ii. 77 General and Phgsiesl Chemistry. The Quantum Theory of Line Spectra. N. BOHR (Danske Vid. Xelsk. Skrzfter nut. mat. Afd. 1922 4 [Sl 101-118; from Chem. Zentr. 1923 iii 1337; cf. A. 1922 ii 801).-The distinction between arc and spark spectra depending on the ratio of the number of electrons- to the nuclear charge is discussed. A cor- relation exists between the arc and spark spectra of elements of difEerent series of the periodic system. The probability of trans- ition between Merent stationary states of an atomic system is also considered with reference to the author’s theories. The series spectra of helium lithium beryllium and the Stark and Zeeman effects are discussed. G. W. R. The Spectrum of Iodine and the Band Spectrum of Mercury. W. GERLACH (Physikal. Z. l923,24,467469).-The nature of the spectrum of iodine is briefly referred to and reasons are advanced for regarding the band 3460 A.as the electron-afhity continuous band spectrum associated with the iodine atom whilst the band 4800 A. is attributable to the electro-negative iodine molecule. Brief reference is made to the resemblance between the R. FRERICHS (27. Physik 1923 20 170-187).-Particulars are given of the determination of the respective wave-lengths and intensities of R and P series of lines in the bands AX 4006 4063 4280 4328 4380 4650 and 4690 A. in the spectrum of copper excited by the introduction of finely powdered copper into an oxy-hydrogen blowpipe flame. The determination of combination relations amongst lines of the respective series indicates that the bands may be classified as follows bands characterised by the same initial energy level of the emitting molecule (n,=O) 4280 4650; (nl=l) 4006 4328 4690; (n1=2) 4063 4380; bands characterised by the same ha1 energy level of the emitting molecules (na=O) 4280 4006; (na=l) 4650 4328 and 4063; (n2=2) 4690 and 4380.The frequencies yo of the respective zero lines associated wifh &he various bands can be very accuratelr calculated by the formula v0 = 23,311.15 + (1658.81n1 - 44-71n1 )-(1903.48n2 - 37-18n$) in whch n and n2 are quantum numbers characterising the respective initial and final energy levels of the emitting molecules. An empirical formula of the Deslandres type permitting the fre- quencies of the heads of the bands to be very accurately calculated and involving half-quantum numbers and five constants is deduced.Two of the constants are shown to be related to the quantum level-numbers n and n2. The origin of the bands is discussed briefly. The calculated value of the moment of inertia of the emitting moleoule is 0.35 x 10-3 gr. cm.2. This result indicates the presence of hydrogen in the emitting system a result which is partly confirmed experimentally. spectra of iodine and mercury vapour. J . S. G. T. The Band Spectrum of Copper. J. S. G. T. VOL. CXXVI. ii. 4ii. 78 ABSTRACTS OF CHEMICAL PAPERS. Combination Relations in the Band Spectra of the Copper Flame. E. BENGTSSON (2. Physik 1923 20 229-236).- Particulars are given of the determination of the respective wave- numbers and intensities of lines in the bands AX 4689 4649 4280 4005 and 3777 8.in the flame spectrum of copper The lines in the various bands except for seven lines in the band 4328 and five in the band 4062 can be arranged in two series and the deduc- tion of combination relations between the respective series in the manner employed by Hulthitn (A 1923 ii 670) indicates that the bands 4280 4005 and 3777 correspond with the same final con- dition of the molecule whilst the bands 4280 and 4649 correspond with the same initial condition of the molecule and similarly in regard to the pair 4005 and 4689. The frequencies yo of the respective zero lines associated with the various bands can be calculated by the formula v0=23552*4+1743-4n-44.15n2- [1977*5n'-37*2n'2] in which n and n' are whole numbers ranging from 0 to 3 and represent quantum numbers associated respec- tively with the initial and final energy levels of the molecule.J. S. G. T. Spark Spectra of Higher Order associated with Mercury. L. BLOCH and E. BLOCH ( J . Phys. Raditzcm 1923 4 333- 348).-Particulars are given of measurements of the respective wave-lengths and intensities of about 600 classified lines in the spectra of Hgn Hