but the protonsonly carry a small fraction of an electronic charge, and must conse-quently be surrounded in the metal by an electronic cloud nearlyneutralising them. This property of dissociating hydrogen mole-cules is shown t o a less degree by nickel, and even less by othermetals of the eighth group and copper. It is clearly a necessarycondition for hydrogenating catalysis, but the catalytic efficiencyfalls off as the heat of combination with hydrogen increases ; hencenickel is a better hydrogenator than palladium.Electron Compounds.-The compounds between transitional andB (A) group metals obey Hume-Rothery's rules, and may appro-priately be called electron compounds, as their crystal structuredepends in the main merely on the ratio of electrons to the atomsthey contain (see above, p.388). Nearly fifty papers have in thepast two years been devoted to X-ray work on these compounds :it is only possible to mention a small selection.The range of alloy system investigated has been much increasedfrom the system TiAl 61 to AgLi. In the latter system H. Perlitz G2has shown that a true y-phase exists between 76.3 and 80.2 atomsyo of lithium. This corresponds to the formula Ag3Li,, and is amost important exception to Hume-Rothery 's rules, as no combin-ation of univalent silver and lithium can yield the electron ratio56 Arkiv Kern>;, Min., Geol., 1933, 11, [A], No.5 , 1; A., 669.5 7 Ibid., No. 10. 58 Ann. Physik, 1933, [v], 16, 174; A., 341.59 2. Physik, 1933, 83, 291; A., 768.60 Ann. Physik, 1933, [v], 18, 150; A., 1110.61 W. L. Fink, K. R. van Horn, and P. M. Budge, Amer. Inst. Min. Met.62 2. Krist., 1933, S6, 155; A., 1007.Eng. (Inst. Met. Div.), 1931, 93, 421 ; A., 1931, 676302 CEYSTALLOGRAPHY .21 : 1 3 . However, its this is the only examplc of a11 alkali y-strue-ture yet known, it is perhaps unfair to expect the same rules to hold.In the system manganese-zinc, studied by N. Parravano andV.Caglioti,a a y-structure is also found, corresponding to theformula Mnl.l~2.gZn~~.~lo.l. This obeys the rule satisfactorily ifwe assume that manganese yields no electrons as do iron andplatinum, or actually absorbs electrons from the zinc.The Mn,A1 and &,A1 systems studied by A. J. Bradley andP. Jones 64 are very complicated ; so is the FeSn system studiedby W. I?. Ehret and A. We~tgren.~~ PeSn at high temperaturesappears to have the NiAs structure. 0. Carlsson and G. Hagg66have confirmed the extremely complex superstructures found bythe author 6' in the CuSn system-the cell of the ?-phase contains500 atoms. S. Stenbeck fiinds,68 in a study of the AgHg and AuHgsystems, that the y-structure appears for a formula Ag,Hg,, exactlyobeying Hume-Rothery's rules. A.Olander 69 has analysed thevery interesting structure of AuCd with a tetragonal deformedczesium chloride structure which changes above 267" to a truec zsium chloride structure .A- and B-Group Metals and Their Compounds.-Until recentlyour knowledge of these was practically confined to the structuresof some of the elements. In the last two years this field has,however, attracted much attention, and we are at last in a positionto form a fairly general picture of its metal chemistry.This differs characteristically from that of transition metals andelectron compounds. Core interaction is much weaker, leadingto large interatomic distances, so that atomic size is an importantdeterminant of crystal type. On the other hand, where B-groupelements, particularly of the fifth and sixth group, are concerned,homopolar- bonds are common, while compounds between A- andB-group elements sometimes appear to be quasi-ionic : e.g., Na,+Pb4-or Mg; '-Se4-.Of the elements themselves, the most interesting new workis on the allotropy of beryllium,70 ~alciurn,7~ and probably of63 Mern.R. Accad. d'ItaZia, 1932, 3, [Chim.], 5 ; A., 1932, 1196; Atti R .Accad. Lincei, 1931, [vi], 14, 166; A., 1932, 116.G4 A. J. Bradley and P. Jones, Phil. iiag., 1931, [vii], 12, 1137 ; A., 1932,116.G6 Z. Krist., 1932, 83, 308; A., 1932, 1081.6 7 J. D. Bernal, Nature, 1928, 122, 54; A., 1928, 822.(i8 2. anorg. Chem., 1933, 214, 16; A., 1006.69 2. Krist., 1932, 83, 145; A., 1932, 986.J.Amer. Chem. Soc., 1933, 55, 1339; A , , 562.F. M. Jaeger and J. E. Zanstra, Proc. K. Akad. Wetensch. Amsterdam,7l F. Ebert, H. Hartmnnn, and H. Peisker, Z . anorg. Chenz., 1933,213, 126 ;1933, 36, 636; A., 1119.A . , 891BERNAL AND CROWE'OOT. 393barium.7z Yttrium 73 proves to have a hexagonal structure exactlyintermediate between strontium and zirconium ; consequently it isnot clear whether it should be classed as an A-group or a transitionalmetal. Enough rare-earth metals have been analysed to make thesystematic position of the group clear. The structures are allhexagonal, but sometimes also cubic close packed : La (37.0),Ce (34.1), Pr (36-3),'* Nd (36.1),'j and Er (36.9) have all the largeatomic volume characteristic of the A-group Na (39), or the heavyB-group metals Pb (30), and much larger than those of the transi-tion metals ranging from Zr (23) to Ni (10.9).(The figures inparentheses give atomic volume in Most interesting from thestructural point of view is gallium, which has been very carefullyredetermined by F. Laves.76 The structure is not, as believed byearlier workers, tetragonal, but orthorhombic. Each atom is sur-rounded by seven others : one at 2.45, and two each at 2-70, 2.75,and 2.79 A. Thus there exist in the solid actual Ga, molecules,the first examples discovered in metallic crystals. These moleculespersist in the liquid, as has been shown by X-ray diffraction," thepattern of which differs completely from that of monatomic liquidmercury. The molecular nature is particularly interesting, as itexplains the low melting point, 37".Hume-Rothery's rule (Ann.Reports, 1931, 28, 294) by which every B-group metal atomshould have 8 - N neighbours, where N is its group number,seems definitely broken. But only apparently so, for an idealtetragonal structure can be made up in which each atom has fiveneighbours, and this by a slight compression distorts into the actualstructure with seven neighbours, just as in the case of grey tin ofco-ordination number 4 passing into white tin with 6. The close-packed structures of indium and lead show that for larger atomsthe generalised metallic attraction is predominant. The transform-ations of arsenic and antimony from the amorphous into the glassystate have been followed by X-rays and electron diffraction.58Our knowledge of the compounds of A- and B-group metals weowe mainly to a series of papers by E.Zintl and his coliabor-72 E. Rinck, Compt. rend., 1931, 193, 1328; A . , 1932, 116.73 L. L. Quill, 2. anorg. Chem., 1932, 208, 59; A,, 1932, 1078.7 4 A. Rossi, Atti R. Accad. Lincei, 1932, [vi], 15, 298; A., 1932, 681.75 L. L. Quill, 2. anorg. Chern., 1932, 208, 273; A., 1932, 1192.Naturwiss., 1932, 20, 472; A., 1932, 797; 2. Krist., 1933, 84, 256.7 7 F. Sauerwald and W. Teske, 2. anorg. Chem., 1933, 210, 247; A., 341.7 8 E. G. Bomen and W. M. Jones, Phil. Mag., 1932, [vii], 13,1029; A., 1932,567; C. W. Stillwell and L. F. Audrieth, J . Amer. Chem. Soc., 1932, 54, 472;A., 1932, 479; G. R. Levi and D.Ghiron, Atti R. Accad. Liricei, 1933, [vi], 17,565 ; A., 1003 ; J. A. Prim, Nature, 1933,131, 760; A . , 667 ; A. Schulze andL. Graf, Metallwirt., 1933, 12, 19; A., 1109.N 394 CRYSTALLOGRAPHY.at or^,^^ and also to C. W. Stillwell,*O A. Baroni?1 and A. Rossi!,The numerous structures studied fall into four main types :(1) The simplest is the CsCl type, which appears for L a g ,LiT1, LiHg, MgAg, MgT1, CaT1, and SrT1.(2) NaTl structure. This is a very interesting new struc-ture. The cell is a double body-centred structure with 8 atomsof each kind. But each kind o€ atom occupies the positionsof a diamond lattice (000, $$$, etc. and 444, Qfg, etc.). Eachatom has therefore four neighbours of its own kind and four ofthe other kind, all a t the same distance.It is clear that sucha structure should only occur when the atomic radii of bothatoms are nearly alike. It is, in fact, found for LiZn, LiCd,LiGa, LiIn, NaIn, and NaTl, where the greatest radius differ-ence is 0.2 A. Thus in these compounds the alkali metalsappear to be compressed down to the size fixed by this B-groupmetal.(3) A number of alloys of formula AB,, where A is Ca, Sr,La, and Ce, and B is Sn, T1, or Pb, form cubic structures basedon a face-centred lattice of B in which one atom is replacedby A.(4) Mg,Ge, Mg,Sn, and Mg2Pb have the anti-CaF, structure.Other more complex structures have been found, as KBi,with the MgCu, structure, a framework of bismuth atoms inthe large holes of which are the potassium atoms ; and Na,Pb,LaAl,, and Li,Sn with large cells.Sub-metallic.Adamantine Compounds.-The structure of the hexaborides offormula B,M, where M = Ca, Sr, Ce, etc., mentioned in the lastreport as possessing an anomalous ion BE, has now been shownby F.Laves83 to be really adamantine, consisting of a covalent-linked B lattice with the metal atoms in the interstices. EachB atom has five B neighbours, one a t 1.66 and four at 1-76 A.,and four metal neighbours. Each metal atom has 24 B neighbours.This 4:24 co-ordination is the highest yet found. Essentiallysimilar compounds have been shown by A. Schleede and M. Well-'s 2. physikal. Chem., 1932, [B], 16, 183, 195, 206; A., 1932, 456; ibid.,1933, [B], 20, 245, 272; A,, 562; 2. anorg. Chem., 1933, 211, 113; A., 472;2.Elektrochent., 1933, 39, 81, 84, 86; A., 341.8o C. W. Stillwell and W. K. Robinson, J . Amer. Chem. SOC., 1933, 55, 127 ;A., 218.81 Atti R. Accad. Lincei, 1932, [vi], 16, 153; A., 18.83 2. pltysikal. Chem., 1933, [B], 22, 114; A., 891.Ibid., 1933, Id], 17, 182, 839; A., 558,1003BEaNAL AND CROWFOOT. 395mannS4 to be formed by the action of alkali metals on graphite.Compounds C,K and C,,K are formed with potassium atomsbetween each layer or between alternate layers of the normalgraphite structure. All kinds of carbon so far examined havebeen shown to possess a graphite structure. The differencesbetween the physical and chemical properties of different varietiesare mainly due to the crystal size, which may in the extreme caseof highly active carbon from benzene be only three layers thickand six condensed benzene rings wide, as U.Hofmann s5 has shown.Other work86 shows the importance of the edges of the crystallitesin absorption and catalytic activity.Homopolar Crystals of Fifth- and Sixth-group Elements.-Throughthe work of the last two years the gap represented by our ignoranceof the structures of this important group of “sulphide ores” hasbeen largely abolished. It is now possible to present a fairly orderedpicture of structure types, and to begin to construct a thio-chemistryequivalent to the oxygen chemistry of the silicates. Thio-structuresdiffer @om oxy-structures, owing to the larger size and greaterpolarisability of the S (Se, P, As, etc.) atoms, so that except foralkali compounds they are not ionic.The packing is generallycloser, owing to increased vitn der Waals forces, but this is offsetby the smaller co-ordination number of the metal atoms-generally4 or 6. The homopolar bonds formed are easily ionised, givingthe whole crystal a metallic or semi-metallic character. Sulphideores have long been studied by means of their reflecting power,and recently F. C. PhillipsS7 has shown that, just as in the caseof ionic crystals, the molecular refractivities calculated from thereflecting power are additive functions of those of the atoms. Thisshows the relative unimportance of structural considerations. Onthe whole, the structures correspond to those of simple oxides orhalides. Thio-salt complex groups, such as SbS,, do not exist inthe crystal as separate individuals, but fuse into chains or net-works. The general scheme can be set out as follows in order ofincreasing sulphur content, References are only given to struc-tures determined in the last two years.84 2.physilcab. Chem., 1932, [B], 18, 1; A., 1932, 903.a5 U. Hofmann and D. Wilm, ibid., p. 401; A., 1932, 1078; U. Hofmannand A. Frenzel, Kolloid-Z., 1932, 58, 7 ; A., 1932, 217; U. Hofmann and W.Lemcke, 8. anorg. Chem., 1932, 208, 194; A., 1932, 1212; S. B. Hendricks,2;. Krist., 1932, 83, 503; A., 12.86 E. Berl, K. Andress, L. Reinliardt, and W . Herbert, 2. physikal. ChenL.,1932,158,273; A., 1932, 217; E. Berl and R. Bemmann, ibid., 162, 71; A.,142; H. Amfelt, Arkiu Mat., Astron.Fysik, 1932, 23, [B], No. 2, 1 ; A., 1932,903; P. M. Wolf and N. Riehl, Angew. Chem., 1932, 45, 400; A., 1932, SO3 ;E. N. Greer and B. Topley, Nature, 1932, 129, 904; A., 1932, 795.87 Min. Mag., 1933, 23, 458396A,B. Ag2S, Cu,S.A,B,, (1). Zn3P2 8812 A.,( ) * A!&3Sb2CRYSTALLOGRAPHY.type also for Zn3As2, Cd3P,, Cd3As, ; cubic, CG =16 mol. per cell. -type,*9 also for Mg3Bi, ; hexagonal anti-Asesquioxide (La,03) structure (for the ionic Mg3P,structure, see p. 400).,.4,B,.AB, (1). Rock-salt type, 6-co-ordinated (Mg, Ca, Sr, Ba, Pb, 3h)(S, Se, Te) ; all the sulphides except those of Pb and Mnare ionic.(2). Types related to (1); HgS (cinnabar), GeS (SnS),91Ag(SbBi)S, (aramaoite).(3). Blende and wurtzite types, 4-co-ordinated (Be, Zn, Cd,Hg, Mn) (S, Se, Te), (Al, Ga) (N, P, As, Sb).It is inter-esting to note that MnS 92 exists in the three forms,rock-salt, blende, and wurtzite, showing that they donot differ fundamentally in nature.(4). Types related to blende ; CuPeS, 93 (chalcopyrite) , KFeS, ; 94Cu3VS4 (sulvanite). This structure, as determined byL. PaulingF5 is very anomalous, for the four metal amtomssurrounding the sulphur are all on one side of it.(5). Nickel arsenide type, ( M i , Fe, Co, Ni) (S, Se, Te, As, Sb).This is the typical form for monosulphides, etc., of theiron group, and is essentially metallic in character.The sulphides of Ni, Cu, Pd, Ptform structures of some complexity and unusual co-ordination. NiS (millerite) has a 5 : 5-co-ordination.CuS (covelline) is a bright blue mineral strongly pleo-chroic ; the structure, determined by I.Oftedal,96 helpsto explain these properties. It is really a compound ofCu,S and CuS, arranged in layers containing free sulphuratoms as well as S, molecules.The structure of PtS (cooperite), determined by F. A.s8 M. von Stackelberg and R. Paulus, 2. physikal. Chew,., 1933, [B], 22, 305 ;8* E. Zintl and E. Husemann, ibid., [B], 21, 138; A., 558.V. Caglioti and G. Roberti, Gazzetta, 1932, 62, 19; A., 1932, 326.91 W. H. Zachariasen, Physical Rev., 1932, [ii], 40, 917; A., 1932, 903.'I2 H. Schnaase, Naturwiss., 1932, 2Q, 640; A., 1932, 986; 2. physikal.C'hem., 1933, [B], 20, 89; A., 341.g3 L. Pauling and L. 0. Brockway, 8. Krist., 1932, 82, 188; A., 1932, 682.34 H.O'Daniel, ibicl., 1933, 86, 192; A , , 1106.95 L. Pauling and R. Hultgren, ibid., 1933, 84, 204; A., 215.9 6 lbicl., 1932, 83, 9; A., 1932, 987.c04s3,9° and Cu,FeS3 (bornite) ; cubic, a = 10 A.(6). Intermediate structures.A., 1003BERNAL AND CROWFOOT. 397B ~ I u I ~ s ~ w , ~ ~ is interesting as showing platinum sixr-rounded by four sulphur atoms in a plane instead of a!tetrahedron. (Pd, Pt, Ni)S with a more complicatedstructure has been called bruggite, being the first mineraldiscovered by X-ray analysis.There may be as many sulphides as oxides ofthis type, which is that of c03s4 and Ni,S,. Fe,S, doesnot, however, appear to be stableYg8 though (FeNi),S,(daubreelite) is found in rneteorite~.~~ Nore complexspinels studied by L.Passerini are MnCr,S,, CdCr,S,,and ZnCr,S,.The extremely important structure oE Sb,S,and Bi,S, has at last been analysed, by W. Hofmann.2The crystals are orthorhombic; a , b, c are 11-20, 11.28,3-83 B., containing four molecules, space group DBX-Pbnm, with all the atoms on the symmetry planes. Themost important feature is the existence of bands ofantimony and sulphur atoms linked by primary valencieslying along the c-axis in the (010) plane. This explainsa t the same time both the needle-shaped habit and theexcellent cleavage, and is most interesting among thesulphides as analogous to the asbestos-like structures inthe silicates.A3B4, Spinel type.AaB,, Stibnite type.ABC2. The structure of Sb,S3 is repeated in some thioantimonatesalso studied by Hofmann; CuSbS, and CuBiS, containSb-S chains: that of AgSbS, and TlAsS, is more com-plicated.Other thioantimonates.Bournonite, C U P ~ S ~ S , , ~ andstephanite, Ag,SbS4,5 have also been studied, but no com-plete structure has been arrived at.ABz, (1). Pyrite and marcasite type (Mn, Fe, Co, Ni, Ru, Rh, Pd,Here the non-metal atoms Os, Ir, Pt) (S,, SAs, As,).9 7 Min. Mag., 1932, 23, 188; A., 1933, 1014.9 8 See refs. 23 and 24 (p. 385).99 F. Heide, E. Herschkowitz, and E. Preuss, Ckem. Erde, 1932, 7, 483; A . ,1 L. Passerini and M. Baccaredde, Atti R. Accad. Lincefi, 1931, [vi], 14, 33 ;2 2. Krist., 1933, 86, 226; A., 1106; J. G . Albright, Phpical Rev., 1931,3 Fortschr. Min., Krist. Petr., 1932, 17, 422; A., 1029; Z.Krist., 19.33,4 I. Oftedal, ibid., 1932, 83, 167; A , 1932, 987.5 R, Salvia, Anal, Pis, Quim., 1932, 30, 416; A,, 1932, 904.1932, 1230.A . , 1932, 114; G. Natta and L. Passerini, ibid., p. 38; A., 1932, 114.[ii], 37, 468; A., 215.$4, 177; A., 214are arranged in molecules. H. Parker and W. White-house 13 have made a very exact Fourier analysis ofFeS,; Fe is found to carry 26 electrons, indicating theessentially non-ionic nature of this compound. Thedistance S-S is 2-14 A., and F e S is 2-26 A. Similarconclusions were reached by M. Biierger for FeAs,,which has the marcasite structure.(2). Cadmium iodide type (Ti, Zr, Sn, Pt) (S,, Se,, Te,).(3). Molybdenum sulphide type, MoS,, WS,. Both of theware layer lattices.It is clear that this survey of the sulphides is not complete, butit does at any rate show the main lines on which sulphur com-pounds are built.Ionic Crystals.The interest in ionic crystals continues unabated-over 250papers have been published on them alone-l-ut as now the mainstructural types have been fixed, interest has shifted to pointswith other than purely crystallographic interest, such as the rela-tions of oxides and hydroxides, the r61e of water, or the structureof complex ions.Halides, Oxides, and Hydroxides.-A new fact has been added tothe structural relations of the alkali halides by G.Wagner andL. Lippert's observation that cmium chloride changes its struc-ture into that of sodium chloride above 450". The small energydifference between these structures is thus experimentally confirmed.The most interesting halide structures which have been inves-tigatedrecently are those of cadmium, mercury, and lead.J. M. Bij-voet and W. Nieuwenkampg have shown that cadmium bromide,which, when sublimed has the cadmium chloride structure, whenprecipitated appears in the form of a statisticad mixture of thechloride and the iodide structure ; the interleavings cannot bemore than of the order of one or two layers thick. Incidentally,0. Hassel lo has shown that the actual structure of cadmium iodidehas a cell double that of the hitherto accepted structure. Completeanalyses have also been made of the structures PbF,,11 PbC1,,126 Phil. Mag., 1932, [vii], 14, 939; A., 13.7 2.Krist., 1932, 82, 165; A., 1932, 682.8 Z. physikab. Chem., 1933, [B], 21, 471; A., 768.B Chem. Weekblad, 1933, 30, 479; A., 892; 8. Krist., 1933, $6, 468.10 2. physikal. Chem., 1933, [B], 22, 333; A., 1004.11 J. A. A. Ketelaar, 2. Krist., 1932, 84, 62; A., 215,12 H. Brmkken, ibid., 83, 222; A., 1932, 1079BEEtNAL ANN CSOWE’OOT. 399PbBr2,13 PbClF,1* PbBrE’.15 The first three are isomorphous.They may be considered as layer lattices which have been distortedinto an approximation to a fluoride 8 : 4-co-ordination structure,owing to the large size and polarisability of the cation. Actually,the co-ordination is 9 : 4 or 5, each lead atom having three bromineatoms at 3.0 and three at 3.2 8., all on one side, while on theother it has one at 3.3 and two at 4-1 A.There is no evidence fora semi-molecular association of lead bromide. The chlorofluoridestructure is particularly interesting as a typical sandwich structure,layers of lead fluoride in fluorite arrangement alternating with a,lead chloride layer lattice of the tetragonal lead oxide type. Curi-ously enough, lead chlorofluoride has been known for a long timeas the mineral matlockite Pb,OCl,, but its X-ray pattern led toits identification, and a new analysis showed that fluorine hadbeen taken for oxygen. H. BraekkenI6 has shown that mercuric:bromide has a somewhat complicated layer structure also withoutmolecules.The trifiuoridesof Al, Fe, Co, Rh, Pd, etc., do not, like the trichlorides, form close-packed layer lattices,l’ but approximate to a cubic lattice whosecorners are occupied by the cations, while the anions occupy themid-point of an edge between two cations.The structure ofaluminium fluoride, analysed by J. A. A. Ketelaar,l* shows arhombohedra1 cell containing two such units distorted in an attemptto fill the large holes in the cell centres. These holes can be occupiedwith water in the zeolite fashion ; 2AlF3,H20 only loses its moleculeof water at a red heat.N. Wooster 19 has pointed out the essential isomorphism betweenferric fluoride and tungsten trioxide (see Ann. Reports, 1931, 28,296), with which rhenium trioxide 2o bas been shown to be iso-morphous. These two structures are the analogues in 6 : 240-ordination of the 4 : 2-co-ordination structures of beryllium fluorideand silica.As in them, the cation can be replaced by one of lowercharge, the extra charge being made up by an ion in the vacantspaces. ThusMore work has been done on AX, compounds.Sii+ 0, K+A13+Si4+0, NephelenePerovskite structureNa+Ta5+O313141516171820W. Nieuwenkamp and J. M. Bijvoet, 2. Krist., 1932, 84, 49; A., 215.Idem, ibid., 81, 469; A., 1932, 460; ibid., 1933, 86, 470.Ibid., 1932, 83, 157; A., 1932, 681.Ibid., 1932, 81, 152; A., 1932, 218.N. Wooster, ibid., 1932, 83, 36; A., 1932, 986.Ibid., 1933, 86, 119; A., 450. l9 Ibid., 84, 320; A., 214.K. Meisel, 2. anorg. Chern., 1932, 207, 121; A , , 1932, 903400 CRYSTALLOGRAPHY.A stranger case is that of the tungsten bronzes studied by W. F. deJong : 21 here the neutral WO, structure takes up a variable amountof metallic sodium in its interstices, giving highly coloured com-pounds whose physical properties would be worth investigating.Another form of substitution is the replacement of two cations byone of higher and one of lower charge.Thus, as Goldschmidtpredicted, BPO, and BAsO, 22 have essentially the cristobalitestructure of silica. This is particularly interesting, as in mostcompounds boron appears as 3-co-ordinated.The cristobalite structure, the cubic high-temperature form ofsilica, has been much studied. T. F. W. Barth23 has shown thatboth in cubic p-, stable above 240", and orthorhombic cc-cristobalite,the oxygen atoms do not lie, as was previously thought, on the linebetween two silicons, but definitely to one side.Thus, in all caseswhere oxygen is found co-ordinated to two cations, we find thatthe lines joining it to them are always inclined at about 140",indicating the effect of covalent binding tending towards a valencyangle of 109". I. Levin and E. Ott 24 first noticed that opal con-tains at ordinary temperatures p- or high-temperature cristobalite.As R. B. Sosman25 has shown, here can be no question of delayedinversion. J. Grieg,26 however, has pointed to the probable ex-planation. In opal, as in some glasses, the cristobalite is closelyadherent to an amorphous matrix. Now, at transition from p- toa-, there is normally a contraction of about 4% by volume. Either,therefore, the p-cristobalite inverts and breaks away from its matrix,as it has been observed to do in certain circumstances, or it failsto invert and continues to exist as the stable form under permanenttension (negative pressure) from the matrix. In the Report'ers'opinion this negative pressure is sufficient to account for theappearance of p-cristobalite rather than any other form of silica-all with smaller specific volumes-in the slow dehydration ofmassive silica gel, even if the temperature has never been abovethe @-+a! transformation point at 198-270".If this is so, it offerssome interesting possibilities for finding the negative-pressure formsof other materials. Through the work of M. von Stackelberg andR. Paulus 88 tcf- 89) we know some of the properties of the ions N3-, P3-.The compounds Mg&, Ca3N2, Mg,P2, etc., are ionic in character and"1 2.Krist., 1932, 81, 314; A., 1932, 450; ibid., 1932, 83, 496; A., 12.22 G. E. R. Schulze, Nafurwiss., 1933, 21, 562; A., 1004.za Amer. J . Sci., 1932, [v], 23, 350; 24, 97; A., 1932, 664, 903.I4 2. Krist., 1933, 85, 306; A., 666; cf. F. P. Dwyer and D. P. Mellor, ,I.PTOC. Roy. Soc. New South Wales, 1933, 66, 378; A., 691.2 5 J , Amer. Chem. Soe., 1932, 54, 3015; A., 1932, 903.2 8 i'bid., p. 2846; A., 1932, 903BERNAL AND CROWFOOT. 401have been shown to possess the anti-C-sesyuioxide type : i.e., theions Mg2 and N3- occupy respectively the positions of 02- and Mn3 t-in Mn203. From this, the radii of N3- and P3- are found to be 1.4and 1435 A,, very slightly larger than those of 0 2 - and S2- respect-ively.The free ions have, of course, far larger radii, but theirextreme compressibility almost compensates for this.We are now beginning to see the place of hydroxides in crystalchemistry. The strongly polar nature of the hydroxyl group (OH)leads in every case so far examined to a position unsymmetricallysurrounded by cations, and in particular to layer lattices. Thestructure of lithium hydroxide determined by T. Fhnst 27 proves tobe the simplest tetragonal layer lattice of the lead oxide type.Aluminium hydroxide (hydrargillite) is also a layer lattice, but thestructure determined by (Miss) H. D. Megaw 28 is far more com-plex. Zinc hydroxide in its stable form, analysed by R. Corey andR. W. G. Wyckoff ,29 is not a layer lattice, but each hydroxyl is sur-rounded by two zinc atoms on one side and two hydroxyl ions on theother, while each zinc has four hydroxyl neighbours.An unstableform exists with a layer lattice, as shown by W. Feitknecht,3O andthis forms sandwich structures with zinc halides, carbonates, andsulphates.In all cases there seems to be a definite attraction between twohydroxyl groups attached of different cations, as shown in theinset, and this attraction increases withthe polarising power of the cation, and@ consequently with the degree of polaris-ation of the hydroxyl group, as shown by@@ 0the decrease in inter-hydroxyl distances (pi.) in the seriesLi(0H) (3.61) Ca(OH), (3.36) Mg(OH), (3.22)Zn( OH), (2.83) Al( OH), (2.79) B( OH), (2.65)It is clear that the apparent neutrality of the hydroxyl groups toeach other in lithium hydroxide has been replaced by a powerfulattraction in the case of boric acid, which is only less than that ofhydrogen-bond-linked oxygens of acids, 2-55 a.(see Ann. Reports,1931, 28, 291). This new kind of bond, which may be called thedouble hydrogen bond, still requires theoretical explanation, as thesimple dipole effect should give increased repulsion rather thanattraction.2 7 Naturwiss., 1932, 20, 124; A . , 1932, 326; %. physlkal. C?LC?L, 1933, [ R ) ,a s 2. Kriist., in press.29 Ibid., 1933, $6, 8; A . , 1003.30 Ibid., 1932, 84, 173; A., 214; Helv. Chisti. Acta, l Y U t , 16, 429; A., Eitt4,20, 65; A., 341402 (~RYS‘I’AT,T,OCi RA P T I Y.A most fascinating field has been opened in the study of theconversion of hydroxides into oxides.The mechanism in magnesiumhydroxide has been already commented on (see p. 386). That ofthe trihydroxides of aluminium and iron is yet more complicatedand interesting. This has been investigated by a great number ofworkers, by X-ray and magnetic methods. Both metals form twoseries of hydroxides and oxides, usually known as a and y. Ineach series are a trihydroxide M(OH),, a monohydroxide MO(OH),and an oxide M20,. Relationships between them are shown in thescheme :a. Y.Structures } Hydrargillite (17.6) ::g,., { unknown y-Fe(OH),, structure unknownA10( OH) Diaspore (14.7) BohmiteFeO(0H) Goethite (17.25) Lepidocrocite (18.2) [Fe(OH),]A1,OZ Corundum (14.2) y-Al,O, (13.9)Fez03 HEematite (16.9) Ferromagnetic Fe203 (17.1) [Fe,O,]The figures in parentheses are the volumes occupied per oxygen atom.value for close-packed oxygen atoms would be 13.7 A.3.TheThe a-series is the more stable with smaller atomic volume; they-series, on the other hand, is the one usually formed from pre-cipitation, and is more chemically active. The structure of thear-trihydroxides has not been determined; it may be similar to thatof zinc hydroxide.That of the a-monohydroxides, diaspore andgoethite, has been established by M. Deflandre 34 (p. 386)3 31 andS. Goldsztaub.s2 It is essentially a structure of hexagonal close-packed oxygens with metal atoms in octahedral holes. Oxygensand hydroxyls are distinguished by being symmetrically or asjm-metrically surrounded by metal atoms.On dehydration, it passesinto a parallel aggregate of corundum or hEmatite crystals, a struc-ture similarly close packed.The y-trihydroxides, hydrargillite and y-Pe( OH),, are layer struc-tures (see above, p. 401); on dehydration, they pass without lossof crystal structure into the monohydroxides bohmite (the chiefmineral of bauxite ores) and lepidocrocite (rubinglimmer or commonrust).s The structure of these is not fully worked out, but it isclearly a layer lattice cleaving readily, probably across doublehydrogen bonds. Its birefringence is strongly negative, 0.57, while31 Cf. K. Takane, PTOC. Imp. A d . Tokyo, 1933, 9, 113; A., 558.32 Conapt. rend., 1932,195, 964; A., 13; W.H. Albrecht and E. Wedekind,2. anorg. Chem., 1931, 202, 205; A., 1932, 134.34 G. Schikorr, 2. anorg. Chem., 1933,212.33 ; A., 581 ; R. D. Williams andJ. Thewlis, Trane. B’araday SOC., 1931, 27, 767 ; A., 1932, 32 ; A. Girard andG. Ch&udron, C m p t . rend., 1933,196, 925; A., 581.S. Goldsztaub, Cornlpt. r e d . , 1931, 198, 633; A., 1931, 1390BERNAT> AN13 CROWFOOT. 403that of goethite is only 0.14 35 (P-386). Lepidocrocite is also formedby the oxidation of Fe(OH),,34 by the normal rusting of iron,35and by the hydrolysis of ferrites36 (MO,Fe,O,) into which it, butnot goethite?' can be reversibly reconverted. On dehydration,bohmite and lepidocrocite pass into a parallel aggregate of y-Al,O,or y-3Fe,03, cubic with a spinel structure 38 containing extra oxygenatoms (see Ann.Reports, 1931, 28, 299). The y-Fe203 structureholds these oxygen atoms so loosely that it loses them in a vacuumeven at 250" and is turned into magnetite, Fe304.36 In oxygen,on the other hand, it changes at about 600" into the stable cr-PezO,form. Lepidocrocite is also, as might be expected from its largervolume, unstable to pressure. 0. Baudisch39 ha8 shown that itmay be converted into a-Fe,03 by grinding at room temperature.Fe(OH), is also unstable, undergoing slow autoxidation even whenkept under anGrobic conditions.The most striking property distinguishing the M- and the y-seriesis ferromagnetism, which is characteristic of the latter. The occur-rence of this phenomenon in purely ionic compounds is of greattheoretical interest, particularly from the point of view of theirmagneto-optics, but so far no single-crystal work has been reported.Apart from the y-series, a number of ferrites show ferromagneticproperties.The nature of these ferrites has been studied, especiallyby H. Forestier 42 and S. Hil~ert.~O They are all of the mixed oxidetype, of general formula mM0,nFe,03, where M is Li,, Na,, K,, Be,Mg, Cu, Sr, Ba, Mn, Fe, Co, Ni, Cu, Zn, Pb, AI213, Sn,,,. Most of theseare ferromagnetic, some with exceptionally low Curie points : forLiFeO, and NaFeO, it lies below - 70". Some have special struc-tures, e.g., LiFeO, with a rock-salt structure with iron and lithiumindistinguishable (see p. 382), NaFeO, 41 with a layer lattice of FcO,'separated by large Na' ions, and a-BeFe,0,42 with a lepidocrocitcstructure with beryllium replacing two hydrogens.The greatmajority have spinel-like structures, but with extremely variablecomposition, Sr0,2Pe2O, showing the same structure as SrO,SFe,O,,35 J. Cates, Trans. Paraday Soc., 1933, 29, 817; A., 1022.36 A. Girard and G. Chaudron, Compt. rend., 1931, 193, 1418; A., 1932,37 A. Krause, Z. Czapska, and J. Stock, 2. anorg. Chem., 1932, 204, 385;38 33. B. Barlett, J . Amer. Ceram. SOC., 1932, 15, 361 ; A., 1932, 904; R.39 0. Baudisch and L. A. Welo, Naturwiss., 1933, 21, 593, 659; A., 1022,40 S. Hilpert and A. Wille, 2. physikal. Chem., 1932, [B], 18,291; A,, 1932,4 1 S. Goldsztaub, Compt. rend., 1933, 196, 280; A,, 215.4% 'El.Forestier and M. Galand, ibid., 1931, 108, 733; A., 1932, 30.133.A., 1932, 481.Brill, 2. Elelctrochem., 1932, 38, 669; A., 1932, 1004.1131.985; S . Hilpert and A. Lindner, ibid., 1933, [B], 22, 395; A., 1234404 CRYSTALLOGRAPHY.suggesting that the structure is simply a stabilised y-Fe,O, structure.It is interesting to note that the maximum of ferromagnetism is notreached for the true spinel M0,Fe,03, but for an iron-richer solidsolution 2M0,3Fe203. Others, such as 2Mt10,3Fe,O3, P-BeO,Fe2O3,and SnO,,Fe,O,, show the or-Fe203 structure, and are not ferro-magnetic. It seemed natural to conclude that the ferromagnetismwas a structural property of the y-series, but there are severalindications that this is not the case,39~40 notably the work of G.Kurdjumov and his collaborators,43 who prepared ferromagneticmFe203 in an alloy.It is more probable that here, as in metallicferromagnetics, structure plays a secondary rble, the internalelectronic state of the ion being primary. In ionic compounds ferro-magnetism seems to be associated with the state of iron functioningas ferrous acid, probably the net acid y-FeO,' H'. It is noticeablethat the oxidation product of Fe(OH), is ferromagnetic or not,according as the medium is alkaline or acid.34The oxide-hydroxide relations of cobalt 44 are essentially similarto those of iron; those of manganese are more complicated, assolid transformations between Mn2+, Mn3+, and Mn4 t are all possible.The transformationsMn(QH), -+ MnO(0H) --+ MnO,Pyrochroite Manganite Pyrolusite (Polianite)a-AlO(OH) TiO,Pseudo-structure R u t h structureWOH),Structureall take place without any essential change in crystal structure.45The products in all these cases are, however, microcrystalline, asthey are formed below their recrystallising temperatures : theyconsequently possess large surface and high chemical reactivity.These oxide-hydroxide transformations are beginning to becomeintelligible through crystal a.nalysis.It should now be possible toattack the related problem of the colloidal hydroxides (e.g., thoseof aluminium, iron, manganese, etc.) so important for physical,bio- and geo-chemistry.Mixed Oxides.- Of mixed oxides ttic spinel structures havealready been discussed (see p.382). The more complex structuresof the pyrochlorite group have also been much studied. Thisstructure bids fair to rival that of spinels in its stability and varietyof composition. Its essential formula is [A]:-*[B]:O7 (see Ann.Reports, 1931, 28, 299), with a cubic cell a z 10.3 8., but the struc-43 V. Danilov, G. Kurdjumov, E. Pluschnik, and T. Stellezky, Naturwim.,44 H. B. Weiser and W. 0. Milligan, ,7. Physical Chem., 1932, 36, 7 2 9 ; A ,,4s cf. Vaux, private coirimunication.1933, 21, 177; A., 340.1932, 485BERNAL AND CROWFOOT. 405ture can take up at least one atom of oxygen and four of waterwithout swelling noticeably. The essential feature of the structureis the 6-co-ordinated Sb5r ion (which may be replaced by Nb5+,Ta5', Ti4+ etc.). As L.Pauling 46 has pointed out, antimony neverappears in any less than 6-co-ordination, and among acid-formingions we should distinguish sharply between the 3-co-ordination of(B)3+, C4 t, N5+, the 4-co-ordination of (Al)3+, Si4+, (Ge)4 r, P5+, As5 r ,S6+, Se6+, CVf, Br74, V5+, Cr6+, Mn7+ and the 6-co-ordination ofSn4I-, Sb5+, Te6b, I7'-, Fe34-, Ti4'-, (Zr)4+, Nb"'-, Ta5i-, Mo6+, W6+,Re7+. (The elements in parentheses are borderline cases; Mo6+and W6+ are anomalous.) G. Natta and M. Baccaredda47 haveshown that not only some but all the calcium atoms in Ca,Sb,O,can be replaced by $Sb,O, so that the structure of Sb,04 reallybelongs to the pyrochlorite group, and should be written asSbi+Sbi+O,, just as Fe304 is Fe2+Fei+04. Further work on thisgroup has been done by 3'.Machatschki 48 and others.498iZicutes.-Since the main outlines of silicate chemistry are known(Ann. Reports, 1931,28,301), most of the work done in the past twoyears has been concerned with anomalous or very complex silicates,of mineralogical rather than of chemical interest. The exceptionis the determination by W. H. Taylor of the structure of sanidiiic,(K,Na)AlSi,O,, which gives the key to that of the petrologicallyessential group the felspars, and thus closes the last important gapin our knowledge of silicate types. As anticipated, it proves to bebased on a somewhat collapsed (Si3A108)1- framework, in theinterstices of which are found the alkali atoms. Among the morecomplicated silicates studied are those of hemimorphite 52 andbertranditeB by T.Ito and J. West, euclase by J. Biscoe andB. Warren,54 and zunyite by L. Pa~ling.5~ All are hydroxy-silicates, but in no case are the hydroxyl ions attached to silicon, butto zinc, beryllium, or aluminium. The silicate grouping is different46 J . Amer. Chem. SOC., 1933, 55, 1895; A., 664.47 Atti R. Accad. Lincei, 1932, [vi], 15, 389; A., 1932, 869; 2. XXst., 1933,85, 271.48 Chem. Erde, 1932, '7, 56; A., 1932, 596; Zentr. Min. Geol., 1932, 33; A . ,1932, 1079; F. Machatschki and 0. Zedlitz, 2. Krist., 1932, 82, 72; A,, 1932,682.49 0. Zedlitz, 2. Krist., 1932, 81, 253; A., 1932, 450; E. Reuning, Chenz.Erde, 1933,8, 186; A,, 1029; G. Aminoff with R. Blix, K . Stienska VetenskapsAlcad. Hand., 1933, [iii], 11, No.4, 3, 14; A., 482.50 E. Schiebold, Ergeb. exakt. NaturwisS., 1932, 11, 352; A., 369; P. Niggliand E. Brrtndenberger, 2. Krist., 1932,82,210; A., 1933,683 ; F. Machatschki,Geol. 3%~. Ptirh:, 1932, 54, 447; A., 1137.61 2. Krist., 1933, 85, 425; A., 892.52 Ibid., 1932, 83, 1; A., 1932, 987.54 Ibid., 1933, 86, 292; A., 1107. 55 Ibid., 84, 442 ; A., 451.53 Ibid., p. 364; A., 13406 CRYSTALLOGRAPHY.in each case. Hemimorphite, Zn4(O€€)2Si,0,,H20, is a hydroushydroxy-disilicate. Bertrandite, 2Be0,2Be(OH)Si0,,Si04, containsthree tetrahedron chains and independent orthosilicate ions.Euclase, AlBe( OH)SiO,, is an orthosilicate containing tetrahedralBeO,( OH) and octahedral AlO,( OH) chains. Each hydroxyl isshared by a beryllium and an aluminium neighbour.Zunyite,[Al]~~02[(OH)~],,[~]'02Si,0,,C1, contains both 6- and 4-co-ordinatedaluminium, as well as a complex (Si5016)12- group and interstitialchlorine.Useful structural work has been done on clays and clay minerals ;66the different varieties are all based on the same Al(OH)Si20, sheets,arranged somewhat differently relatively to each other. The zeoliteshave been much studied, and the structure types proposed by Paul-ing (Ann. Reports, 1931, 28, 302), based on various (SiAl)O, frame-works, have been confirmed and extended. The natrolite grouphas received the most attention. W. H. Taylor and W. Jackson 57have determined the structure of natrolite, Na2A12Si,0,0,2H,0, andedingtonite, BaA12Si,010,4H20, the related meta-natro1ite.M Ash-c r ~ f t i t e , ~ ~ mesolite,sO and thomsonite have been studied by otherworkers.E. Schiebold 62 ha5 determined the structure of scapolite,E. Kozu that of cancrenite, both zeolites containing C03. T. F. W.Barth 64 has studied the cuboctahedral structure of the sodalitegroup, and in conjunction with E. Posnjak, an interesting set ofstructures derived from cristobalite (see p. 400) by replacing halfthe Si4+ ions successively by A13+ and Ca2+ without essential changeof structure, the Naf ions needed for neutrality falling into theinterstices of the structure, thus :(SiSi0,)O CristobaliteNa+(AlSiO,)l- darnegeiteNa2+ ( CaSi0,)2-From a physical standpoint, the last structure is an orthosilicatewith separate SiO, groups ; crystallographically it has a SiCaO,s6 J.W. Gruner,Z. Km'st., 1932,83,75, 394; A., 1932, 987; 1933,45; 1933,85, 345; A., 892; W. W. Jackson and J. West, ibid., p. 160; A., 451 ; U.Hofmfmnn, K. Endell, and D. Wilm, i b d . , 86, 340; see also ( 6 ) , p. 380.Ibid., 1933, 86, 53; A., 1004; W. H. Taylor, C. A. Meek, and W. W.Jackson, ibid., 84, 373; A., 451; F. Halla and E. Mehl, ibid., 1932, 83, 140;A., 1932, 987.68 M. H. Hey (with F. A. Bannister), Min. Mag., 1932, 23, 243; A., 141.6o M. H. Hey, ibid., p. 421.61 J. Wyart, Compt. rend., 1931,193, 666; A., 1932, 12.E. Sohiebold and G. Seumel, 2. K~ist., 1932, 81, 110; A., 1932, 218.63 S. KBzu and K. TakanB, Proc. Imp. Aoad, Tokyo, 1933,9, 56,105; A , , 658.6 4 2. KGt., 1932, 83,405; A., 13; Amer.Min., 1932,17,466; A., 482.Idem, ibid., 1933,23, 305; A., 588BERNAL AND CROWFOOT. 407framework.change of crystal type is shown in the classical felspar series :Albite Na(AlSi,O,)- --+ Anorthite Ca(Al,Si,0J2-Complex Ions.-The work of the last two years, represented byover 100 papers, now makes it possible to review the types of com-plex ion in a wider and more rational scheme. In the first place, itis necessary to have a physical definition of a complex ion. Themost general is : a charged group of atoms (closed or open) for eachatom of which the forces attaching it to the group are greater thanthe forces attaching it to other atoms. Applying Pauling’s rules tothe simplest case of an ion of the type Bb+X:-, we can see that thiswill always be a definite complex ion if b/n>x/2, for here the partialvalency joining B to X, being greater than half the charge of X, noother valency can equal it.For oxy-ions the rule is simplest :an atom will always give a stable oxy-ion when its maximum valencyis greater than its co-ordination number ; e.g., C4+03, P5+04, 17+06.For fluoride ions, twice the valency must be greater than the co-ordination number; e.g., B3+F4, Si4+F6. This rule gives a lowerlimit to the stability. Ions with weaker internal forces can, how-ever, exist in f avourable circumstances, i.e. , in the presence of largeions such as K’ or Cl’. This is the case with co-ordinated ions ofthe type Ni(NO,):-. Oxy-ions of this type, however, inevitablyhydrolyse in water, e.g., Sn0;- -+ Sn(OH)i-.The ions for whichb/n = x/2 tend to polymerise by sharing X atoms. Thus (B3+03)3-The same transition of physical properties without9 Bb ‘7 ‘0 0’ \O,B,O/ = (BO,)’ ions,75 (Si4+O4)” gives0 .(Si0J2- chains, etc. ; AlFi- and WOE- give framework structures.Where there are homopolar forces inside the ion, as in organic andquasi-organic ions, e.g., the dithionate ion, the question of ionicstability merges into that of molecular stability.Ions can be divided into mononuclear and polynuclear types.The former are of the general form B&[n = (0),1,2,3,4,6] ; thelatter consist of mononuclear ions bound together either (a) throughX atoms (generally 0), or (b) by covalent bonds between B atoms.The (a) type are represented by pyro- and per-ions and by thesilicates; ( b ) by most organic ions.We limit by conventioninorganic polynuclear ions to those which are fairly compact andsymmetrical in structure. The simple mononuclear ions are sym-metrical and linear, triangular, tetrahedral, or octahedral in shape.From them we can derive asymmetrical ions, either by removing Xatoms, leaving a reduced ion with electron pairs (Zachariasen’408 CRYSTALLOGRAPHY.principle, see Ann. Reports, 1931, 28, 297), or by choosing differentkinds of X atoms, by substituting, e.g., F, S, or N for 0 in an oxy-ion.65 The substitution of hydrogen to form a hypo-ion is anextreme case of this. The substitution of hydroxyl for oxygen orsimple addition of hydrogen ion, leading to an acid ion or acid, doesnot change the form, or, in general, the symmetry of the ion, owingt o the partial ionisation of the hydrogen in a hydrogen bond.Two types of anomaly are apparent : those such as SeBrg- andVPl-, where the co-ordination is normal but the valency too low,and those such as W0:- and OsOj, where the valency is normal butthe co-ordination too low.Crystal-structural investigations have been able to throw con-siderable light on the nature of co-ordinated ions.A. Ferrari andC. Colla6' have measured a series of salts containing the ions[Co(N0,),l3- and [Ni(N0,)6]4-. All form cubic face-centred cells,a z 10 A., determined by the complex ions, various numbers ofK,NH,,Ba,Pb, etc., filling the interstices of the structure. Completeanalysis was not attempted, but the published dimensions of thecell exclude, in the Reporters' opinion, the possibility of an electro-static co-ordination Co+ -ONf and are only compatible with a0 covalent Co-NO co-ordination.The square planar configuration of the PtX, and PdX, ions, whichcan be considered as derived by removing two X atoms from theoctahedral PtX6 ion, has been fully confirmed by structure analyseson single crystals.R. M. Bozorth and L. Pauling68 find planarconfiguration in Mg2PtC1,,7H20, while E. G. Cox in a series ofpapers 69 demonstrates it in [Pd(NH,),]Cl,,H,O, [Pt(NH,)4]C1,,H,0,Pt(NH,)2C12,70 and most clearly of all in the " green salt of Magnus "[Pt(NH3)4][PtC14],71 where the + and - planar ions are piledalternately along the tetrad axis of the crystal.-065 H.Seifert, 2. Krist., 1932, 81, 396; A., 1932, 449.66 L. Sieg, Z. anorg. Cherct., 1932, 207, 93; A., 1932, 903; J. L. Hoard andB. N. Dickinson, 2. Krist., 1933, 84, 436; A., 451.6 7 A. Ferrari and C. Colla, Atti R. Accad. Lincei, 1931, [vi], 14, 435; A.,1932, 483; ibid., 1933, [vi], 17, 390; A., 666; cf. L. W. Strock, 2. Krist.,1933, 86, 42, 186, 270; A., 1004, 1106, 1107.68 Physical Rev., 1932, [ii], 39, 537; A., 1107.6D H. D. K. Drew, F. W. Pinkard, W. Wardlaw, and (in part) E. G. Cox, J.,1932, 988, 1004; A., 1932, 562 ; E. G. Cox, ibid., p. 1912; A., 1932, 797;E. G. Cox and G. H. Preston, ibid., 1933, 1089; A., 1040.70 F. Rosenblatt and A. Schleede, Nuturwws., 1933, 21, 178; A., 342.71 E. Hertel and K.Schneider, 2. anorg. Chern., 1931, 202, 77; A., 1932,315; E. G. Cox, F. W. Pinkard, 14'. Wardlaw, and G. EI. Preston, J., 1932,2527BERNAL AN11 C!ROWFOOT. 409The stxucturo of K[Ag(CIN),] 1 1 : ~ been studied b)y ?J. L. Hoard ''and the Ag(CN), groups found to be linear: this may be regardedas the last stage in removing substituents from a BX, complex. Asimilar linear arrangement has been found by H. M. Powell 73 in thedialkylthallium ions.We are not yet in a position to classify the polynuclear ions.Those of the pyro-type sharing oxygen (or fluorine) are fairly familiarthrough the silicates. The most valuable addition is the threeco-ordinated chain ion studied by W. H. Zachariasen 74 in calciummetaborate (see p. 407). The homopolar type is represented by thedithionate ion.G. V. Helvig,75 G, Hiigg,76 and M. L. Huggins 77are now all agreed that the ion consists of two pyramidal SO, groupsjoined point to point through the sulphur atoms in octahedralconfiguration. The hypophosphate ion, O,P-POi-, is probablyisomorphou~.~~ I n connection with the heteropoly phospho-tungstate ion P(W,O,,);- mentioned in the Report (p. lOS), it shouldbe noted that, if the structure proposed is correct, it would requireus to modify our conceptions of the laws of ionic structure, for one02- ion has as neighbours three W6+ ions and one P5+ ion, giving itan equivalent charge of 49 electrons instead of its usual two, whilcother oxygen ions receive a charge of only one electron.The Salts.-Apart from information on the nature of ions, muchwork of crystal structural interest has been done on the salts.InABX, compounds with flat ions, optical anisotropy is an importantproperty, and has been much studied.79 S. B. HendricksgO haspublished a full account of the changes of structure in ammoniumnitrate due to ion rotation (see Ann. Reports, 1931,28, 291). V. M.Goldschmidt 81 has shown a long series of isomorphisms between72 2. Krist., 1933, 84, 231; A., 215.73 H. M. Powell and (Miss) D. M. Crowfoot, Nature, 1932, 130, 131; A . ,1932, 904.74 Proc. Nut. Acad. Sci., 1931, 17, 617; A., 1932, 114; UT. H. Zachariasenand G. E. Ziegler, 2. Krist., 1932, 83, 354; A., 13.75 2. Krist., 1932, 83, 485; A., 13.76 2. physikal. Chem., 1932, [B], 18, 327; A., 1932, 986; 2.Kyist., 1932,83, 265; A., 1932, 1079. 7 7 Ibid., 1933, 86, 384.7 8 I?. Nylen and 0. Stelling, 2. unorg. Chem., 1933, 212, 169; A., 664.79 W. H. Zachariasen, Physical Rev., 1931, [ii], 37, 1693; A., 211; H. Bras-seur, 2. Krist., 1932, 83, 493; A., 10; K. s. Krishnan avd A. C. Dasgupta,Indian J. Physics, 1933,17, 49; A., 1109 ; P. Le ROUX, Compt. rend., 1933,196,394; A., 336; S. B. Hendricbs, W. E. Deming, and M. E. Jefferson, 2. KTist.,1933, 85, 143; A., 448.S. B. Hendricks, E. Posnjak, and F. C. Kracek, J . Amer. Chem. Soc.,1932, 54, 2766; A., 1932, 986; J. M. Bijvoet and J. A. A. Ketelaar, ibid.,p. 625; A,, 1932, 445.V. M. Goldschmidt and H. Hauptmann, Nach. CTes. Wiss. Giittingen,1932, 53; A , , 1932, 1079410 CRYSTALLOGRAPHY.BOZ-- and (30;:- ions.Thus ScBO,, InBO,, and YtBO, have thestructure of MgCO,, CaCO, (calcite) and NaNO,, while LaB0, hasthe structure of CaCO, (aragonite), BaCO,, and KNO,.Nm important type structures recently established are :ABX,, NH,N0,82 and RbNO, ;83 A,BX,, Li,S0,,S4 Na,SO,(Ag,SO,) ; 85A,BX,, Li,P04,86 which shows great resemblance to Mg,SiO,(olivine). Much work has been done on the apatite series 87 (seeAnn. Reports, 1931,28, 297) ; the mineral constituent of bone andteeth is shown to be hydroxy-apatite, Ca3( OH)Ca,(PO,),. Thehydroxy-ion is very tightly held, and cannot be removed below1400". The structures of the beautiful copper carbonate mineralsazurite, Cu,( OH),(CO,),, and malachite, Cu,( OH),C03, have beenstudied by M.Brasseur.88 The structures are complicated, and aretrue hydroxy-carbonates, not sandwich structures, CuCO,Cu( OH),,AS usually written. Other polyionic salts, such as northupiteYsgNa3MgCI(C0,),, and it8 related salts, have been studied with somesuccess.Our knowledge of water of crystallisation has been much increasedby the determination of the structures of Be(H,O),SO, andNi(H,O),SO, by C. A. Beevers and H. Lip~on.~Q Ineachcase the watermolecules are co-ordinated to the cation, the co-ordination numberbeing determined by the ionic size. The water molecules act asdistributors of the ionic charge, each touching two oxygens ofdifferent SO, groups carrying an electrostatic valency of S and & inthe beryllium and nickel sulphates respectively, each sulphateoxygen being accordingly touched by two water molecules belongingto different beryllium ions or three belonging to different nickel ions,so that in each case Pauling's rules are exactly obeyed.OtherC. D. West, J . Amer. Chem. SOC., 1932, 54, 2256; A., 1932, 903.83 L. Pauling and J. Sherman, 2. Krist., 1933, 84, 213; A., 215.84 J. G. Albright, ibid., 1932, 84, 150; A,, 215.85 W. H. Zachariasen and G. E. Ziegler, ibid., 1932, 81, 92; A., 1932, 218;8 6 F. Zambonini and F. Laves, ibid., 83, 26; A., 1932, 986.W. H. Zachariasen and K. Herrmann, ibid., 82, 161; A., 1932, 681.S. B. Hendricks, W. L. Hill, K. D. Jacob, and M. E. Jefferson, Ind. Eng.Uhem., 1931, 23, 1413; A., 1932, 359; S. B. Hendricks, M. E. Jefferson, andV. M.Mosley, 2. Krist., 1932, 81, 352; A,, 1932, 494; M. A. Bredig, H. H.Franck, and H. Fiildner, 2. EEektrochem., 1932, 38, 158; A., 1932, 469; A.Schleede, W. Schmidt, and H. Kindt, ibid., p. 633; A., 1932,1217 ; G. Tromel,2. physikal. Chem., 1932,158,422; A., 1932,481; R. Klement and G. Tromel,2. physiol. Chem., 1932,213,263 ; A., 290 ; M. A. Bredig, ibid., 1933, 216, 239.2. Krist., 1932, 82, 111, 195; A., 1932, 682.H. Shida and T. Watanabe, Compt. rend., 1931,193,1421 ; A., 1932, 218 ;B. Gossner and I. Koch, 2. Kriet., 1931, 80,455; A., 1932, 11; T. Watanabe,Sci. Papers Inst. Phys. Chem. Res. Tokyo, 1933, 21, 40; A., 802.2. Krist., 1932, 82, 297; A., 1932, 681; ibid., 83, 123; A., 1932, 986;R .B. Corey and R. W. G. Wyckoff, ibid., 1933,84,477 ; A., 451BERNAL AND CROWFOOT.41 1hydrates atudied are NaBr,2H20, Na1,2H20,91 3CdS0,,8H,0,92Zn,,( OH) (AsO,),, 12H,0,Y3 and finally Na 7F( PO,), ,19H20?* whichcrystal has a cubic cell, a = 27.86 8., containing 40 molecules ofthe above composition.Moleczclar Crystals.Although X-ray methods were from the beginning used to studymolecular and, in particular, organic crystals, it is only in the lastfew years that they have begun to yield information of positive valueto chemists. This was because, owing to the intrinsic difficulties oforganic crystal analysis, the results obtained were in the case ofsimpler compounds too rough to do more than confirm the chemicalevidence, while compounds where information of any kind wouldbe of use were generally too complicated in structure to yield tothe X-ray methods of the time.Now, however, improvements in crystal analysis technique and amore extensive knowledge of the nature of molecules and of intra-and inter-molecular forces have changed all this, and resultsalready achieved justify the claim of crystal analysis as an importantand ultimately essential adjunct to structural chernistr~.~~ Everystage in the crystal analysis of a molecular compound may be of valuefrom the chemical standpoint.These stages may be summarisedas follows :(1) X-Ray photograph of the crystal, either as powder or singlecrystal. This is a very powerful means of identification. Theangular positions and the intensities of the reflexions are equivalentto so many characteristics, each of the value of a melting point orother single physical determination.It has been used, for instance,in showing the substantial identity of adenine hydrochloride with apreparation of vitamin-B, (p. 424).(2) Determination of cell size. This, in conjunction with a know-ledge of the density, furnishes the most accurate method of absolutemolecular-weight determination. The limitations of this methoddepend only on that of the density, which is difficult with very smallor cracked crystals. The weight obtained is in general some simpleexact multiple, e.g., 2, 4, 6, 8, of the true molecular weight, but thismultiple can be found at the next stage.(3) Determination of space group and molecular symmetry. This9 1 W. A. Wooster, Nature, 1932, 130, 698; A., 12.92 L.Egartner, F. Halle, and E. Schwarz, 2. Krist., 1932, 83, 422; A,, 12.93 J. Drugman and M. H. Hey (with I?. A. Bannister), Min. Mug., 1932, 23,94 E. W. Neuman, 2. Krist., 1933, $6, 298; A., 1107.95 J. D. Bernal, " Crystal Structure of Complex Organic Compounds " (a175; A., 1932, 1015.lecture), Metropolitan Vickers. Manchester412 CRYSTALLOURAPHY.is effected from X-ra,y photographs combined with a determinationof polarity by piezo- 96 or pyro-electric methods,97 or by etchfigures. The first use to which this information can be put is thedetermination of the maximum molecular size. The minimum sizecannot be determined, as all the molecules in a cell may not occupycrystallographically equivalent positions, even when no polymeris-ation has taken place.Thus, it is possible to distinguish betweenpure substances and molecular compounds, as in the cases of calci-ferol and quinhydrone (see pp. 423, 420). The actual molecularsymmetry is, of course, of stereochemical importance in many cases,but here again the crystal analysis only gives a minimum symmetry,as molecules in general will not be able to pack so as to exhibit theirhighest symmetry (cf. benzene, p. 419).For this theX-ray evidence may be supplemented by evidence from physicalproperties-habit, cleavage, etc.-but most importantly by a know-ledge of the optical or magnetic anisotropy of the crystal. In amolecular crystal where the intramolecular forces are strong and theintermolecular forces weak-as in nearly all organic crystals-themolecule behaves in the crystal very much as in a gas, and con-sequently the anisotropy of the crystal is a mere product of that ofthe individual molecules and of their mutual orientations.Now the optical (or magnetic) anisotropy of a molecule is closelyrelated to its shape-long molecules are positive, flat moleculesnegative, and round or compact molecules quasi-isotropic. Ifmolecules pack parallel in a crystal, the optical character of thecrystal is that of the molecules.If, as is more rarely the case, theiraxes lie in a plane, the character is inverted, i.e., long moleculesarranged at right angles give negative birefringence. If there is noselected common direction, quasi-isotropy results for every type ofmolecule.It is nearly always possible from such indications to find thedisposition of the molecules in the crystals, and hence their size andshape, even when the molecular structure is quite unknown, as hasbeen done, e.g., for vitamins B, and C (ascorbic acid, see p.424).Once the size and shape of the molecule are known, many questionswith respect to its structure can be answered. The method isessentially a negative one. Of proposed structures, some can beexcluded as having inadmissible shapes, though at this stage it isimpossible to be sure that a given admissible structure is the correctone. It was in this may that the new sterol formula was firstdeduced (see p. 423). The greatest difficulties are encountered(4) General arrangement of molecules in the cell.96 A.Hettich and H. Steinmetz, 2. Physik, 1932, 76, 688; A., 1932, 904.9 7 A. J. P. Martin, Min. Mag., 1931, 22, 519BfiRNAL AND OROWBOOT. 41 3where (u) the molecules are geometrically and optically quasi-isotropic, (b) intermolecular forces are large, as in poly-alcohols andacids, (c) the molecule is not rigid but protean in shape, owing to theexistence of internal rotations. Thus the sugars and simple vegetableacids, where all these disadvantages occur together in varying degree ,have not yielded anything significant to X-ray analysis, comparedto that yielded by aromatic or paraffinoid types.Once the rough positionsof molecules in a crystal are known, the exact positions of the atomscan be found by a laborious process of trial and error, so as to givea good agreement between observed and calculated intensities ofreflexion.(6) Determination of electronic densities (Fourier analysis) (seeAnn.Reports, 1929, 26, 280). Starting from approximate atomicpositions, the phases of the diffracted beams can be found. Thisgives their complex amplitude, and hence by Fourier analysis acomplete picture (usually a two-dimensional projection) o€ theelectronic distribution can be obtained (see Figs. 1 and 2). Promthis, interatomic distances and valency angles can be read off, andthese final values are, owing to the method of successive approxim-ation used, independent of any original assumptions about thestructure of the molecule.This chain of operations is divided clearly into two parts, vix.,(1)-(4) and (5) and (6). The first is straightforward measurement,involving little calculation, and for any crystal species may takebetween a day and a week.The second calls for much experienceand intuition, refined measurements of intensity, and heavy corn-putations , running into several thousands of operations for onecrystal, and the time taken must be reckoned by months. It isplain, therefore, that far fewer crystals can be determined by thiscomplete intensive method than by rapid preliminary methods moreadapted to extensive surveys. The two sides of the work are inter-dependent. Extensive surveys suggest which crystals will yieldthe most useful results if intensively analysed, while the results ofintensive analysis provide data on intra- and inter-moleculardistances useful in building up rough models of crystal structures,particularly in stage (4).The chemical implications of these two aspects are also different.Intensive analysis provides fundamental information on stereo-chemistry and the nature of the chemical bond.It is one of thechief means of transformation from the classical qualitative, topo-logical, chemistry of the nineteenth century to the quantum-mechanical, metrical chemistry of the present day. The extensivemethod is more closely connected with the practical development of(5) Determinution of atomic positions414 CRYSTALLOGRBPHY.chemistry in analysis and synthesis, particularly in dealing withcomplex biochemical problems.Here the chemist and the crystallo-grapher must work in intimate collaboration. A large number ofrelated substances and derivatives are prepared and examined, andfrom their crystallographic relations more can be discovered thanfrom an elaborate analysis of any one of them. There is no doubtthat, if chemists were more aware of the possibilities of thesemethods, there mould result an enormous saving of time and effort.The work isa continuation of that of (Sir) W. H. Bragg a t the Royal Institution,chiefly by J. M. Robertson and (Mrs.) K. Lonsdale. Including someearlier work, particularly that of It. W. G. Wyckoff, thc only crystalsfor which complete analyses are available are shown in Tablc I.TABLE I.Complete analyses of organic crystals are still rare.Cryst,al.Urea 1 6 .....................Thiourea l7 ...............HexamethylbenzeneDwene 35 ..................Naphthalene .........Anthracene 48 ............p-Diphenylbenzene 44Distances, A.T -7 ---Intramolecular.. Interm olecu l a .C-N 1-33 NH, ...O= 3.20c-0 1-18C--N 1.36 NHS.. . HZN 3.85C=S 1.64 NH, . , . S= 3.45C-C arom. 1-42C-C aliph.-arom. 1.48C-C arom. 1.41 CH, ... H,C 3.87C-C a1iph.-arom. 1-47C-C arom. 1-41, CH . . . HC 3.60C-C arom. 1-41 CI-I . . . HC 3-77C--C between rings 1.48 CH ... HC 3-92With each crystal type are given the intramolecular distances whichcan be derived from it. These, together with the 1-54 A. distancederived for the aliphatic bond from diamond, and the value 1-42 A.for graphite, represent all the determinations of interatomic distancesjustifiable by X-ray measurement.Values are quoted by N. V.Sidgwick 08 for many other bonds, but, except where they depend onspectral data for symmetrical diatomic molecules, they are of thenature of predictions only. This is particularly the case for distancesderived from electron diffraction, except for the simplest moleculesof the BX, type, where they are least liable to interpretative error.A summary of this work is given by L. Bru.'The table also contains values of intermolecular distances, butthese have not the same precise significance. For most organicmolecular compounds covered by neon-like electronic orbitals (otherforms occur only in compounds containing sulphur, chlorine, etc.)the distance between atoms in neighbouring molecules is determinedprimarily by residual (van der Waals) forces leading to a common# * " The Covalent Link in Chemistry," p.81 et 8q.(Mre.) H. Lonsdale, PTOC. Roy. SOL, 1929, 1a8, [A], 494; A., 1929, 750BERN& BND CROWFOOT. 415value of about 36---4.0~\. This is only slightly greater than thedistance 3.41 8. between the hydrogen-less molecular sheets ofgraphite. The presence of hydrogen atoms does not affect thedistance much, as they are embedded in the electronic cloud of theatoms to which they are bound. If, however, they are part of a localdipole, as in hydroxy- or amino-groups, the interatomic distancemay be reduced to 3-3.2,&., or in the extreme case of acids, ahydrogen bond may reduce it to 2-55 8.All other work on molecular compounds is either of crystalphysical interest, as showing how known molecules pack in a crystal,or of direct analytic interest, as leading t o a structural formulawhich still remains to be proved by chemical methods.The resultsmay be classified under : (1) Simple molecular compounds, includ-ing not only inorganic molecules, but organic compounds withcompact molecules (pentaerythritol derivatives, etc.) . (2) Long-chain molecules. (3) Ring molecules, simple, multiple, and con-densed.Simple MoZecukcr CrystaZs.-Much work has been done on hydridcstructures.l Although the essential picture is that of close packing(Ann. Reports, 1931, 28, 303), yet in almost very case there existlow-temperature forms with lower symmetry, usually tetragonal.The transitions in the solid are clearly due to dipole rotation, as hasbeen shown by C.S. Hitchcock,2 by the appearance of high dielectricconstants a t the transition temperatures, the only temperature atwhich the molecules are neither freely rotating nor fixed, and can thusrespond to the electric field. Similar effects have long been knownfor ice and the alcohols, but curiously enough are not found for solidammonia. M. Ruhemann has shown that a-nitrogen (above 35" K.)is hexagonal close packed, as against the cubic close packing ofa-nitrogen. The molecular polyhalides of the types BX,, BX,,BX,, BX, have been studied in the solid by X-raysY4 and in the liquidand vapour both by X-ray and electron 6 methods.Electron1 G. Natta, Mem. R. Accad. d'Italia, 1931, 2, [Chim., 31, 31; A., 1932, 326;Gmzettcc, 1933, 63, 425; A., 1003; B. Ruliemann and F. Simon, 8. physikal.Chem., 1932, [B], 15, 389; A., 1932, 325.2 C. S. Hitchcock and C. P. Smyth, J . Arner. Chem. SOC., 1933, 55, 1296,1530; A., 448, 663; K. Clusius, 2. Elektrocliem., 1933, 39, 598; A., 1000.2. Physik., 1932, 76, 368; A., 1932, 796.4 0. Hassel and H. Kringstad, Teknisk Ukebiad, 1931, 78, 230; A., 1932,326; Z . physikal. Chem., 1932, [B], 15, 274; A., 1932, 325.5 R. W. James, Physikal. Z., 1932, 33, 737; A., 12; L. Bewilop, ibid.,p. 658; A., 1932, 1078; W. van der Grinten, ibid., p. 769; A., 12; ibid., 34,609; A., 1003.6 L. 0. Brockway and L. Pauling, Proc.Nat. Acad. Sci,, 1933,19,68; A.,341; H. Braune and 5. Kmke, 2. physikal. Chem., 1933, [BJ, 21, 297; A.,658; Natumk., 1933, 21, 349; A., 658416 CRYSTALLOGRAPHY,methods have also been used by K. Wierl and L. Bru for a wholeset of simple molecules, hydrocarbons, halides, ethers, etc. In allthese cases the structure determination depends on the interpre-tation of a small number of diffuse diffraction rings; it can only beof value in the simplest cases, and then only for the measurement ofrough interatomic distances and valency angles. A very interestingapplication of electron diffraction is, however, shown by A. Stockand R. Wierl’s 8 work on the compound B,N,H,, which shows thesame pattern as benzene, and should therefore be formulated as in(I).Other such cyclic inorganic compounds have been investigated.H H(PNCI,), and (PNCl,), have been shown by F. M. Jaeger to havethe structures like (11) ; S,N, and H,S,N, lo are apparently morecompact. Most interesting, however, is further light on the struc-ture of sulphur.ll E. Hertel l2 has shown, by examination of thetrigonal molecular compounds CHI3,3S,, As13,3S,, that the S,molecules not only have a plane of symmetry but must be piledabove each other in the direction of the c-axis at a mutual distanceof 4.44 A. This is only compatible with a disc-shaped molecule ofexternal diameter z 6.6 and thickness z 4.4 d. The most probableconfiguration is therefore a cyclic molecule derived from a cubeflattened into a puckered octagon.A more careful analysis of thiscompound would give more information on the nature of the sulphurmolecule than a study of the element itself.G. Wagner and G. Dengel l3 have investigated the tetrahalides ofpcntaerythritol, finding simple cells with halogen atoms lying inplanes, but the Reporters have reason to believe that the real struc-tures are more complicated. The only other apparent exception foundby X-ray analysis to the tetrahedral carbon atom is that of dibenzyl-idenepentaerythritol (see Ann. Reports, 1929, 26,304), the moleculeR. W i d , Ann. Physik, 1932, [v], 13, 453; A., 1932, 670; J. Hengsten-berg and L. Bru, Anal. Pis, Quim., 1932,30,341; A , , 1932, 798; L. Bru, ibid.,p. 483; A., 1932, 1078; ibid., 1933, 31, 115; A., 890.* A.Stock and R. Wierl, 2. anorg. Chem., 1931, 203, 228; A., 1932, 215.9 F. M. Jaeger and J. Beintema, Proc. K . Akad. Wetensch. Amsterdam,10 F. M. Jaeger and J. E. Zanstra, ibid., 1931, 34, 782; A., 1932, 797.11 J. J. Trillat and H. Forestier, BUZZ. SOC. chim., 1932, [iv], 51, 248l a 2. phy8ikal. CImnnt., 1931, [B], 15, 51; A., 1932, 114.l3 Ibid., 1932, [B], 16, 382; A., 1932, 564.1932, 35, 756; A., 116.A.,1932, 452BERNAL AND CROWFOOT. 417of which was found by F. M. Jaeger to have three diad axes of~ymrnetry.1~ One of the Reporters l5 has recently found that thisis incorrect, as the crystal is really trigonal, not hexagonal, and inthe molecule consequently only one diad was of symmetry, whichis fully compatible with the tetrahedral carbon atom.R.W. G. Wyckoff and R. B. Corey have continued their work onurea and its derivatives thiourea l7 and methylurea,l8 and havefurnished valuable precise interatomic measurements (see Table I)and also P curves for carbon, nitrogen, and oxygen. F. V. Lenel19has continued his work on the polypeptides. Glutathione, thebasic respiratory peptide, has been examined by J. D. Bernal,20who has shown that its three amino-acid units are arranged in astraight chain.E. G. Cox 21 is making progress in the field of the sugars, but sofar the difficulties have been too great to put forward any completestructure.Long-chain Compounds.-A. Miiller 22 has shown that the cyclo-paraffins C,H,, (n = 12-30) are essentially similar to the normalparafhs in structure, the flexible ring being drawn out under theinfluence of residual forces into a double chain with a small ringinvolving 1-3 carbon atoms at each end. The cross-section of thestraight part of the ring is 37.5 almost exactly double that of afree paraffin, vix., 18.3 k2.More remarkable are the apparentlygreater intermolecular forces, especially in the low-carbon members :cycZododecane has, for instance, a density of 0-945 compared to 0.77for dodecane. The rotation of the carbon chains of the paraffins inthe solid state has been well established by his further work,23showing that it is complicated by other forms of polymorphism forhydrocarbons higher than C24. These rotational forms have beenstudied optically for hydrocarbons a’nd alcohols by N. Yannaquis 24and J.D. Berna1,25 where they appear as uniaxial hexagonal crystals.l4 See also “ Handbuch v. Stereochemie.”l5 2. Krist., (in preparation).16 Ibid., 1931, 81, 102; A., 1932, 218.1 7 Ibid., p. 386; A., 1932, 451.Ibid., 1932, 81, 224; A., 1932, 451.20 Biochem. J., 1932, 26, 75; A., 1932, 636.21 J., 1932,138,2535; A., 1932, 326, 1192; 2. Kmkt., 1932,84,45; A., 216;E. G. Cox and T. H. Goodwin, J., 1932, 1844; A., 1932, 798; 2. Krist., 1933,85, 462; A., 892; G. W. McCrea, Proc. Roy. SOC. Edin., 1932, 51, 190; A.,1932, 327.Ibid., 1933, 85, 132; A., 451.22 Helv. Chim. Acta, 1933, 16, 155; A., 267.23 PTOC. Roy. SOC., 1933, [A], 138, 514; Nature, 1932, 129, 436; A., 451.24 Compt. rend., 1933, 196, 784; A., 451.26 Nature, 1932, 129, 870; A., 1932, 798; Z.Krist., 1932, 83, 153; A.,1932, 987.REP.-VOL. XXX. 4is CRYSTALLO GltAPH Y .E. Ott 26 has shown that, in general, mixtures of hydrocarbons aidof alcohols show mixed crystals in the solid with occasional 1 : 1molecular compounds. Separation can, however, be effected bydistillation in high vacuum, as shown by (Miss) H. S. Gilchrist and(Miss) B. Karlik.27 Other hydrocarbons, together with ketonesand secondary alcohols, have been studied by S. H. Piper.28 Poly-morphism seems a general phenomenon both in these, and in theacids, esters, and salts studied by F. D. la Tour 29 and 0thers.3~S c a l e0 I 2 3 4 SA,L--FIG. 1.-Projection along b axis showing mutual relation of molecules,Most of this work has been done by measurement of long spacingsonly.W-e badly need a complete Fourier analysis of typicalhydrocarbons, acids, etc.Ring CompozcncZg.-Ring compounds, particularly polyphenyl and26 E. O t t and D. A. Wilson, Scieme, 1933, 78, 16; A., 896; E. O t t and F. B.2 7 J . , 1932, 1992; A., 1931, 927.28 8. H. Piper, A. C. Chibnall, S . J. Hopkins, A. Yollard, J. A. B. Smith,and E. F. Williams, Biochem. J., 1931, 25,, 2072; a., 1932, 250.29 Compt. rend., 1932, 194, 622; A., 1932, 327; Ann. Physique, 1932, [x]18, 199; A., 1932, 1192; J. Thibaud and F. D. la Tour, J . Chim. phyeique,1932, 29, 153; A., 1932, 904.P. E. Verkade and J. Coops, Proc. K . Akad. Wetensch. Amsterdam, 1933,36, 76; A., 559; D. Coster and A. v. d. Ziel, ibid., 1932, 35, 91; A., 1932,682; P.A. Thieasen and E. Ehrlich, 2. physikal. Chem., 1932, [B], 19, 299;A., 116; ibid., 1933,165,453; A., 1004; J. W. C. Phillips and S . A. Mumford,.I.. 1932, 898; A., 1932,451: T.Malliin.J., 1031, 2796; A . , 1932, 326; Trans.b'araday Soc., 1933, 29, 977 ; d., 1107.Slttgle, J . Physical Chem., 1933, 37, 257; A., 342BERNAL AND CROWFOOT. 41 9condensed systems, have been found extremely amenable to crystalanalysis, owing to the rigidity and large anisotropy of their molecules.The optical and magnetic anisotropy has been specially studied byS. B. Hendri~ks,~~ I. Obreimov,32 and K. S. Krishnan.33 Benzeneitself has a structure, determined by E. G. Cox?* which is somewhatanomalous, being intermediate between that of a simple molecularScaleI 1 ...I._..I 1 I I A . 0 1 2 , ! 4 5FIG. 2.-Projection along c axis showing mutual relation of molecules.Each contour line represents two electrons per A.2.crystal-the molecule centres are on a pseudo-cubic close-packedlattice-and that of a typical aromatic structure, as the moleculesare not piled like plates but inclined almost at right angles to eachother, giving a strongly positively birefringent crystal. The ring is31 S. B. Hendricks and M. E. Jefferson, J . Opt. SOC. Amer., 1933, 23, 299 ;A., 1104.92 I. Obreimov and A. Prichotjko, Physikal. 2. 8owiet Union, 1932, 1, 203;A., 1932, 674.33 Nature, 1932, 130, 605; A., 10.3* Proc. Roy. Soc., 1932, [A], 135, 491; A., 1932, 461420 CRYSTALLOGRAF'RY.almost certainly planar, but the best evidence for this comes fromthe study of more complex aromatic structures, in particular durene,1 : 2 : 4 : 5-tetramethylbenzene, for which a complete Fourieranalysis was carried out by J.M. Robertsona5 (see p. 413 andFigs, 1 and 2). It is interesting to note that the mutual repulsionof methyl groups in the o-positions made them deviate from theradial orientation. Other benzene derivatives have been lessthoroughly studied. p-Dichlorobenzene has been already mentioned(p. 382). The plane of symmetry reported for m-dinitrobenzene 36(see Ann. Reports, 1931,28,306) has now been shown by piezoelectricmeasurement 96 not to exist. Of particular interest are the struc-tures of quinone, corrected from optical and magnetic measurementsby K.S. Krishnan and S. Banerjee?' and that of quinhydronedetermined by 0. R. Foz and J. Palacios.38 The cell is monoclinic,space group Cih - PZ,/n, but in it there is only one molecule ofquinone, and one of quinol, and these must be derivable from eachother by a symmetry axis. The only coizclusion is that here we havcto do with no molecular compound, but with a new type of chainpolymeride held together by hydrogen bonds :This would explain both the peculiar oxidation-reduction propertiesof the substance, and its extreme optical pleochroism. The crystalsare practically black for vibration directions along the macro-molecule, even when transparent and showing Newton's colours forthe other vibration. In vapour and solution, ring polymerisationmay take the place of chain polymerisation.The most complex benzene derivative studied is the mineralmellite, A12C,,0,2,18H,0, which has a large tetragonal cell with16 m0lecules.3~ E.Halmoy and 0. Hassel's work in cyclohexanederivatives 40 shows iairly conclusively that here the atoms do notlie in a flat ring.The essential nature of polyphenyl structures has now beenestablished. Diphenyl itself has been measured by J. Dhar 41 andby L. W. Pickett 42 by the X-ray method, and the molecular orient-35 Proc. Roy. SOC., 1933, [ A ] , 141, 594; A., 1108.36 S. B. Hendricks and G. E. Hilbert, J. Amer. Chern. SOC., 1931, 53, 4280 ;37 Nature, 1933, 131, 653; A., 557; W. A. Caspari, Proc. Roy. Soc., 1932,A., 1932, 327.A , 136, 82; A,, 1932, 682.Anal.Pis. Quim., 1932, 30, 431; A., 1932, 904.:it) T. F. W. Barlb and C. J. Ksanda, Amer. Min., 1933, 18, 8; A., '328.I" %. physikal. Chena., 1932, [B], 16, 234; 17, 268; A., 1932, 451, 708.J1 Indian J. Physics, 1932, 7,43; A., 1932, 798.42 Nature, 1933, 131, 613; A., 451BERNAL AND CROWFOOT. 421mtion determined optically by K. S. Krishnan.43 L. W. Yickett44has also published a Fourier analysis of p-diphenylbenzene (cf.Hertel) ,45 and preliminary results for pdiphenyldiphenyl40 (bis-diphenylyl). In these compounds the structure found shows thatthe benzene rings are flat hexagons which in any one molecule lie inthe same plane. The value for the distance C-C between the ringswas 1.48 pi., ie., intermediate between those observed in aromaticand in aliphatic carbon atoms.The implication is that the Ph-Yhbond does not easily admit of rotation. This is further confirmed bythe analysis of s-triphenylbenzene by (Mrs.) K. L0nsdale,~6 wherea11 rings lie in a plane.In a similar way J. M. Robertson 47 has established the structureof the basic condensed ring systems naphthalene 47 and anthracenq48and with J. Iball 49 extended this study to chrysene and 1 : 2 : 5 : 6-(1 ibenzanthracene. Condensed ring systems form flat rigid mole-cules, which with increasing number of rings tend to pack more andmore in parallel sheets approximating to the structure of graphite.Stilbene, tolane, and o-azotoluene have been described by 1LI.Pra~ad.~* The crystals all show similarities with those of azo-benzene. The -N=N- and -CEC- are equivalent in lengthwithin the errors of experimental measurement, -C=C- is some-what longer.To this group also belong the aromatic disulphidesand diselenides measured by L. Egartner, F. Halla, and R. Scha-~ h e r l . ~ l The series included diphenyl and dibenzyl disulphides anddiselenides, and dibenzoyl disulphide. The parameters of the sulphurand the selenium atoms were determined, and their radii from thedistance between the two centres found to be in both cases 1.19 A,,rather larger than the values S = 1-04, Se = 1-13 A. found in theelements. This is probably the half length of the S-S (Se-Se) singlebond. The straight character of dibenzyl derivatives is maintainedfor similar multiple-linked phenyl groups such as dibenzylidene-benzidine or diethyl pp’-xylylidenebisaminocinnamate, as shown by43 Nature, 1932, 130, 313; K.S. Krishnan, B. C. Guha, and S. Banerjes,Phil. Trans., 1933, [ A ] , 231, 2315; A., 340.44 Proc. Roy. SOC., 1933, [A], 142, 333; A., 1235.4 5 E. Hertel and G. H. Romer, 8. physikal. Chem., 1933, [B], 21, 292 ; A.,46 Nature, 1934, 133, 67.4 7 Proc. Roy. SOC., 1933, [A], 142, 674.48 8. Krist., 1933, 84, 321; A., 216; Proc. Roy. SOC., 1933, [A], 140, 79;49 Nature, 1933, 132, 750.50 Phil. Mag., 1933, [vii], 16, 639; A., 1107; M. Prasad and K. V. Desai,ibid., 1932, [vii], 13, 600; A., 327.5 1 2. phyaikal. Ohm., 1932, [B], 18, 189; A., 1932, 987.666.A., 558422 CRYSTALLOGRAPHY.thc work of' J . 1). H w i d 52 and K. Herrinmn53 on tho crystalstructures of substances giving rise to liquid crystals.E.Hertel* and his co-workers have examined a number of thesubstitution products of trinitrobenzene, and also several molecularcompounds.55 It seems a pity that in this work the optical proper-ties of the crystals generally are not measured, as these should assistin the structure determination far more than the theories of mole-cular packing put forward. This is particularly the case with theinteresting observations on colour dimorphism in which one mightcorrelate change of colour with the actual alteration in the mutualarrangements of chain and ring systems rather than with chemicalinteraction. Among condensed ring systems, J. Hengstenberg andJ. Palacios56 have shown that the two anthracene molecules indianthracene are probably linked a t the 9 : 9'-positions, the anthra-cene ring system being slightly bent at thc centre to enable thislinking to take place.There has been a, certain increase in X-ray measurements takenrather for purposes of identification of the compounds than forstructure determinations.A very complete series is that of deriv-atives of ephedrine and $-ephedrine:' where the measurementsshould prove sufficiently complete to lead ultimately to the exactstructures.In the last two years a number of important groups of biochemicalcompounds, the sterols, the bile acids, the female and the malesexual hormones, as well as other substances, pre-mandiol, digitoxin,bufotoxin, etc., have been shown to be based on a common condensedring system.The establishment of this common skeleton was dueinitially to X-ray investigation of one of its more complex members,calciferol (vitamin D). The chemical proofs of the new sterolformula are dealt with in another part of these Reports (p. 198), butit is here of some importance to state precisely how the methods ofcrystal analysis were of use in settling difficul6 points or indicatingpossible lines of attack. The following summary is arranged inroughly chronological order.52 J. D. Bernal and (Miss) D. Crowfoot, Trans. Faraduy SOC., 1933, 29,1032; A., 1107.53 K. Herrmann and A. H. Krunimacher, 2. K~ist., 1932,81, 317; A., 1932,450.6 p 3:. Hertel and G. H. Romer, Z. physikal. Chem., 1933, [BJ, 22, 267; A .,1004.5 5 E. Hertel and K. Schneider, ibid., 1932, [ B ] , 18, 436; A., 1932, 1080;E. Hertel and G. H. Romer, ibid., [B], 19, 228; A., 116; ibid., 1933, [B], 22,380; A., 1004.5 6 Anal. Pis. Quirn., 1932, 30, 5 ; A., 1932, 327.G 7 K. Briickl, 2. Krist., 1932, 81, 219; A., 1932, 451; B. Gossner andH. Neff, Z. Krist., 1933, 85, 370; A., 892; ibid., 86, 32; A., 1004BXRNAL AND CROWFOOT. 423( I ) 'l'bt: first crystal structural investigatioii of calciferol aid itsrelated compounds 58 and of cholesterol and ergosterol showed (a) onthe evidence of the space group and number of molecules in the cell,that calciferol was a simple substance (see p. 412) ; (b) on evidenceof cell size and optically determined molecular orientation, that thesterol molecule must occupy a space 5 x 7.2 x 17-20 A.and mustform in the crystal a double layer similar to that of long-chainalcohols. This showed that the accepted skeleton (111) must bewrong in at least two essential points : (i) the sharing of one carbonatom by rings I, 11, and 111, for this would lead to a much thickermolecule, and (ii) theattachmentof thelong chain at C,,, which wouldlead to a much shorter molecule than that observed. Purther,(iii) the double layer pointed to a hydroxyl group placed terminallyin the molecule. (This was confirmed later by an examination of$-cholesterol, which with a hydroxyl j n ring I1 shows no doublelayer.)(2) Meanwhile, a concurrent but fortuitously uiidertaken investig-ation of cestrin 59 pointed strongly to a structure with a phenanthrenenucleus of approximately the same molecular width and thickness.A little later it was apparent that pregnandiol,60 which was alreadyknown to be related through the bile acids to the sterols, had alsovery similar molecular dimensions.(3) These were the considerations from crystallography that,added to the serious chemical defects of the old sterol formula, ledRosenheim and King61 to propound the new formula (IV) sub-sequently amended to (V), which accounted for all the featuresoccurring in the crystal analysis.\/I\/I(111. )Subsequent work has consisted, for the most part, in attempting(4) The general trans-nature of the skeleton probable from theB* J.D. Bernal, Nature, 1932, 129, 277; A., 1932, 327.59 J.D. Bernal, Chem. and Ind., 1932, 51, 259; cf. C . Gaudefroy, Compt.Go J. D. Bernal, Chem. and Ind., 1932,51,466 ; A., 1932, 736 ; Nature, 1939,G1 Chem. and Id., 1932, 51, 464 ; A., 1932, 736.to fill in details of the structure.rend., 1932, 195, 623; A., 1932, 1079.129, 721; A., 1932, 658424 CRYSTALLOGI.RAPHY.thinness of the molecule was made more certain by an examinationof cis- and truns-hexahydrochrysenea 62 synthesised by G. R. Ramageand R. Robinson, of which only the former showed no resemblanceto a sterol derivative.(5) The attempt to settle the nature, 5- or 6-membered, of ringI V has led to much work on the hydrocarbons derived from thesterols, in particular Diels's hydrocarbon CI8Hl6. While it hasbeen impossible to determine its structure, X-ray and opticalmethods have shown that it is not cycZopentenophenanthrene,@and its resemblance to retene suggested the structure y-methyl-cyczopentenophenanthrene.This was subsequently synthesised byE. Bergmann64 and by G. A. R. ICon65 and appears to be indis-tinguishable from the natural product.(6) Much further work is in progress, in particular on the positionof the double bonds. Here, both X-ray and optical theory66suggest but cannot prove that in ergosterol these lie in rings Iandlor 11.Similar but less extensive work has been done on vitamins B,,B4,67 and C. The structural formula of the last, now known asZ-ascorbic acid (VI), was established, as in the case of the sterols,This is still in opposition to current chcmicd opinion.5%HOCH, /%.;.*')-.I \=/=\=/=\,/=\,/I ,"=C,W.1 HOHO OH (VII.)by a close collaboration of E. G. Cox's68 crystal analysis withorthodox chemical methods.In another important group, the polyenes, which includes caro-tene and vitamin A, a very complete study of methyl bixinhas been carried out by H. Waldmann and E. Brandenberger.89These crystals have remarkable optical properties ; the refractiveindices of 1-630,1.649, and > 2.63-for which last vibration directionthe crystal is almost black-fix the orientation of the moleculeabsolutely in the cell, and determine the nature of the molecule as acis-polyene chain. This is presumably the chain of the completely62 J. D. Bemal, Chem. and Ind., 1933, 52, 288.64 E.Bergmann and H. Hillemann, Ber., 1933, 66, 1302; A., 1164.O 5 Chem. and Ind., 1933, 52, 951.66 J. D. Bernal and T. M. Lowry, Chem. and Ind., 1933, 52, 10.67 J. D. Bernal and (Miss) D. Crowfoot, Nature, 1933, 131, 911; A . , 768.Gs Nature, 1932, 130, 205; A., 1932, 987.69 2. Kriat., 1932, 82, 77; A., 1932, 682.Ibid., p. 729BERNAL AND CROWPOOT. 425polymerised rubber which has recently been obtained in singlecrystals by E. W. Wa~hburn.~~Fibre Xtructures.A discussion of X-ray work on fibres requires a separate report :here we can only mention the most salient advances of the last twoyears. The whole subject has been admirably summed up byW. T. Astbury in “ Fundamentals of Fibre Structure ” (Oxford,1933), and in shorter articles.71The simplest fibre substances studied are the polyoxymethylenes.Crystallographic research on these, chiefly carried out by H.W.Kohlschiitter and L. Sprenger72 and E. S a ~ t e r , ~ ~ has been ableto show precisely how the polymerisation takes place, and hasprobably given the clue to the mechanics of polymerisation in allchain polymeride systems. Trioxymethylene crystallises in thehexagonal system with six-ring molecules arranged in order alongthe hexad axis :On exposure to light in the presence of free formaldehyde, nucleiappear in the crystals, and starting from them fibrous polyoxy-methylene appears, the fibre axis lying along the hexagonal axis,and with a 9-fold repeat pattern corresponding to the 6-fold repeatof the trioxymethylene.All that has happened, quite clearly, is aninterchange of homopolar bonds proceeding in a ‘‘ zip ” fashion,made possible by the previous apposition of the appropriate mole-cules of the ring polymeride by crystal forces. In this way thecrystals of ring polymerides autocatalyse their transformation tochain polymerides. This mechanism may prove of very generalapplication. The apposition of chain polymeride molecules neces-sary to form fibres may be produced by flow and shear, as has beenvery beautifully shown by W. H. Carothers and J. W. Hill74 for70 Physical Rev., 1931, [GI, 38, 1790; A., 1933, 371.71 Science Progress, Oct. 1933, 210; W. T. Astbury and H. J. Woods, Phil.Trans., 1933, [A], 232, 333; J. Text. Inst., 1932, 23, 17; A., 1932, 451.72 Z .phydml. Chem., 1932, [B], 16, 284; A., 1032, 721; cf. W. Kern,2. phyeikal. Chem., 1932, [B], 18, 417; A., 1932, 1080; ibid., 1933, [B],KoEloid-Z., 1932, 61, 308; As, 14.21, 161, 186; A,, 666; cf. 2;. Krist., 1932, 83, 340; A., 13.7 4 J . Amer. Chern. Soc., 1932, 54, 1579; A., 1932,601.0 426 CRYSTALLOGRAPHY.poly-esters. A thread of glassy poly-ester is instantaneously con-verted by strain into an oriented fibre.Of carbohydrate polymerides, the literature on cellulose is toovoluminous even to be referred to, but some understanding of themechanism of cellulose formation in Nature has been derived by thestudy of the cell walls of single cells of the alga Valonia ~entricosa,'~which have been shown to consist of a double spiral network ofcellulose chains intersecting nearrly a t right angles to give a " multi-ply " structure of maximum strength.This is probably the casein all vegetable cell walls, save that in elongated fibre cells the angleis reduced to a few degrees. In wood, K. Preudenberg 76 has shownthat the cellulose network is reinforced by interstitial lignin.The starches have been much studied, particularly by J. R. Katz 77and his co-workers. The difference between starch and celluloseis due t o the steric inability of the former, derived from a-glucose,to build straight chains leading to a nearly amorphous structure.The most generally significant advance, however, is in our under-standing of the essential nature and classification of proteins.W. T.Astbury has now shown that the protein, besides its character-istic peptide chain which can exist in the a-(contracted) orp-(expanded) form, has also a characteristic thickness and width.The thickness 4.65 8. between NH and CO in neighbouring chainsis roughly the same for all proteins, but the width corresponds tobonding between the side chains, and consequently varies with theirlength, being only 4.5-6.1 A. in silk fibrin,78 and up to 9.8 A. inkeratin. Further, the side chains can remain unattached, as in thesoluble proteins like ~valburnen,~~ can be attached by ionic linkagesin the gelatinsYs0 myosin, and muscle,s1 or by permanent peptide7 5 W. T. Astbury, T. C. Marwick, and J. D. Bernal, Proc. Roy. SOC., 1932,[B], 109, 443; A., 1932, 438.7 6 Papier-Fabr., 1932, 30, 189; A., 1932, 606; R.Thiessen, Incl. Eny.C'hen~., 1932, 24, 1032; A., 216.7 7 J. R. Katz and A. Weidinger, Rec. trav. claim., 1931, 50, 1133; A., 1932,149; J. R. Katz and J. C. Derksen, Naturwiss., 1932,20,851; A., 14; M. Samecand J. R. Katz (with R. Klemen), 2. physikal. Chem., 1933,163, 291 ; A., 464.7 8 0. Kratky and S. Kuriyama, 2. physikal. Chern., 1931, [B], 11, 363;A., 1931, 415 ; I. Sakurade and K. Hutino, Sci. Papers Inst. Phys. Chern. Res.Tokyo, 1933, 21,266; A., 1108; J. R. Katz and A. de Rooy, Naturwiss., 1933,21, 559; A., 1004; Rec. trav. chim., 1933, 52, 742; A., 893.7s W. S. Miller, K. G. Chesley, H. V. Anderson, and E. R. Theis, J. Amer.Leather Chem. ASSOC., 1932, 27, 174; A., 807.B0 J. R. Ketz, J.C. Derksen, and W. F. Bony Rec. trav. chin&., 1931, 50,1138; A., 1932, 123; J. J. Trillat, Ann. Inst. Pasteur, 1932,48,400; J . Chim.physiqzce, 1932, 29, 1; A., 1932, 466; 0. Gerngross, K. Herrmann, and R.Lindemann, Kolloid-Z., 1932,60,276; A., 1932, 1088; K. Hess and C. Trogus,Biochem. Z., 1933,262,131 ; A., 893 ; W. T. Astbury and W. R. Atkin, Nature,1933, 132, 348; A., 1108.a1 G. Boelim and H. H. Weber, KoZZoid-Z., 1932, 61, 869; A., 14BERNAL AND CROWFOOT. 427or -S-S- linkages in the sclero-proteins such as keratinF2 givingrise to a semi-rigid ladder-like structure. The extremely complexphysical properties of the proteins, particularly their behaviour onswelling with water 84 or adsorption on have alreadybeen- qualitatively analysed by this model.Here we have un-doubtedly the key to the mechanism of movement, if not to the veryexistence, of living matter.After the proteins, the lipoids are of the greatest biological interest.Here we find liquid-crystal structures playing an important d e , asemphasised by F. This is particularly the case in nerve:’which has been studied by X-rays in active, narcotised, and de-generate states. Thus the applications of X-ray technique arealready passing the boundary of biochemistry into physiology andpathology.Liquids and Classes.The methods of X-ray analysis of liquids and amorphous solidsdeveloped by Debye and Prim (see Ann. Reports, 1929, 26, 306)have been the basis for an extensive study of types of liquid struc-ture. The two characteristics of a liquid are (a) the close contactbetween the molecules, leading to a relatively dense approximatelyclose-packed structure ; ( b ) the irregular nature of the packing overdistances more than a few molecular diameters. Both thesecharacters are modified by the temperature and the nature of themolecules. Higher temperatures produce rotary and translatorymolecular movements, but molecular clustering still occurs up toand even slightly above the critical point, as shown by the work ofF. H. W. No11 88 on ether. The effect of the nature of the moleculesis more varied. X-Ray analysis distinguishes between small-anglediffraction due to intermolecular, and large-angle diffraction due tointramolecular, interference. Further, the small-angle diffractionis either normal, indicating close packing, or abnormal, indicating alooser arrangement of molecules, generally brought about by dipoleattractions. The analysis of these offers greater difficulty as it can82 W. T. Astbury, Trans. Faraduy SOC., 1933, 29, 193; A., 216.s3 J. B. Speakman and M. C. Hirst, ibid., p. 148; A., 227; W. T. Astburys4 (Miss) D. J. Lloyd and H. Phillips, T’rans. E’aruday SOC., 1933, 29, 132;85 A. H. Hughes and E. K. Rideal, PTOC. Roy. Xoc., 1932, [A], 137, 62; *4.,86 Trans. B’araday SOC., 1933, 29, 1016; A., 1108; 2. Krist., 1932, 82,87 H. Handovsky, KoZZoid-Z., 1933, 62, 21; A., 313; G. Boehm, ibid.,a8 Phy8i.d Rev., 1932, [GI, 42, 336; A., 14.and T. C. Marwick, Nature, 1932,130, 309; A., 1932, 1080.A., 226.909.379; A., 1932, 798.p. 22; A., 313428 CRYSTALLOGRAPHY.on1.y be done indirectly by assuming a mean molecular distribution,and calculating the diffraction to compare with that observed. Ithas, however, been attempted in one case, that of where thebasic arrangement is tetrahedral (quartz-like) rather than close-packed.Greater anisotropy of molecules leads to clumping in parallelaggregates, as occurs in nematic liquid crystals which show whenoriented a fibre diagram. Finally they may arrange themselves inlayers, generally curved, as in smectic liquid crystals which show oneset of crystal plane refiexions. (Between these and true crystals arestill other intermediate stages; see below.) The general scheme isshown below:1. Atomic liquids : Hg, Na, K, Rb,2. Close-packed molecular liquids : GEL,.^^ CCl,,Ql C,H,Cl,,g3 (C2H6)20,88 S,.3. Associated molecular liquids : Hg0,93 C,H,,.0H.944. Nematic liquids.965. Smectic liquids : thallium st,earntr.SGThe precise intimate structure of liquids is still in dispute.According to G. W. Stewart,g7 a liquid consists of a mass of smallquasi-crystallites, the so-called cybotactic groups , containing 10-100 molecules, not necessarily permanent, and separated by a numberof unattached molecules. Another view 98 contends that theseaggregates have only a statistical and not a physical existence, andthat every molecule in the liquid is immediately surrounded by thesame quasi-regular arrangement of molecules.We have no space to discuss the great developments of the pasttwo years in the study of liquid crystals (see Ann. Reports, 1931, 28,280). Luckily the whole subject is amply covered by the FaradaySociety’s volume “ Liquid Crystals and Anisotropic RTelts.” 99The most important facts that have emerged are the clear dis-tinction between nematic liquids where neighbouriiig molecules aremerely oriented in parallel, and smectic liquids where they are alsoarranged in regular planes, inside of which they either form a two-J. T. Randall and H. I?. Rooksby, Nature, 1932,130,473; A., 1932, 1192.F. Ehrhardt, ibid., p. 605; A,, 1932, 987.s~ J. D. Bernal and R. H. Fowler, J . Chem. Physics, 1933, 1, 515.s1 H. Menke, Physikal. Z., 1932, 33, 693; A., 1932, 986.O3 E. Amaldi, Physikal. Z., 1931, 82, 914; A., 1932, 11.94 W. C . Pierce, Physical Rev., 1931, [ii], 38, 1409; A., 1932, 12.O 5 K. Hernaann, A. H. Krwnrnacher, and K. May, 2. Pliysz‘ll, 1931, 73,419; A., 1932, 214.K. Eemenn, Trans. Faradccy SOC., 1933, 29, 972; A., 1107.9 7 IndianJ. Physics, 1933, 7, 603; A., 891.s8 Cf. Trane. Faraday SOC., 1933, 29, 1070 et scq.yo Ibid., p. 883; J. D. Bernal, Nature, 1933, 132, 86ItTCRNAT, AN 1) CROWFOOT. 429dimensional liquid (true smectic) or are arranged in lattices (crystal-line smectic), which are still free to move over each other. Theurious fo cal conic structures exhibited by smectic crystals havebeen shown by (Sir) W. H. Bragg to be an equilibrium structurefor molecular layers lacking rigidity.Glasses have plainly the same internal molecular structure asliquids : they differ by their relative immobility, with the consequentappearance of rigidity, or more correctly, of a long characteristictime for plastic flow. (Liquids also have an internal rigidity, asshown by B. V. Deriagin,2 but the characteristic time is short.)Therniodyrmrnically, glasses differ from undercooled liquids by theabsence of internal equilibrium, as discussed by I?. Simon: E.Berger,4 and G. Tarnm~~nn.~ The liquid-glass transformation is notsharp, but inside a small temperature range the slope of variousproperty-temperature curves changes markedly. Gels are diphasicsystems, one of which is generally a glass, though sometimescrystalline .6W. H. Zachnriasen7 has discussed the conditions under whichglasses are formed, equating them with the possibility of formingthree-dimensional frameworks. This, in the case of ionic glasses,occurs when, in thc notation of p. 407, b/n 2 z/2, e.g., for boricanhydride, silica, tungsten trioxide, beryllium fluoride. Glassesmay also be expected where metallic and homopolar structures haveapproximately the same energy, e.g., for selenium and arsenic.The actual atomic arrangements in vitreous silica have beendetermined by B. Warren; it is simply a completely irregularthree-dimensional network of SiO, tetrahedra sharing oxygens.Other glasses have also been studied by x-ray^.^The utilisatjon of electronic and protonic waves for crystal, liquid,and molecular analysis has grown into too large a subject to bediscussed in this report (see Ann. Reports, 1929, 26, 284). The mainmethods and results, however, have been described in W. I,. Bragg’s“ Crystallinc Stat(\.’’ Owing to the low pcnetrating powor of1 Trans. Furaduy ~S’oc., 1933, 29, 1066; *4., 1105.Physikat. 2. Soviet Union, 1933, 4, 431.2. anorg. Chem., 1931, 203, 219; A., 1932, 217.J. Amer. Ceram. SOC., 1932, 15, 647; A., 115.G. Tammann and A. Elbrachter, 2. anorg. Chem., 1932, 207, 2 6 8 ; A.,1932, 996.6 J. C. Derksen, Collegium, 1932, 838; A., 125.J . Amer. Chem. Soc., 1932, 54, 3841; A., 12; Physical Rev., 1932, [ii],38, 185; A., 1107.8 Z. Krist., 1933, 86, 349.M. Hirata, Scl;. Papers Inst. Phys. Chem. Res. Tokyo, 1932, 18, 237; A . ,1932, 1015430 CRYSTALLOGRAP€IY.electrons, the chief applications have been to thin films, metalsurfaces, and free molecules, regions where X-rays, for the oppositereason, give little information. Thus, the two methods are veryhappily complementary. J. D. B.D. M. C.J. D. BERNAL.D. M. CROWFOOT.B. W. ROBINSON.W. A. WOOSTER

ISSN:0365-6217    

DOI:10.1039/AR9333000360

出版商:RSC    

年代:1933

数据来源: RSC

摘要:

INDEX OF AUTHORS' NAMESABEL, E., 52..Abeles, A., 264.Acker, W., 326.Ackermann, P., 96.Ackermann, W., 300.A4dam, N. K., 212, 342.Adams, C. A., 279.Adams, F. H., 265.Adams, L. H., 295.Adams, R., 255, 256, 267, 258, 259.Adel, A., 80.Adell, B., 23.Adkins, H., 138, 139, 140, 141, 142,143, 259.Adler, E., 252, 353.Ageev, N., 384.Agulhon, H., 278.Ahmet, H., 327.Albrecht, O., 262.Albrecht, W. A., 327.Albrecht, W. H., 402.Albright, J. G., 397.Alder, K., 226, 227, 229, 233.Alexeevrt, E. N., 163.Allen, N., 293.Allison, F., 349.Allmand, A. J . , 49.Allsopp, C. B., 75.Almasy, I?., 54.Alter, C. M., 84.Alyea, H. N., 52.Amagat, P., 185.Amaldi, E., 428.Ambler, H. R., 297, 298, 300, 301.Aminoff, G., 405.Anderson, C. D., 355, 356, 368.Anderson, H.V., 426.Anderson, T. F., 131.Andress, K., 395.Andrews, T. S., 326.Angelstti, M., 258.Antropoff, A. von, 96.Anwar-Ullah, S., 130.Arakawa, S., 313.Arbusov, A. E., 94.Arbusov, B. A., 94.Archibald, E. H., 85.Archibald, F. M., 279.Armstrong, G., 37, 285.Arnrtl, V., 293.Amfelt, H., 395.h e s o n , E. J., 298.Asahina, Y., 218, 220, 321, 222, "23,224, 225.Asai, T., 310.Asano, J., 224, 225.Ascher, E., 91.Aschheim, S., 341.Ashley, M., 347.Asmus, R., 166.Asmussen, R. W., 96.Astbury, W. T., 426, 426, 137.Astin, S., 281.Aston, Y. W., 82, 83, 84, 346,Asundi, R. K., 65.Atanasiu, J. A., 284, 291, 293.Atkin, W. R., 426.Atkins, W. R. G., 364.Audrieth, L. F., 393.Auerbach, F., 254.i+pr, P., 353.Ault, R.G., 175, 336.AumBras, M., 296.Austin, J. B., 365.Austin, P. R., 180.Auwers, K. von, 137.Avsejevjtsch, G. I?., 288.Axionnaz, R., 320.348.347,Babcock, H. D., 348.Baccaredda, M., 397, 405.Bach, A., 165.Bacharach, A. L., 337.Bacher, R. F., 77.Bachmann, W. E., 182.Bachoulkova-Brun, R., 2S1.Back, E., 76.Badami, J. S., 77.Bader, G., 316.Blicklin, E., 378.Bkkstrom, H. J . L., 52.Baeckstrom, S. E., 292.Baer, E., 329.Baernstein, H. D., 333.Baeyer, A. von, 212.Bailey, C. R., 80.Bain, (Miss) A. M., 261.Bainbridge, K. T., 83, 346, 347, 348.Baird, D. K., 175, 336.Bairstow, S., 40.Baker, J. W., 44.Balrtnescu, G., 280.Ball, F. K., 320.Baltzly, R., 186.Banchetti, A., 289.Banerjee, S., 420, 421.43432 INUEX OF AUTHORS’ NAhIES.Banga, I., 343.Bannister, F.A., 397, 406, 411.Barbieri, G. A., 101.Barker, E. F., 80, 347.Barlett, H. B., 403.Baroni, A., 394.Barratt, S., 129.Barrett, C. S., 386.Barth, T. F. W., 382, 400, 406, 420.Bartlett, E. P., 14.Bartlett, J. H., 70.Bass, A., 186.Bassett, H., 102.Basu, I<. I?., 310.Bates, J. R., 51.Batuecas, T., 87, 88.Baudisch, O., 403.Baucirexler, H., 84.Baumann, W., 80.Baumgartner, H., 25-1.Bawden, A. T., 290.H a m , C. E. H., 51.Baxter, G. P., 83, 84, 86, 87.Baxter, R. A., 302.Baylis, J. R., 289.Beal, G. D., 280.Beale, C. H., 182.Beamer, C. M., 279.Beans, H. T., 291.Beard, H. C., 288.Bearden, J. A., 378, 380.Beaumont, A. B., 322.Bech, A., 299.Beck, G., 358.Becker, B.C., 256.Becker, F., 244.Becker, K., 389.Becker, R., 39, 366.Beckham, L. J., 302.Beevers, C. A., 98, 410.Beintema, J., 110, 416.Beisswenger, O., 259, 260.Belcher, D., 285, 286.Belchetz, L., 47.Belikova, A., 299.Hell, F., 94, 257, 261, 263.Bell, R. A., 287.Bell, R. P., 15, 20, 30.Bemmann, R., 395.Benedetti-Pichler, A. A., 279.Bengough, G. D., 55.Bennett, 186.Bennett, G. M., 44.Bennewitz, K., 37.Benz, F., 148.Benz, P., 147.Berchet, G. J., 133, 134, 135, 136, 179.Berg, 277.Berg, R., 282, 284.Berger, E., 429.Bergmann, E., 214, 424.Bergstrom, F. W., 54.Rerl, E.. 303, 395.Bernd, J. D., 34, 199, 214, 392, 411,Bernhauer, K., 318.Bernreuther, F., 49.Bertho, A., 310.Best, F. L., 287.Best, R.J., 285.Bethe, H., 72.Beuthe, H., 54.Bewilogua, L., 415.Bezzi, S., 163.Bhatnagar, S. S., 106.Bienfait, H., 292.Biesalski, E., 299.Bigelow, N. M., 340.Bijvoet, J. M., 398, 399, 409.Biltz, W., 91, 124, 386.Birch, T. W., 167, 337.Birch-Hirschfeld, L., 307.Birge, R. T., 348.Birkenbach, L., 86.Birkinshaw, J. H., 319.Biscoe, J., 405.Bishop, E. R., 270, 271.Bjerrum, N., 15, 16, 17.Bjurstrom, T., 391.Blacet, F. E., 48, 297.Black, C. A., 14.Black, (Miss) M. M., 64.Blackett, P. M. S., 355, 356, 367, 358.Blackie, A., 302.Blake, M. A., 322.Blayden, H. E., 24.Bleakney, W., 28, 29, 33.Bleyer, B., 159.Blix, R., 405.Bloch, A. M., 49.Bloch, R., 366.Block, R. J., 334.Blomquist, A. T., 135.Blum-Bergmann, 0 ., 183.Bodendorf, K., 164, 279.Bodenstein, M., 49, 50.Bodforss, S., 289.Bodger, W.H., 09.Boehm, E., 46.Boehm, G., 426, 427.Bohm, J., 379.Boehm, T., 279.Boekenoogen, H. A., 137.Bottger, W., 289, 291.Bogdan, M., 40.Bogdanova, O., 299.Bohnholtzer, W., 294.Boldyreff, A. W., 292.Bollinger, G. M., 295.Boltunov, J. A., 289.Bon, W. F., 426.Boname, A., 293.Boncyk, L., 18.Bond, P. A., 18.Bonhoeffer, K. F., 32, 42, 46, 47,Bonstedt, K., 201, 211.417, 422, 423, 424, 426, 428Booth, H. S., 127.Borelius, G., 365.Boren, B., 391.Borgeaud, P., 211.Rorinski, P., 299.Born, M., 15, 341, 367.Borsche, W., 192, 193, 195.Bortels, H., 307.Bossert, K., 201.Bowden, F. P., 37.Bowen, E. G., 393.Bowen, E. J., 54, 56.Boyland, E., 339.Bozhenko, A., 325.Bozorth, R.M., 408.Bradley, A. J., 380, 383, 392.Brttdy, 0. L., 99, 100.Eraelrken, H., 379, 398, 399.Bragg, (Sir) W. H., 111, 357, 429.Bragg, W. L., 360, 384, 429.Braida, A., 128, 129, 131.Brandenberger, E., 405, 424.Brann, B. F., 290.Brasseur, H., 369, 409, 410.Braun, J . von, 162, 208, 242.Braun, W., 239.Braune, H., 14, 415.Bray, W. C., 299.BrdiEka, R., 270.Bredig, G., 264.Hredig, M. A., 410.Breit, G., 77.Breitschneider, O., 389.Bresler, F., 14.Bresler, S. E., 54.Bretsznajder, S., 55.Brewer, F. M., 98, 99.Brickwedde, F. G., 28.Bridgman, P. .W., 368, 373.Briegleb, G., 20.Brill, R., 403.Brilliantov, N. A., 375.Brings, T., 366.Rrinkman, R., 53.Brinkmann. E.. 328.Briscoe, H.’V.A., 83, 91.Britton, XI. T. S., 285, 286, 288, 289,Brockmann, XI., 151, 152, 153, 154,Brockway, L. O., 93, 96, 396, 415.Broderick, A. E., 143.Broderick, S. J., 287.Brodkob, F., 107.Brodsky, A. E., 16.Bronsted, J. N., 17, 18, 20, 21, 24, 53.Brown, G. W., 32.Uru, I,., 414, 416.Bruchhausen, F. von, 247.Bruckl, K., 422.Briinnich, J. C., 289.Brukl, A., 280.Brunovski, B. K., 389,296.156, 336.433Bruzau, (Mme.), 264.Buchbock, G., 285.Buck, J. S., 141.Bucknall, W. R., 11 1.Budge, P. M., 391.Buerger, M., 398.Bussem, W., 386.Bulian, W., 362.Bumm, H., 389.Burgers, W. G., 389.Burgsmuller, W., 374.Burk, D., 305, 306, 307, 308.Burrell, G. A., 302.Burstall, F. H., 100, 101, 106, 241.Burt, F. P., 86.Burton, H., 52.Buss, G., 80.Butemndt, A., 216, 341, 343.Butler, J.A. V., 21, 37, 285.Butz, L. W., 334.Cadenbach, G., 18.Caglioti, V., 376, 392, 396.Cairns, A. C. H., 51.Caley, E. R., 31.Calhane, D. F., 292.Callan, T., 291, 295.Cambi, L., 106.Campbell, J. S., 77, 348.Campbell, 35. A., 285.Canter, F. W., 222.Carlson, J. F., 358.Carlsson, O., 392.Carothers, W. H., 133, 134, 135, 136,Carr, I?. H., 335.Carrington, H. C., 175, 336.Carter, J. M., 27.Carter, J. S., 19.Case, E. M., 328.Cassar, H. A., 279.Cassie, A. B. D., 80.Catenacci, M., 288.Cates, J., 403.Cavanagh, B., 284, 385, 291.Cawood, W., 88.Chabre, P., 157.Chadwell, H. M., 47.Chadwick, J., 351, 353, 355, 356.Challenger, F., 317.Chamberlin, D. S., 297.Chambers, L.A., 54.Chan, S. B., 86.Chandlee, G. C., 24.Chapin, R. M., 277.Chapman, 186.Chapman, A. W., 178, 188.Chapman, D. L., 50.Charles, J. H. V., 319.Chas, C. Y., 356.Chase, E. F., 19.Chatfield, J. N., 290.179, 425434 INDEX OF AUTHORS' NAMES.Chaudron, G., 402, 403.Cheng Ling Liu, 53.Chesley, K. G., 426.Chevallier, A., 157.Chibnall, A. C., 418.Chilton, D., 79.Christian, W., 159, 160.Christiansen, J. A., 297.Chrzaszcz, T., 317.Chu, T. T., 266.Chuang, C. K., 210.Claassen, A., 389.Clapp, M. H., 290.Clark, W. M., 285.Clarke, B. L., 291, 292.Claxton, G., 19.Clay, J., 358.Clemo, G. R., 232, 233, 234, 237, 238.Cliff, I. S., 163.Closs, J. O., 289.Clusius, K., 116, 415.Clutterbuck, P. W., 319.Cobb, A., 286.Cockcroft, J.D., 350, 353.Coehn, A,, 391.Coffman, D. D., 135, 136.Colin, H., 315.Colles, W. M., 103.Collie, B., 20.Collie, C. H., 28.Collie, J. N., 236.Collins, A. M., 135.Compton, A. H., 358.Cornpton, K. G., 287, 292.Conant, J. R., 165.Connor, R., 138.Cook, J. W., 190, 191, 207, 214, 339,Cook, R. G., 25.Cook, R. P., 328.Coops, J., 418.Copaux, H., 300.Cope, I. S., 103.Corballini, 31 8.Corbellini, A., 257, 258.Cordes, H., 90.Corey, R., 401, 410, 417.Cornish, R. E., 28.Coster, D., 418.Covert, L. W., 141.Cowley, M. A., 140.Cowperthwaite, I. A., 292.Cox, D. C., 292.Cox, E. G., 108, 167, 172, 408, 417,Cox, R. F. B., 276.Craggs, H. C., 49.Craig, L. C., 241.Crane, H. R., 352.Craven, E. C., 278.Crawford, M.F., 77.Crtmford, S. R., 37.Church, J. M., 177.340, 343.419, 424.Cremer, E., 32, 33.Criegee, R., 166.Crooker, A. M., 77.Cross, P. C., 80.Cross, R. J., 300.Crowell, W. R., 293.Crowfoot, (Miss) D. M., 104, 214, 409,Cupery, M. E., 256.Cupples, H. L., 300.Curd, F. H., 219, 224.Curie, (Mme.) I., 352, 354, 356,Curie, M., 344.Currie, T., 225.Curry, J., 47.Curtis, W. E., 79.Cushing, R. E., 292.Cutcliffe, E. F., 103.Czapska, Z., 403.422, 424.358.Damiens, A., 89, 116.Demon, G. H., 48.Dane, E., 199, 202, 203, 204, 205,206, 207, 208.Danielli, J. F., 341.Daniels, F., 48, 52.Denilov, 183.Danilov, S. N., 134.Danilov, V., 404.Dann, W. J., 337.Danneel, R., 208.Darbishire, F. V., 324.Dasgupta, A.C., 409.Dasler, W., 138.Daugherty, J. P., 303.Dauphin&, J. A., 333.Davidenkov, N. N., 373.Davidson, G. F., 286.Davies, C. W., 24, 294.Davies, R. M., 372.Davis, H. S., 302, 303.Dawson, H. M., 19, 53.Daynes, H. A., 303.Deaglio, R., 362, 364.Debucquet, L., 282.Debye, P., 18, 369.Deckert, W., 302.Dee, P. I., 352.Do Eds, F., 288.Deflandre, M., 386, 402.Dehlinger, U., 383, 387, 388, 389.De Jong, W. F., 400.Delbanco, A., 17.Del Fresno, C., 293.De Mandrot, R., 373.Dember, H., 361.Deming, W. E., 409.Demole, V., 337.Denbigh, K. G., 91.Dengel, G., 416.UenigBa, G., 277INDEX OF AUTHORS’ NAMES. 435Denina, E., 295.Dennis, L. M., 116.Dennison, D. M., 76, 79, SO, 347.Deriagin, B. V., 429.Derksen, J. C., 426, 429.De Rooy, A., 426.Desai, K.V., 421.De Turk, E. E., 326.Deuticke, 327.Deutsch, A., 165.Dhar, J., 420.Diamond, H., 42.Di Capua, A., 318.Dice, (Miss) M. E., 27.Dickens, P., 294.Dickinson, B. N., 408.Dickinson, R. G., 108.Dickinson, S., 287.Diels, O., 199, 213, 226, 227, 229, 233,Dijk, J. A., 144.Dillon, R. T., 302.Dillon, T., 225.Ditt, M., 282.Dittrich, E., 299.Diwoky, F. F., 143.Dodd, E. N., 286, 289.Dodds, E. C., 340, 341.Dodge, B. F., 277.Doisy, E. A., 216.Dole, M., 36, 285, 286, 292.Dollins, C. B., 270, 271.Dommerich, K. H., 374.Doody, T. C., 31.Dorcas, M. J., 83.Dorfman, M. E., 27.Drakeley, T. J., 299.Drew, H. D. K., 107, 408.Drewski, K., 293.Drier, R. W., 389.Drikos, G., 265.Dripps, R.D., 178.Drugman, J., 411.Drumm, P. J., 165.Drummond, J. C., 157, 158.Druschinin, W. W., 54.Dubois, D., 286, 287.Dubois, J., 302.Du Bridge, L. A., 287.Dubsky, J. V., 282, 283.Dull, H., 140.Dunwald, H., 361.Duffendack, 0. S., 272.Duhn, T. G., 321.Dull, M. F., 47.Du Nouy, P. L., 285.Dunkel, M., 74.Dunn, J. L., 187.Dunstan, W. R., 191.Duntze, R., 219, 220.Duquenois, P., 283.Dushman, S., 57.Dutoit, P., 291.342, 343.Du Vigneaud, V., 334.Dvoleitzka-Gombinska, (Mme.), 181.Dwyer, F. P., 400.Dyer, H. M., 334.Eastman, E. D., 284, 294.Ebert, F., 128, 131, 389, 392.Ebert, M. S., 101.Eckling, K., 380.Edgar, G., 299.Edgcombe, L. J., 298.Edner, A., 374.Egartner, L., 411, 421.Eggeling, H., 102.Ehmann, L., 214.Ehrenstein, M., 240, 241.Ehret, W.F., 392.Ehrhardt, F., 428.Ehrlich, E., 367, 418.Eichmann, O., 80.Elbrtichter, A., 429.Elder, L. W., 286.Ellinger, P., 159.Elliott, K. A. C., 263.Ellis, C. D., 355.Ellis, D., 311.Ellis, E. W., 297.Elsasser, W., 354.Embden, G., 327.Emde, H., 140, 255.Emmert, E. M., 320.Endell, K., 406.Endo, K., 19.Engelhard, E., 361.Ensfellner, L., 239.Ephraim, 96.Erdey-Griiz, T., 37.EremQv, M. A., 370.Erler, W., 389.Ernst, T., 401.Emera,, J., 369.Ertel, L., 140.Erxleben, H., 196, 198.Escher, H. H., 149.Estermann, I., 78.Eucken, A., 14, 39.Euler,H.von, 152, 157, 168.Evans, J., 278.Evans, R. W., 282.Evans, S. F., 79.Evans, U. R., 55.Everest, A. E., 217.Evering, B. L., 47.Evers, W.L., 177.Evrard, V., 282.Ewald, 360.Ewins, A. J., 236.Eyring, H., 30, 31, 32, 34, 49, 50, 53,367.Fagelston, I., 303.Failey, C. F., 17, 1436 INDEX OF AWTHBICS’ NAMES.Fairn, M. W., 277.Falkenhagen, H., 36.Faltis, F., 242, 245, 246, 248, 249,Farkas, A., 29, 30, 33, 42.Farkas, L., 29, 30, 33, 41.Farmer, E. H., 142.Farnsworth, M., 80.Fauroux, P., 316.Fearon, W. R., 278.Feather, N., 352.Feigl, F., 282.Feitknecht, W., 401.Feldhaus, A., 247.Fenwick, F., 285, 288, 289, 290, 291.Fermi, E., 77.Feussner, O., 271.Fichter, F., 163.Field, R. H., 277.Fieser, L. F., 191.Fieser, M., 191.Fink, W. L., 391.Finkelnburg, W., 64, 79.Fischer, E., 217, 218, 221, 222.Fischer, F. G., 140.Fischer, Hans, 251, 252, 253, 254.Fischer, Hellmut, 283.Fischer, H.0. L., 218, 221, 329.Fischer, J., 115, 126.Flanzy, M., 278.Flatt, R., 293.Fleck, E. E., 340.Fleck, H. R., 281.Fleischer, I. F., 286.Fletcher, C. J. M., 40.Fletcher, L., 285.Fleury, P., 166.Florkin, It., 250.Flosdorf, E. W., 54.Fock, V., 63.Focke, A. B., 379.Fond, G. R., 274.Foord, S. G., 304.Foote, F., 385.Forbes, G. S., 86.Forestier, H., 403, 416.Forscey, L. A., 130.Forssman, S., 309.Fosbinder, R. J., 286, 287, 280.Foulk, C. W., 290.Fouretier, G., 285, 292.Fowler, A., 64.Fowler, R. H., 22, 34, 70, 361, 428.Fox, J. J., 287, 301.Fox, 5. M. C., 103.Foz, 0. R., 420.Franck, H. H., 410.Franck, J., 34, 39.Frang, G., 299.Franke, W., 165.Frantz, H. W., 27.Franz, H., 96.Frauendorfer, H., 249.250.Frauenhof, H., 96.Frazer, J.C. W., 101, 299.Fredenhagen, K., 18, 128.Frederick, R. C., 300.Fr6ederickaz, V., 373.French, A. J., 289.Frenzel, A., 395.Freudenberg, K., 80, 426.Freund, M., 236, 344.Prey, F. E., 47, 302.Fricke, R., 96.Fridenson, A., 342.Friedman, H. B., 25.Friend, N. A. C., 167.Friess, H., 46.Frisch, R., 78.Froschl, N., 140.Frommer, M., 284, 285.Frost, A. A., 31.Frost, A. H., 34.Frost, A. V., 89.Frost, O., 89.Frumkin, A., 37.Fuldner, H., 410.Furth, R., 354.Fuhrmann, F., 314.Fuller, J. E., 306.Fuller, M. L., 386.k’ulton, J. W., 275.Funk, H., 282.FUOSS, R. M., 289.Eurman, N. H., 284, 289, 290, 291,Furry, W. H., 358.Furter, M., 209.Fuzikawa, F., 220, 222, 223, 224.G&dke, W., 342, 343.Galand, M., 403.Galinowsky, F., 236.Galley, R.A. E., 142, 191.Gamow, G., 77.Gans, D. M., 352.Gantz, E. St. C., 273.Gapon, E. PIT., 354.Garner, W. E., 55.Gassner, G., 324.Gatterer, A., 304.Gatty, O., 15.Gaudefi*oy, C., 423.Gaunt, J. A., 62.Gautier, A,, 297.Gavrilescu, N., 338.Gayler, (Miss) M. L. V., 389.Gehm, G., 391.Gehring, A., 327.Geib, K. H., 48, 92.Geiger, 363.Geilmann, W., 91, 280.George, C. M., 20.Georgia, F. R., 277, 314.Gerb, L., 109.293INDXX OF AUTHORS’ NAMES. 437Gerbes, O., 297.Gercke, A., 208.Gericke, P. H., 247.Gerischer, W., 329.Gerlach, W., 272.German, W. L., 296.Germann, F. E. E., 297.Gerngross, O., 281, 426.Gettler, A. O., 279.Ghiron, D., 393.Gibbs, F.B., 50.Gibbs, R. C., 57, 77.Gibson, C. S., 103, 266.Gibson, F. H., 289.Giehrmann, H., 299.Giese, H., 95.Gilbert, E. C., 286.Gilbert, F. L., 105.Gilchrist, (Miss) H. S., 418.GiHllan, E. S., 30.Gillespie, L. J., 14.Gilman, E., 288.Gird, J., 316.Girard, A., 342, 402, 403.Glasstone, S., 19, 284, 286.Gleu, K., 95.Glissmann, A., 52.Glocker, R., 383.Goens, E., 365, 372, 373.Gore, J., 285.Gotte, E., 295.Goetz, A., 365, 366, 377, 379.Goetze, G., 324.Goldberg, A,, 265.Goldberg, M. W., 207, 213, 214, 215.Goldman, F. H., 17, 23.Goldovski, A., 326.Goldschmidt, V. M., 409.Goldsztaub, S., 402, 403.Gem, A. S., 55.Goode, K. H., 292.Goodwin, L. F., 279.Goodwin, T. H., 417.Gorbach, G., 313.Gore, G., 132.Gortikov, V.M., 288.Gossner, B., 410, 422.Gotts, R. A., 100.Goubeau, J., 85.Goudsmit, S., 57, 58, 76, 77.Gould, A. J., 28, 29, 33.Gow, (Miss) E. R. L., 178.Gowlett, F. W., 292.Grdanin, M., 320.Grace, N. S., 77.Gradham, R. P. D., 378.Graf, L., 389, 393.Gray, L. T. M., 130.Green, J. R., 300.Greene, C. H., 87.Greer, E. N., 395.Gregorini, 318.Greiff, L. J., 22.Greville, G. D., 286, 287.Grieg, J., 400.Griffin, S. W., 300.Griffith, R. O., 52.Griffiths, J. A. G., 51.Grigg, P. P., 50.Grimm, H. G., 120.Grinberg, M., 356.Grinten, W. van der, 415.GroBk, B., 285.Groll, H. P. A., 102.Grondahl, L. O., 360, 363.Gronwall, T. H., 22.Gross, C. E., 308.Gross, P., 18.Gross, P. M., 19.Grossmann, H., 282.Groves, L.G., 287, 301.Griinberg, A. A., 107.Griinsteidl, E., 274.Grussner, A., 174, 175.Grundmann, C., 148, 150, 154.Gruneisen, E., 365.Gruner, E., 90.Gruner, J. W., 406.Gschaider, B., 210.Gudguen, E., 315.Guntelberg, E., 25.Guerlain, J. P., 185.Guggenheim, E. A., 24, 25, 26.Guillemin, V., jun., 62.Guiteras, A., 210, 211.Guittonneau, G., 311.Gull, H. C., 301.Gurevitsch, H., 239.Gurney, R. W., 37, 361.Gutzeit, G., 281.Guye, P. A., 297.Gyorgy, P., 159.Gysinck, T., 289.Haagen-Smit, A. J., 196, 198.Haarmann, W., 328.Haas, P., 315, 316.Haase, G., 294.Haddow, A., 341.Hagg, G., 385, 392, 409.Hagen, F. D., 266.Hahn, A., 328.Hahn, F. L., 284, 285, 295.Halberstadt, S., 280.Haldane, J. S., 300.Hales, H.R., 55.Hall, N. F., 292.Hall, It. E., 295.Halla, F., 406, 411, 421.Halmby, E., 420.Hamblet, C. H., 24.Hamburg, H., 282.Hhmilton, J. M., 289.Hammerschmid, H., 18.Hammett, L. P., 38, 98438 INDEX OF AUTHORS’ NAMES.Hammick, D. L., 52, 122.Hand, D. B., 250.Handovsky, H., 427.Hanemann, H., 386, 390.Hansen, L. A., 23.Hanson, N. H. W., 130.Hantzsch, A., 107.Harder, A., 92.Hardy, D. V. N., 139.Hardy, J. D., 347.Hardy, R. K., 19.Haring, H, E., 287.Harkins, W. D., 352, 354.Harlos, W., 300.Harmon, J., 334.Harned, H. S., 295.Harper, S. H., 214.Harris, L. E., 279.Harris, L. J., 167, 337.Harrison, G. B., 286, 287.Harteck, P., 46, 48, 50, 92.Hartel, H. von, 47.Hartelius, V., 317.Hartley, (Sir) H., 16.Hartman, R.J., 286.Hartmann, H., 389, 392.Hartree, D. R., 57, 63, 64.Hartridge, H., 300.Hartung, W. H., 141.Harwood, H. F., 282.Hashimoto, A., 224.Hwlam, J., 280.Hasler, M. F., 379.Haslewood, G. A. D., 207, 216.Hassel, O., 398, 415, 420.Hasselt, J. F. B., 148.Hatcher, W. H., 162.Hatt, H. H., 182.Hauer, E., 282.Hauptmann, H., 134, 409.Haurowitz, F., 214.Haworth, W. N., 167, 168, 175, 196,Hayashi, H., 225.Hayashi, M., 190.Hazel, F., 286.Head, F. S. H., 107.Heczko, T., 292.Heesch, H., 387.Heide, F., 397.Heilbron, I. M., 157, 158, 159, 198,210, 212, 213.Heine, U., 373.Heisenberg, W., 60, 354.Heitler, W., 26, 67, 70, 111.Helfenstein, A,, 150.Hellstrbm, H., 250.Helmholtz, L., 367.Helwig, G. V., 104, 409.Henderson, M.C., 350, 361.Hendricks, S. B., 382, 395, 409, 410,Hengstenberg, J., 416, 422.336.419, 420.Henry, T. A., 191.Henstock, H., 276.Henville, D., 279.Hepburn, T. F. G., 286.Hepp, H. J., 47.Herbert, R. W., 168, 171, 175, 336.Herbert, W., 395.Herfeld, A., 281.Hergenrother, R. C., 365, 379.Herrk, H. T., 318.Herrmann, K., 410, 422, 426, 428.Herschkowitz, E., 397.Hertel, E., 408, 416, 421, 422.Herz, W., 19.Herzberg, G., 65, 67, 68, 69, 70, 76.Herzen, G. M., 353.Hess, K., 426.Hess, R., 254.Hesse, O., 220, 221.Hettche, O., 313.Hettich, A., 412.Heuberger, A., 140.Heuck, K., 219, 220, 224.Hevesy, G. von, 274, 344, 384.Hewett, C. L., 207, 214, 339, 340, 343.Hey, M. H., 406, 411.Heyn, W., 134.Heyrovskf, J., 268.Hickinbottom, W.J., 186.Hickman, K., 54.Hieger, I., 339.Hilbck, H., 277.Hilbert, G. E., 420.Hildebrand, J. H., 26, 27.Hill, J. W., 425.Hill, R., 250.Hill, S. E., 287, 314.Hill, T. G., 315, 316.Hill, W. L., 410.Hillemann, H., 214, 424.Hilpert, S., 403.Hiltner, W., 289, 291, 296.Himus, G. W., 302.Hinshelwood, C. N., 40, 51, 54, 55,Hippel, A. von, 370.Hirabayashi, O., 290.Hirai, K., 307.Hirakata, T., 221, 226.Hirata, M., 429.Hirsch, P., 167, 296.Hirst, E. L., 167, 168, 169, 171, 172,175, 336, 337.Hirst, M. C., 427.Hitchcock, C. S., 415.Hoagland, D. R., 320.Hoar, T. P., 38.Hoard, J. L., 105, 408, 409.Hobbie, R., 274.Hock, H., 164.Hodges, F. W., 54.Hodgson, W. R., 19.Hglzer, H., 283.111INDEX OF AUTHORS' NAMES. 43%Honigschmid, O., 82, 83, 84, 85, 86,87, 91.Hosli, H., 214.Hofeditz, W., 46.Hoff, R.W., 277.Hoffer, M., 165.Hofmann, E., 299.Hofmann, U., 395, 406.Hofmann, W., 375, 397.Hohlweg, W., 341.Holbert, J. R., 325.Holgersson, S., 382.Holley, K. T., 321.Holmes, J., 278.Holmqvist, A., 289.Holschneider, F. W., 232, 833.Holt, 31. L., 289.Homeyer, A. H., 176.Honda, K., 390.Hopkins, E. W., 327.Hopkins, S. J., 418.Horiuti, J., 29, 30, 33, 80.Horn, E., 47.Horn, F., 333.Horn, K. R. van, 391.Homer, G. K., 308.Horrobin, S., 291, 396.Horstmann, C. A. I,., 303.Hossfeld, R., 204.Hostetter, J. C., 384, 290.H o d , A. L., 177.Hovde, F. L., 51.Howis, C. C., 178.Howk, B. W., 325.Howland, F., 162.Hubers, P.J., 337.Huckel, E., 36, 46, 69, 73.Huckel, W., 208.Hiihn, W., 109.Hiirbin, M., 137.Huffmann, J. R., 297.Huggins, M. L., 409.Hughes, A. H., 54, 427.Hughes, E. D., 100.Hughes, W. S., 286, 286.Hultgren, R., 396.Hume-Rothery, W., 385.Hund, F., 57, 60, 65, 67, 69, 71, 74.Hunt, H., 18.Hunter, A., 333.Hunze, R. B., 290.Hurd, L. C., 282.Husemann, E., 396.Hussey, S. C., 266.Hutchison, A. W., 24.Hutchison, W. K., 52.Hutino, K., 426.Etylleraas, E. A, 62.Iball, J., 421.Ibmz, J., 293.Ida, M., 225.Ihara, S., 223.Iimori, S., 345.Ing, H. R., 235, 236.Inge, (Miss) L., 370.Tngham, J. W., 286.Ingold, C. K., 42, 52, 177, 178, 181.Inhoffen, H. H., 210, 211.Inosemzev, S. L., 324.frineu, D,, 194.Irmisch, G., 208.Tser, M., 19.Ishikawa, F., 293.Ito, T., 405.Ivanfieva, E.G., 293.Iwasaka, K., 307.Jackson, D. A., 77.Jackson, R. W'., 334.Jackson, W. W., 406.Jacob, K. D., 410.Jacobi, H., 106.Jacobi, R., 206, 215.Jacobs, W. A., 340.Jacobson, H. G. &I., 321, 327.Jacobson, R. A., 133, 134, 135.Jaeger, F. M., 110, 389, 392, 416, 417Jaenckner, W., 126.James, R. W., 415.James, T. C., 130.Jander, G., 294, 295, 296.Jansen, B. C. P., 337.Janson, S. E., 266.Janssen, G., 326.Jay, A. H., 365, 380, 381, 383.Jefferson, M. E., 410, 419.Jeglinski, H., 277.Jenkins, F. A., 89, 347, 348.Jenny, H., 327.Jensen, M. A., 292.Jessop, G., 125.Jette, E. R., 385.Jeu, K. K., 52.Jevons, 57, 77.Jewell, W., 335.Jilek, A., 280.Jimeno, E., 293.JoffB, (Mme.) A., 361.JoffB, A.F., 361.Johansson, C. H., 365.Johnson, J. R., 180, 184.Johnson, R. C., 65.Johnson, S. W., 337.Johnson, T. H., 358.Johnson, W. R., 47.Jolibois, P., 285, 292.Joliot, F., 352, 354, 356, 358.Joly, J., 345.Jones, A. O., 278.Jones, G., 86, 286, 295.Jones, G. C., 277.Jones, G. G., 125.Jones, P., 392440 INDEX OF AUTHORS' NAMES.Jones, P. T., 292.Jones, W. M., 393.Jost, W., 130, 384.Jbzefowicz, E., 17.Juliard, A., 300.Jullien, P., 181.Jung, W., 293.Junge, C., 296.Juza, R., 385.Kabat, E. A., 283.Kading, H., 96.Kahlenberg, L., 289.Kahler, H., 288.Kallman, H., 347.Kallmann, 77.Kallmann, O., 159.Kameda, T., 285, 296.Kamienski, B., 289, 292.Kanaoka, Y., 224.Kanematsu, T., 225.Kantor, T., 293.Kantzer, 49.Kapeller-Adler, R., 316.Kapfenberger, W., 82, 87.Kaplan, B.B., 286.Kapustinski, A. F., 367.Karagunis, G., 265.Kargin, V. A., 294.Karlik, (Miss) B., 418.Karrer, P., 145, 146, 147, 148, 149,150, 152, 153, 157, 167, 168, 171,173, 335.Karsten, B. J., 129.Kassner, J. L., 290.Katagiri, H., 318.Katz, J. R., 426.Kaufmann, H. P., 144.Kawai, S., 276.Kaya, S., 389.Kaye, A. L., 96.Kazanski, B. A., 142.Keggin, J. F., 105, 111.Keilin, D., 250.Keilling, J., 311.Keim, R., 132.Kelland, N. S., 51.Keller, W., 162.Kempter, K., 87.Kemula, W., 273.Kendall, A. I., 308.Kendall, J., 89.Kern, W., 425.Kerridge, (Mrs.) P. M. T., 286.Ketelaar, J. A. A., 398, 399, 409.JLharasch, M.S., 53, 166.Kidd, H. V., 178.Kiesslhg, W., 329.Kikoin, I., 361.Kikuchi, S., 83.Kilpatrick, M., jun., 19.Kimball, G. E., 50, 53.Limura, K., 373.Cindstrtim, A. L., 385.Cindt. H.. 410.Gng,'A. s., 348.Gng, C. V., 23, 53.Cing, H., 99, 199, 208, 213, 215, 341,342, 423.Gng, J.. G., 298.Kinnersley, H. W., 338.Kinsey, B. B., 30, 351.Kipphan, H., 210.Kipphan, K. F., 300.Kirby, J. E., 135.Kirkbride, F. W., 48, 54.Kirkwood, J. G., 20, 22.Kirsch, G., 353.Kiss, A. von, 52.Kitahara, K., 318.Klanhardt, F., 264.Klassen-Nekludova, 31. V., 373.Kleiderer, E. C., 256.Klein, G., 277.Klemen, R., 426.Klement, R., 410.Klemm, W., 102, 106.Klingerhoefer, W. C., 46, 50.Klinkott, G., 116.Klockmann, R., 285.Kluyver, A.J., 318.Knauer, F., 78.Knoke, S., 415.Knowles, F., 325.Kobe, K. A., 298.Kobeko, P. P., 370.Kobel, M., 330.Koch, F. K. V., 16.Koch, I., 410.Koberich, F., 386.Kogl, F., 196, 197, 198.Koelsch, C. F., 184.Koenig, F. O., 37.Kording, P., 342.Kbster, W., 389.Kostlin, (Mme.), 326.Kofler, L., 277.Kogsrt, H., 290, 202.Kohler, E. P., 164.Kohlschutter, H. W., 425.Kolbe, A., 242, 247.Kollek, L., 134.Koller, G., 222.Kolthoff, I. M., 284, 285, 291, 294,295, 296.Komkek, K., 282.Kommes, C. E., 138.Kon, G. A. R., 214, 343, 424.Kondo, H., 244,247,248,249.Konig, F., 326.Konovalovs, R., 239.Kopfermann, H., 77.Kopsch, U., 46.Kordatski, W., 285.Hordes, E., 382INDEX OF AUTHORS’ NAMXS. 441ICorenman, I. M., 280, 302.Kornfeld, L., 192.Koschara, W., 159.Koser, S.A., 308, 309.Kostuk, E., 278.Kostytschev, S. P., 307.Kotake, Y., 332.KOZU, E., 406.KGzu, S., 406.Kracek, F. C., 409.Kraft, 327.Kraft, K., 169, 171.KrajEinovid, M., 141, 279.Kramers, H. A., 22.Kratky, O., 27, 380, 426.Kratschmer, L., 326.Kraus, C. A., 24.Krause, A., 403.Krauss, F., 107.Krebs, H. A., 332, 333.Krefft, 0. T., 128.Kkpelka, J. H., 83.Kringstad, H., 415.Krishnan, K. S., 409, 419, 420, 421.Krishnaswami, K. R., 82.Kronig, 68, 77.Kronig, R. do L., 49.Krueger, A. C., 289.Kriiger, F., 391.Kriiger, K. H., 96.Krueger, P. A., 177.Kruger, P. G., 77.Krug, H., 131.Krumholtz, P., 282.Krummacher, A. H., 422, 428.Ksanda, C. J., 420.Ku, Z. W., 80.Kubasowa, W., 19.Kiihas, E., 246, 250.Kuhne, J., 79.Kursten, 191.Kuffner, F., 239, 249.Kuhn, G., 303.Kuhn, P., 146.Kuhn, R., 145, 147, 148, 149, 150,151, 152, 153, 154, 155, 156, 159,160, 165, 250, 262, 264, 336.Kuhn, W., 366.Kultzscher, M., 322.Kung, T.T., 356.Kunze, P., 358.Kurdjumov, G., 404.Kuriyama, S., 426.Kurtschatov, B. V., 370.Kurtschatov, I. V., 370.Kussmann, A., 389.Kwasnik, W., 91.Laar, J. J. van, 15, 26.Laass, F., 128.Labrousse, F., 327.Ladeck, F., 249.Lahiri, T. K., 105.Laki, K., 343.Lamb, A. B., 299.Lambrey, M., 303.LaMer, V. K., 17, 22, 23, 25.Lammed, 0. M., 285.Lang, E. P., 287.Lang, R., 280.Lange, 363.Lange, E., 18, 284.Lange, J., 166.Lange-Pozdeeva, I. P., 311.Langer, R., 206.Langer, R.M., 354.Langlois, D. P., 177.Langmuir, I., 21, 54, 122, 297.Lannung, A., 17.Lanzing, (Miss) J. C., 296.Lapicque, C., 362.Larson, A. T., 14.Idamson, E., 15, 16, 19, 23.Lasarev, W., 347.Latimer, W. M., 345, 349.La Tour, F. D., 418.Laughlin, K. C., 177.Lauritsen, C. C., 352.Lauterbach, H., 291.Lautsch, W., 46.Laves, F., 387, 388, 393, 394, 410.Lavin, G. I., 51.Lavrov, I. N., 293.Law, D. J., 291.Lawrence, E. O., 30, 350.Lawson, W., 340.Lea, D. E., 354.Leach, R., 327.Lease, E. J., 140.Lebeau, P., 89, 116, 131.Lebedev, S. V., 142, 144.Le Blanc, M., 361, 389.Lederer, E., 150, 153.Lederer, H., 187.Leder-Packendorff, L., 139.Leermakers, J. A., 47, 52.Le FBvre, R. J. W., 279.Leffmann, H., 278.Legard, A.R., 52.Lehl, H., 92.Lehrman, L., 283.Leighton, P. A., 48, 54, 297.Leithe, W., 249.Lemaitre, G., 359.Lemberg, R., 255, 316.Lemcke, W., 395.Lemon, J. M., 275.Lemon, J. T., 102.Lenel, F. V., 417.Lennard-Jones, J. E., 63, 66, 69, 76,111.Leo, W., 362.Leopoldi, G., 283.Le Rolland, P., 372.Le ROUX, P., 409442 INDEX or AUTHORS’ NAMES.Lesslie, (Miss) AT. S., 357, 258, 261,Leuchs, H., 262.Levi, A. A., 189, 190.Levi, G. R., 393.Levin, B., 266.Levin, I., 400.LBvy, (Mme.) J., 181.Lewis, B., 52.Lewis, G. N., 28, 29, 30, 31, 32, 34,Lewis, H. B., 332, 334.Libby, W. F., 345.Lichtenstadt, L., 194.Lieben, F., 309.Ziebhafsky, H. A., 62.Lifschitz, I., 290.Linde, J. O., 389.Lindemann, R., 426.Linderstrom-Lang, K., 19.Lindner, A., 403.Lineweaver, H., 305, 306, 308.Linggood, F.V., 311.Linhard, M., 18.Linser, H., 277.Linstead, R. P., 53.Lipmann, F., 328.Lippert, L., 398.Lipson, H., 98, 410.Little, E., 291.Liversedge, S. G., 278.Livingood, J. J., 77.Livingston, M. S., 30, 360.Ljunggren, G., 299.Lloyd, (Miss) D. J., 427.Lowe, L., 309.Lohmenn, K., 327, 328.London, F., 15, 70, 111.Lonsdale, (Mrs.) K., 414, 421.Lorenz, R., 26.L’Orsa, F., 147.Lotmar, L., 80.Low, G. W., 289.Lowig, E., 326.Lowry, T. M., 105, 125, 212.Lubarskaja, L. S., 322.Lucas, H. J., 302.Luttringhaus, A., 211, 213.Luft, F., 115.Luis, (Miss) E. M., 185.Lumdegardh, H., 327.Lungulescu, E., 307.Lurie, E., 14.Lurie, S., 163.Lustgarten, S., 280.Lux, A.R., 177.Lyle, A. K., 293.263.Lettr6, H., 205, 210.64, 71, 349, 351.McAlevy, A., 163.McCrea, G. W., 417.MaeDonald, G. D., 297,Macdonald, It. T., 2!1, 30, 3 1 .MacDougall, B. H., 291.McDougall, J., 63.McEachern, D., 328.McElvain, S. M., 140.Macfarlane, A., 15, 16.McFarlane, A. S., 287.MacGillivray, W. E., 302.McGuinness, A., 225.McHaffie, I. R., 14.Machatschki, F., 382, 405.McHatton, L. P., 125.MacInnes, D. A., 285, 286, 29-3.McKay, W. B., 183, 184.McKellar, A., 348.McKenzie, A., 178, 184, 185, 264.McKeown, A., 52.Maclagan, N. F., 286, 287.McLonnan, J. C., 77.Maclennan, W. H., 21.McMahon, E., 256.McMath, (Miss) A. M., 932.MeMorris, J., 131.McNab, M. C., 53, 166.McNair, L. C., 301.MacNevin, W.M., 85.Macoun, J. M., 277.Madelung, E., 368.Mahlmann, K., 107.Magistris, H., 335.Maier, A., 140.Mainz, H., 165.Mairlot, E., 293.Maiweg, L., 204.Majer, V., 269.Makgill, R. H., 300.Makower, B., 62.Malaprade, L., 166.Malkin, T., 418.Malone, G. B., 276.M.alvea, B. B., 288.Manderscheid, H., 333.Manegold, E., 295.Mann, F. G., 109.Mannich, C., 137.Manske, R. H. F., 315.Marcon, J., 296.Margaria, R., 53.Maroney, W., 31.Marrian, G. F., 216, 341.Martin, A. J. P., 412.Martin, A. R., 20.Martin, J. C., 320.Martin, W. M. K., 300.Mertius, C., 168, 204, 205.Marvel, C. S., 135, 266.Marwick, T. C., 426, 427.Masaki, K., 290, 293.Mason, C. F., 23.Matano, C., 384.Methewson, C. H., 386.Matthiii, R., 375.Mauthner, F., 187, 221INDEX OF AUTHOICS’ NAMES.443Mavcrick, C: ., 88.May, K., 428.May, 0. E., 318.Mayer, J. E., 367.Mayo, F. R., 53, 166.Maze, P., 320.Maz6, P. J., jun., 320.Mazzuchelli, A., 289.Mecke, R., 80.Medvedev, S. S., 163.Meek, C. A, 406.Meer, M., 47.Megaw, (Miss) H. D., 365, 380, 381,389, 401.Mehl, E., 406.Mehl, R. F., 386.Meier, R., 327.Meiklejohn, A. P., 338, 339.Neinert, R. N., 46.Meisel, K., 399.Meisenheimer, J., 259, 260.Meister, R., 161.Meitner, (Frl.) L., 353, 356.Meksyn, D., 354.Mellor, D. P., 400.Melville, H. W., 51, 53.Mongdehl, H., 324.Menke, H., 27, 428.Menschick, W., 201.Menschikov, G., 240, 241.Menzel, D. H., 348.Menzel, W., 90, 96, 116, 128, 131,Menzies, R. C., 103.Mertens, W.K., 313.Messiner-Klebermass, L., 330.Metcalf, G. M., 287.Meuwsen, A., 86.Meyer, G., 287.Meyer, J., 102.Meyer, K., 324.Meyer, K. H., 101.Meyer, L., 301.Meyerhof, O., 327, 328, 329.Meyers, E. L., 23.Michailov, G., 373.Micheel, F., 169, 171.Migita, M., 182.Milas, N. A,, 163.Miller, W. S., 426.Milligan, W. O., 404.Mills, A. K., 184.Mills, W. H., 100, 261, 263.Minaeff, M., 264.Mitchell, D. M., 278.Mitchell, D. T., 135.Mnich, E., 325.Moller, E. F., 165.Moelwyn-Hughes, E. A., 61, 52.Monch, G., 361.Moffitt, W. G., 280.Mohr, O., 277.Moissan, H., 132.132.Moles. E., 86, 87.Monnier, R., 281.Monossohn, A., 18.Montgomery, C. W., 54.Moore, T., 336.Moore, T. S., 101.Morales, R., 277.Morf, R., 147, 150, 157, 167, 168,335.Morgan, G.T., 100, 101, 106, 110,111, 138, 241, 276.Morgan, J. L. R., 285.Morley, J. F., 167.Morris, W. C., 127.Morrison, A. I,., 213.Morrison, J., 286.Mortimer, F. S., 277.Morton, C., 286, 287, 292.Morton, R. A., 157, 158, 169.Mosca, C., 299.Moser, F. H., 182.Moser, L., 280.Mosley, V. M., 410.Motzoc, (Mlle.) M. D., 280.Moureu, C., 238.Mousseron, M., 316.Moycho, W., 313.Moyer, W. W., 178, 261.Muchlinsky, W., 295.Mudrak, A., 314.Mugge, O., 376.Muller, A., 115, 366.Mueller, E., 85.Miiller, Erich, 290, 291, 292, 293,Muller, F., 286, 287.Muller, H., 282.Mueller, J. H., 334.Mugdan, M., 300.Mulders, E. M. J., 303.Muller, F. M., 311, 312.Mulliken, R. S., 56, 65, 66, 67, 69,Mumford, S.A., 418.Murooka, T., 293.Murphy, G. M., 28, 348.Murray-Rust, D. M., 295Murschhauser, €I., 299.Musgrave, F. F., 40.Muzhe, 0. T., 386.Myles, J. R., 184, 185.294.70, 71, 72, 73, 74, 75, 111.Nadler, E., 194, 195.Nagai, T., 282.Nahmias, M. E., 380.Nakamiya, Z., 210.Nakatsuka, Y., 264.Name, R. C. van, 290, 291Narita, S., 249.Nasledov, D., 362.Natelson, S., 186.Nathan, W. S., 43444 INDEX OF AUTHORS' NAMES.Satta, G., 397, 406, 415.Neddermeyer, S. H., 356.Neff, H., 422.Negelein, E., 250.Nemenov, L., 362.Nernst, W., 50.Neuberg, C., 277, 310, 330.Neuburger, M. C., 388, 389.Neuman, E. W., 23, 24, 411.Neumann, F., 248.New, R. G. A., 122.Newitt, D. M., 49, 297.Newson, H. W., 352.Nicholls, J. R., 279.Nichols, M.L., 281.Nicol, H., 299.Nicolski, B. P., 293.Niederl, J, B., 186, 270.Niederlander, K., 21 1.Niel, C. B. van, 311, 312.Nielson, N., 317.Niemann, J., 192, 193, 195.Nieuwenkamp, W., 398, 399.Nieuwland, J. A., 133, 134, 135.Niggli, P., 387, 405.Nightingale, G. T., 325.Nilsson, R., 328.Nishiyama, Z., 390.Nix, F. C., 361.Noack, K., 251.Noetzel, O., 279.Noll, F. H. W., 427.Nonomura, S., 222.Nordb~), R., 287.Norkina, S., 239.Norris, W. V., 80.Norrish, R. G. W., 48, 50, 51, 54,Noskov, M., 361.Notthafft, A., 153.Nowacki, W., 387.Noyes, W. A., 79.Nuka, P., 282.Nygaard, E. M., 164.Nylen, P., 94, 409.79.Oberembt, H., 247.Obreimov, I. W., 375, 419.O'Brien, H., 287.Occhialini, G. P. S., 355, 356, 357,O'Daniel, H., 396.ohman, A., 385.Olander, A., 392.Oftedal, I., 397.Ogg, R.A., 54.Ohta, Z., 225.Oliphant, M. L. E., 30, 33, 350, 351.Olsen, J., 308.Olson, A. R., 31, 49.Omaki, T., 225.Onsager, L., 22,358.Oppenauer, R., 174, 175.Oppenheimer, J. R., 357.Orcutt, F. O., 293.Orekhov, A., 181, 185, 239, 240, 241.Orndorff, W. R., 281.Ornstein, L. S., 89, 348.Orowan, E., 366, 376.Osswald, E., 389.Ott, E., 298, 299, 400, 418.Otto, I. G., 271.Owen, B. B., 17.Owen, E. A., 380, 384.Oxford, A. E., 319.Packendorff, K., 139.Pacsu, E., 33.Page, I. H., 201.Pahl, M., 344.Paine, S. G., 311.Palacios, J., 420, 422.Palmer, L. S., 146.Pamfilov, A. V., 293.Paneth, F., 46, 47.Panizzon, L., 163.Paoloni, C., 292, 295.Paris, R., 166.Parker, H., 398.Parker, H.C., 289.Parkes, A. S., 341.Parks, L. R., 288.Parravano, N., 392.Partington, J. R., 93.Partridge, H. M., 287.Pass, A., 282.Passerini, L., 397.Passinski, H. A., 285.Passmore, R., 338.Patterson, H. S., 88.Patterson, W. I., 259.Patzsch, H., 277.Pauli, W., 76.Pauling, L., 45, 57, 67, 72, 76, 93,96, 97, 98, 105, 106, 396, 405, 408,410, 415.Paulus, R., 396, 400.Pavolini, T., 283.Paxinos, S. A., 319.Peachey, S. J., 125, 261.Pearce, R. H. H., 365.Pearson, J., 279.Pearson, T. G., 47, 93.Pease, R. N., 51.Peel, J. B., 83.Peerlkamp, P. K., 49.Peisker, H., 392.Peltzer, J., 280.Penzoldt, 279.Pepe, R. O., 225.Percival, E. G. V., 169, 171, 175,336.Perkin, A. G., 217.Perlitz, H., 388, 391INDEX OF AUTHORS’ NAMES.445Pernot, R., 181.Perquin, L. H. C., 318.Perrin, J., 354.Perrott, C. H., 185.Perucca, E., 364.Peschel, J., 148.Peters, M. A., 191.Peters, R. A., 338, 339.Peterson, B. H., 23.Peterson, W. H., 293.Peterson, W. R., 165.Petit, A., 294.Petrfi, F., 214.Pettet, A. E. J., 276.Petzold, W., 124.Pfau, A. S., 218, 219.Pfeiffer, G., 222.Pfeiffer, P., 263, 264.Pfutzer, G., 320.Pfundt, O., 295, 296, 299.Philipp, K., 353, 356.Phillips, F. C., 395.Phillips, H., 427.Phillips, J. W. C., 418.Philpot, J., 37.Piatti, L., 373.Pickett, L. W., 420, 421.Pickett, T. A., 321.Pickup, L., 380, 384.Pictet, A., 241.Pierce, W. C., 428.Pieters, H. A. J., 299.Pietsch, E., 92.Pikl, J., 247.Pincussen, L., 285.Pines, C.C., 278.Pinkard, F. W., 107, 108, 408.Pinkhof, J., 291.Piper, S. H., 418.Pire, L. R., 87.Pirie, N. W., 333.Pischinger, E., 326.Pizzi, C., 257.Placinteanu, J. J., 354.Placzek, G., 72, 79.Plant, S. 0. P., 98.Plate, A. F., 142.Platt, B. S., 288.Platz, H., YO.Plesset, M. S., 357.Pluschnik, E., 404.Podbielniak, W. J., 302.Poe, C. F., 314.Poethke, W., 294, 296.Pohland, E. 300.Polanyi, M., 29, 30, 32, 33, 47, 49.Pollard, A., 418.Pollitzer, F., 14.Poole. H. H., 364.Poole; J. H. J., 345.Pope, C. G., 292.Pope, (Sir) W. J., 109, 126, 261, 266.Popov, S., 23.Porges, N., 317.Porter, M. D., 99.Posnjak, E., 382, 406, 409.Posternak, T., 206.Potts, J. C., 49.Potts, T. T., 295.Poulson, E., 157.Pouzergues, J., 281.Powell, H.M., 104, 409.Prasad, M., 421.Prescott, C. H., 297.Preston, G. D., 389.Preston, G. H., 107, 108, 408.Preston, J. M., 296.Preuss, E., 397.PrBvost, C., 166.Prichotjko, A., 419.Prideaux, E. B. R., 124, 131, 132.Pring, (Miss) M. E., 290.Prins, J. A., 393.Pritze, M., 94.Procter, R. A., 139.Przibram, K., 375.Ptizyn, B. W., 107.Pugh, W., 293.Pummerer, R., 163.Pycock, E. R., 53.Quehl, K., 263.Quill, L. L., 393.Quimby, S. L., 372.Quin, J. I., 251.Rabinovitsch, E., 56, 64, 76, 79.Rabinovitsch, M., 239.Rae, N., 296.Rains, H. C., 189.Raistrick, H., 318, 319.Ramage, G. R., 232, 233, 234, 424.Ramart-Lucas, (Mme.), 185.Ramsperger, H. C., 52.Randall, H. M., 79.Randall, J.T., 428.Randall, M., 15, 19.Rank, D. H., 29, 80.Raper, K. B., 317.Raper, R., 232, 233, 234, 237, 238H,audnitz, H., 147, 148, 214, 251.Haw, R., 279.Ray, P., 283.Ray, S. N., 167, 337.Rayleigh (Lord), 303.Read, J., 262.Redmond, J. C., 281.Reeve, L., 297.Regener, E., 358.Reggiani, M., 290.Reichardt, H., 47.Reichstein, T., 174, 175.Reid, E. E., 276.Reif, G., 279446 INDEX OF AUTHORS’ NAMES.Reif, w., 283.Reifer, J., 280.Reihlen, H., 109.Reindel, F., 210, 211.Reinfurth, (Frl.) E., 277.Reinhardt, L., 395.Reinicke, R., 387.Reissaus, G. G., 290.Retter, K., 52.Rettger, L. F., 306.Reuning, E., 405.Reuter, F., 250.Rexer, E., 373, 375, 376.Reynolds, R. J. W., 167, 171Rhines. F. N., 386.336.Ricard; P., 315.Rice, F.O., 47, 48.Rice, 0. K., 40.Richards, 85, 86.Richter, D., 162.Rideal, E. K., 54, 427.Rieche, A., 161.Riedl, E., 273.Riehl, N., 96, 395.Riemann, H., 205.Righellato, E. C., 294.Riley, H. L., 167, 281.Rimington, C., 251.Rinck, E., 393.Rinne, F., 375, 427.R!ppel, A., 317.Ritchie, M., 50, 82.Ritschl, R., 77.Rittenberg, D., 29, 32.Rius, A., 293.Roberti, G., 396.Roberts, D. C. V., 40.Roberts, E. J., 288.Roberts, H. S., 284, 290.Robertson, A., 194, 217, 218, 219,Robertson, A. E., 302.Robertson, G. R., 286.Robertson, I. W., 302.Robertson, J. M., 420, 421.Robinson, A. L., 23.Robinson, B. W., 366, 379.Robinson, P. H., 357, 262.Robinson, P. L., 91, 93.Robinson, R., 44, 219, 239, 255, 319,Robinsoii, R.A., 286, 288.Robinson, W. K., 394.Roche, A., 388.Roche, J., 288.Rochow, E. G., 116.Rodda, J. L., 386.Rodebush, W. H., 46, 47, 50.Rodowskas, E. L., 101.Rohl, H., 373, 389.Romer, G. H., 421, 422.Roger, M., 181, 183, 184.Rogers, H. E., 115.222, 224.424.Holfe, A. C., 51.Rollefson, G. K., 49, 50, 54.Roller, P. S., 282, 384.Roman, W., 92.Rona, P., 165.Rooksby, H. P., 428.Rose, W. C., 335.Rosebury, F., 287.Rosen, R., 302.Rosenblatt, F., 109, 110, 408.Rosenbohm, E., 389.Rosenhall, G., 391.Rosenheim, A., 94, 95, 109.Rosenheim, O., 199, 208, 213, 215,341, 342, 423.Rosenthaler, L., 280.Rossi, A., 393, 394.Roth, H., 154.Roth, W. A., 292.Rothhaas, A., 251, 313.Rother, E., 296.Rothrock, H. S., 177.Rothstein, E., 42.Rotschy, A., 241.Rougebief, G., 313.Roughton, F.J. W., 53.Roussin, A. L., 380.Roxburgh, H. L., 51.Rudolph, E. A., 212.Rudy, H., 160.Ruff, H., 299.Ruff, O., 90, 91, 96, 115, 116, 128, 129,131, 132.Ruhemann, B., 414.Ruhernann, M., 381, 410.Rummel, K. W., 42.Russell, A., 225.Russell, H. N., 59.Russell, W. W., 275.Rutgers, J. J., 342.Rutherford, (Lord), 30, 360, 351, 355,Rutterford, G. V., 99.Ruyssen, R., 300, 304.Ruzicka, F. C. J., 214.Ruzicka, L., 137, 207, 209, 212, 213,Ryder, H. M., 397.Rydon, H. N., 63.Rygielski, J., 273.RyBSnek, A., 280.Rysselberghe, P. van, 22.356.214, 215.Sachs, G., 376, 380, 386.Sachse, H., 361.Sachsse, H., 41.Sachtleben, R., 83, 84, 85, 86, 91.Saenger, H., 108.Sakaguchi, K., 318.Sako, S., 260.Salrurada, I., 426.Salazar, M.T., 87HDXX OM AUTHORS' NAMES. 44%Salkind, J . S., 144.Salley, D. J., 51.Salmon-Legagneur, F., 185.Salomon, H., 147, 148, 167, 168.Salstrom, E. J., 36.Salvia, R., 397.Samant, K. M., 210.Samec, M., 426.Sand, H. J. S., 292.Sandor, G., 313.Sandulesco, G., 342.Santos, A. C., 245, 247.Santos, J. A., 381.Sasaki, K., 389.Sauerwald, F., 393.Saunders, F., 309.Saunders, F. A., 69.Sauter, E., 379, 425.Sazerac, R., 281.Scalione, C. C., 299.Scatchard, G., 19, 22, 27.Schaafsma, A., 49.Schacherl, R., 421.SchLifer, K., 383.Schall, B. M., 291.Schattenstein, A. J., 18.Scheffer, F. E. C., 303.Scheffler, B., 221.Scheloumov, A., 307.Schenk, P.W., 50, 90.Scherb, E., 298, 299.Scheuer, O., 86.Schiebold, E., 379, 405, 406.Schiedt, E., 389.Schikorr, G., 402.Schimmer, F., 311.Schischakov, N. A., 293.Schlapfer, P., 299.Schlatter, C., 88.Schleede, A., 109, 110, 394, 408, 410.Schlichting, O., 206, 315.Schlosser, C., 384, 293.Schmalfuss, K., 324.Schmid, E., 373.Schmidt, E., 239.Schmidt, E. H., 144.Schmidt, H., 54.Schmidt, O., 303.Schmidt, W., 389, 410.Schmidt-Nielssen, S., 157.Schnaase, H., 396.Schneider, E., 251.Schneider, K., 408, 422.Schonberger, W., 202.Schonfeld, H., 374.Schopf, C., 219, 220, 224, 239.Schopp, K., 150, 157, 167, 168, 335.Scholl, A. W., 24.Scholtz, M., 249.Scholtze, R., 247.Scholz, E., 203.Schonweltl, 36 I.Schlesinger, N., 19.Schopel, H., 361.Schoonover, I.C., 293.Schoonover, J., 286, 28'7.Schoorl, N., 367.Schorstein, H., 294, 296.Schottky, W., 361, 364.Schrarnek, W., 295.Schroder, H. J., 375.Schroder, M., 306.Schroder, O., 386.Schroeter, G., 194.Schtern, V. J., 144.Schubnikov, A., 376.Schubnikov, L. W., 375.Schuck, B., 282.Schiiler, H., 77.Schuette, H. A., 140, 279.Schutze, W., 375.Schuftan, P., 303.Schulenburg, W., 202.Schultze, R., 361.Schulz, G., 285.Schulz, J., 37.Schulze, A., 393.Schulze, G. E. R., 400.Schulze, R., 284, 285.Schumacher, H. J., 52, 53.Schumb, W. C., 24, 300.Schwab, G. M., 46.Schwartz, A,, 208.Schwartz, E., 284.Schwartz, K., 284.Schwarz, E., 411.Schwarz, K., 19, 293.Schwarz, R., 95.Schwarzenbach, G., 168, 286, 287.Schweitzer, E., 272.Schwiersch, H., 386.Scripture, E.W., 83.Searle, N. E.', 258.Seeman, H. J., 359.Segrb, E., 77.Seibert, F. M., 302.Seifert, H., 408.Sekito, G., 389.Seljakov, N., 373.Sella, G., 295.Selwood, P. W., 31.Semerano, G., 370.Semichon, L., 278.Sennewald, K., 53.Sessions, A. C., 330.Seuferling, F., 92.Seumel, G., 406.Severyns, J. H., 300.Seward, R. P., 24.Sexl, T., 354.Sfiras, J., 181.Shaeffer, W. E., 83.Shah, C. C., 93.Shambaugh, N. F., 332.Sharp, P. I!., 291.Shaw, (Miss) k'. R., 258448 INDEX OF AUTHORS’ NAMES.Shead, A. C., 281.Shearer, G., 116.Shenstone, A. G., 77.Sheppard, A. F., 184.Sherman, A,, 32, 367.Sherman, J., 45, 53, 410.Shida, H., 410.Shive, J. W., 320, 321.Shoppee, C.W., 44.Shriner, R. L., 276.Shukla, S. N., 303.Shukov, I. I., 288, 289.Shull, G. O., 92.Sidgwick, N. V., 56, 93, 98, 99, 100,101, 103, 104, 110, 111, 114, 115,120, 256, 414.Siedel, W., 251, 253.S!eg, L., 408.Silberschatz, S., 300.Simeon, F., 272.Simmonds, C., 277.Simmons, W. H., 279.Simon, E., 310.Simon, F., 415, 429.Simonart, P., 319.Simons, J. H., 47.Simonsen, J. L., 103.Simpson, J. C. E., 198, 210, 213.Sinchir, H. M., 339.Sinozaki, H., 293.Sixt, J., 300.Skinner, W. W., 300.Skobeltzyn, D., 358.Skoog, I?., 131.Slagle, F. B., 418.Slanina, F., 318.Slater, J. C., 20, 60, 63, 64, 72, 368.Sloan, D. H., 350.Smekal, A,, 373, 374.Smiles, S., 188, 189, 190.Smith, C. R., 241.Smith, D.W., 386.Smith, E. R., 28.Smith, F., 169, 171, 175, 337.Smith, G., 319.Smith, (Miss) I. A., 264.Smith, J. A. B., 418.Smith, J. C., 53.Smith, L., 278.Smith, M. P., 279.Smith, R. A., 186.Smith, R. W., 272.Smith, W. W., 89.Smolczyk, E., 284.Smyth, C. P., 115, 415.Smythe, C. V., 329.Snell, F. R., 52.Snow, C. P., 75.Sobotka, H., 265.Sokolov, S. I., 285.Soltan, A., 352.Sommerfeld, A., 67, 120.Sonn, A., 221.Sorin, P., 372.Sorokin, V., 299.Sorum, C. H., 286.Sosman, R. B., 400.Sosnick, B., 15.Soule, B. A., 282.Spacu, G., 281.Spacu, P., 281.Splith, E., 192, 194, 195, 236, 239,242, 247, 249.Spiith, W., 273.Spahr, R. J., 133.Speakman, J. B., 427.Spedding, F. H., 349.Spencer, J. F., 290.Sperling, G. F., 373.Sperling, K., 391.Spinks, J.W. T., 302.Spivey, E., 53.Sprenger, L., 425.Spring, F. S., 157, 198.Spychalski, R., 293.Stacey, M., 175, 336.Stackelberg, M. von, 396, 400.Stadie, W. C., 286, 287.Stadler, A., 214.Staehler, A., 85.Stahl, A. L., 321.Stahly, E. E., 176, 177.Stange, O., 199, 205.Stanley, W. M., 256, 257, 259.Stanner, E., 19.Stansby, M. E., 275.Starkweather, H. W., 87.Staub, EI., 370.Steacie, E. W. R., 162.Stefanovski, V. F., 294.Steiger, B., 283.Stein, C. P., 129.Steineck, H., 165.Steiner, A., 285.Steiner, K., 374.Steinmetz, H., 412.Stellezky, T., 404.Stelling, O., 94, 293, 409.Stenbeck, S., 392.Stenzel, W., 389.Stephan, M., 18.Stern, O., 37, 78.Sternberger, H. R., 182.Stevens, T.S., 187.Stevenson, R. J., 218, 219.Stewart, A. M., 251.Stewart, G. W., 428.Stillwell, C. W., 393, 394.Stock, A., 302, 403, 416.Stockings, W. E., 302.Stoddart, L. M., 91.Stoess, U., 317.Stoll, M., 145, 147, 148.Storch, E. A., 186.Storey, H. EL, 327.Stepp, w., 277INDEX OF AUTHORS’ NAMES. 449Stortenbeker, W., 130.Stoughton, R. W., 256.Stowe, M. V., 18.Strasser, E., 165.Strassmann, F., 14.Straumanis, M., 377, 378.Straus, F., 134.Strebel, E., 14.Striebel, H., 87.Strock, L. W., 408.Struss, E. F., 138.Stuart, N. W., 323.Style, D. W. G., 47, 130.Suciu, G., 281.Sueda, H., 288.Sue, P., 280.Sugden, J. A., 287.Sugden, S., 93, 94, 102, 106.Suneson, R., 316.Susemihl, W., 164.Sutherland, G. B. B. M., 80.Sutton, L.E., 43, 104, 111, 122.Svensson, E., 346.Svirbely, J. L., 167, 337.Swanback, T. R., 321, 327.Swearingen, J. S., 297.Swinehart, C. F., 127.Szalay, A., 54.Szalkowski, C. R., 280.Szebelledy, L. von, 289.Szego, L., 106.Szent-Gyorgyi, A., 54, 167, 337, 343.Tabart, A., 181.TBnzler, K. H., 293.Tait, T., 89.Takahashi, T., 147, 145, 310.Takane, K., 402, 406.Takvorian, S., 344.Talmud, D. L., 54.Tammann, G., 429.Tamura, K., 276.Tanaka, K., 310.Tananaev, I., 294.Tanase, Y., 225.Taylor, H. S., 28, 31, 34, 42, 51, 367.Taylor, J. B., 78.Taylor, T. W. J., 40, 130.Taylor, W. H., 405, 406.Teegan, J. A. C., 295.Teichmann, H., 361.Teller, E., 76, 79.Tenniswood, C. R. S., 237.Tertsch, H., 376.Teske, W., 393.Thanheiser, G., 294.Theilacker, W., 260.Theile, W., 374.Theis, E.R., 426.Theobald, L. S., 282.Thewlis, J., 402.Thibaud, J., 418.RZP.-VOL. xxx.Thiel, A., 286.Thiessen, P. A., 367, 418.Thiessen, R., 426.Thom, C., 317.Thomann, G., 207, 209, 213, 215.Thomas, J. S., 83, 84.Thomas, M. D., 300.Thomas, R. W., 140.Thompson, B. J., 287.Thompson, H., 216, 341.Thompson, H. W., 49, 51.Thompson, L. G., 306.Thompson, M. R., 286.Thomson, D. W., 21.Thomson, T., 187.Thon, N., 354.Thorpe, T. E., 278.Thrupp, T. C., 311.Tiedjens, V. A., 322.Tiffeneau, M., 181, 186.Tilk, W., 106.Tillmans, J., 167, 337.Tincker, M. A. H., 324.Tisza, L., 80.Titani, T., 47.Tjabbes, B. T., 95.Todd, A. R., 319.Todd, G. W., 354.Toeniessen, E., 327.Tonnies, B., 198.Tokuoka, M., 269.Tolansky, S., 77.TomiEek, O., 282.Tomita, M., 244, 247, 248.Tongberg, C.O., 165.Topley, B., 30, 395.Tornau, O., 324.Tourtellotte, D., 332.Townend, R. V., 53.Traubenberg, H. R. von, 354.Trautz, M., 300.Treadwell, W. D., 291, 292, 295.Treibs, W., 165.Trillat, J. J., 416, 426.Troberg, B., 293.Tromel, G., 410.Trogus, C., 426.Tronstad, R., 293.Trtilek, J., 283.Truchet, R., 137.Tschesche, R., 200.!lhkamoto, T., 220.Turner, E. E., 257, 258, 161, 263.Twyman, F., 272.Ubbelohde, A. R., 55.Ulich, H., 295.Uemura, T., 288.Uhlenbeck, G. E., 58.Undheim, B., 62.Ungemach, O., 137.Unger, H. J., 79, 80.450 IKDES OF AUTHORS' NAMES.Urban, F., 283.Urey, H. C., 28, 29, 3.'.Urey, H.%I., 348.UrmBnczy, A., 32.Usanovich, M. I., 2991.Usher, F. L., 86.Uspenskaja, L. P., 29'33.Valeur, A., 238.Valla, R. K., 281.Van Arendonk, A. M., 256.Vance, J. E., 53.Vargha, L. von, 173.Vaubel, R., 167.Vaughn, T. H., 133, 134.Vaux, G., 404.Veen, A. G. van, 313.Velculesco, A. J., 291, 293.Velliiiger, E., 288, 289.Velluz, L., 282.Venus-Danilora, E. D., 134.Vering, F., 316.Verkade, P. E., 415.Vernon, E. L., 52.Vernon, W. H. J., 100.Veselovski, B., 368.Vickers, A. E. J., 287.Vigoureux, P., 373.Villars, D. S., 76.Villiger, A., 212.Vleck, J. H. van, 75, 80.Vlodrop, C. ran, 144.Vogel, J. C., 289.Vogt, R. R., 133.Voicu, J., 307.Voigt, A., 124.Volmer, M., 37, 40.Volqvartz, K., 17.Vorlander, D., 277.Vranjican, D., 141.Vreeswijk, J.*4., jun., 348.Wada, M., 333.Waddington, G., 52.Wagenaar, M., 280.Wagner, C., 361.Wagner, G., 398, 416.Wagner-Jauregg, T., 54, 159, 160.WaEl, M. H., 47.Wahl, W., 100.Waibel, F., 271, 361, 364.Waldbauer, L., 273.Waldeland, C. R., 142, 259.Waldmann, H., 424.Walker, H. L., 389.Walker, O., 150, 164, 167, 335.Walker, 0. J., 303.Waller, I., 62.Wallis, E. S., 178, 261, 265.Walther, A., 370.Walton, E. T. S., 350, 352.Warburg, O., 158, 160, 250.Ward, A. M., 281, 282.Ward, G. E., 318.Wardlaw, W., 107, 108. 111. 408.Ware, J. O., 336.Warming, E., 21.Warren, B., 405, 429.Warren, L. A, 188, 189, 190.Wartenberg, H. von, 116.Washburn, E. W., 28, 425.Watanabe, IN., 221.Watannbe, T., 410.Waterman, H.I., 1-44.Waters, R. B., 194. 222.Watkin, J. E., 326.Watkins, J. S., 50.Watson, H. B., 43.Watson, H. E., $8.Watts, v. M., 323.Weber, H. H., 426Webster, E. T., 15!).mTebster, W. L., 366.Wedekind, E., 402.Weerts, J., 380, 318'3.Wehner, G., 389.Wehrli, H., 150.Weidinger, A., 426.Weidlich, H. A., 416, 341.Weigerts, TV., 54.Weiler, G., 286,Weill, P., 181.Weinhold, R., 264.Weisberg, H., 283.Weiser, H. B., 401.Weiss, L., 291.Weiss, R., 264.Weissbach, K., 244.Weissberger, A., 166.Weisse, G. von, 291.Weissflog, J., 324.Weitendorf, K. F., 296.Weizel, W., 64, 69, 177.Wellmann, M., 394.Welo, L. A., 403.Went, F. ,4. F. C., 196.Werder, F. von, %lo.Wessely, F., 192, 194, 196.West, C. D., 410.West, J., 113, 381, 405, 406.West, W., 80.WesterhoiY, H., 96.Westerman, B.D., 335.West,gren, A., 380, 390, 392.Westmeyer, H., 77.Westwater, W., 27.Westwood, J. B., 285.Wettstein, A., 150.Wheland, G. W., 45.Whitby, L., 100.White, A., 334.White, H. E., 77.White, M. G., 360.Whitehouse, W., 398INDEX OF AUTHORS' NAMES. 451TWritenack, T. A, 292.Whitmore, F. C., 176, 177.Whytlaw-Gray, R., 89, 91, 348.Wiardi, P. W., 337.Wibaut, J. P., 337.Wick, G. c., 78.Wiedersheim, V., 207.Wiegnnd, W., 147.Wieland, H., 162, 163, 199, 202, 203,204, 205, 206, 207, 208, 215, 310.Wierl, R., 416.Wiesner, 13. P., 341.Wiest, P., 389.Wiester, H. J., 300.Wigner, E., 42, 68, 72.Wilcox, L. V., 291, 293.Wild, W., 88.Jt'ilkinson, D. G., 210.Wjkinson, $1. R., 300.Willard, H. H., 86, 289, 200, 291,Wi1Ie, A., 403.Wille?, E. J. B., 304.Williams, E. F., 418.Williams, J. W., 23, 24, 292.U7illiams, R. D., 402.Williamson, A. T., 51.Willstiitter, R,., 145, 140.Wiliii, D., 395, 406.Wilson, A. H., 361.W-ilson, (Miss) B. &I., 286.Wilson, C. T. R., 359.Wilson, D. A., 418.Wilson, E. B., 290.Wilson, E. D., 363.Wilson, P. W., 293.IVinans, C. F., 141.Windaus, A., 198, 200, 204, 205, 209,Winkler, L. W., 300.Winn, A. G., 52.Winterfeld, K., 231, 232, 235.Wintersberger, K., 83.Winterstein, A., 146, 147, 148, 149,151, 153, 154.Withrow, J. R., 258.Witnier, E. E., 68.Wolbling, El., 283.Wold, K., 20.Wojcik, B., 140, 141.Wolf, L., 293.Wolf, P. M., 395.Wolfe, R. A,, 272.Wolfenden, J. H., 30.Wolk, L. J. van der, 296.Wood, R. W., 46, 377.Wood, S. E., 27.Woodburn, H. M., 177.Wooclhead, M., 89, 348.Woodruff, E. H., 259.Woods, H. J., 425.Woolcock, J. W., 295.Wooster, N., 399.292.810, 211, 212, 213, 264.Wooster, W. R., 411.Wooten, L. A, 2!11, 292.TVormwell, F., 55.Wrann, S., 246, 250.Wrede, E., 46.Wrede, F., 251, 313.Wrigge, F. W., 01, 289.Wright, A. H., 289.Wright, L. O., 277.Wright, N., 79.Wiihrer, J., 279.Wulff, P., 285.Wurm, O., 282.Wustrow, W., 302.Wyart, J., 406.Wyatt, G. H., 107.Wyckoff,R. W. G., 137, 401,410,417.Yakubchik, A. O., 142.Yamaseki, K., 209.Yanagita, AT., 223, 225.Yannaquis, N., 417.Yano, Y., 247.Yant, W. I?., 302.Yeh, W., 345.Yoe, J. H., 281.Yoshimura, J., 345.Yost, D. M., 92, 96, 131.Young, (Miss) M. W., 101.Young, V. H., 326.Yuan, H. C., 255, 256.Yuching Tu, 378.Young, H. A., 349.Zaayer, (Frl.) $1.. 144.Zacharias, J., 373.Zachariasen, W. H., 73, 97, 105, 113.396, 409, 410, 429.Zakomorny, M., 317.Zambonini, F., 410.Zamotorin, M., 384.Zanstra, J. E., 110, 392, 416.Zartman, W., 259.Zartman, W. EL, 143.Zawadzki, J., 55.Zedlitz, O., 405.Zeile, K., 250.Zelinski, N. D., 139, 143.Zener, C., 62.Ziel, A. van der, 418.Ziegler, G. E., 97, 409, 410.Ziegler, K., 46.Zilva, S. S., 337.Zinserling, K., 376.Zintl, E., 85, 86, 92, 393, 386.Zubrys, A., 153.Zuckerkandl, F., 330.Zwicky, F., 365, 376, 378.Zwikker, J. J. L., 280.Zvegintzov, M., 52.Zvjagintsev, 0. E., 359

ISSN:0365-6217    

DOI:10.1039/AR9333000431

出版商:RSC    

年代:1933

数据来源: RSC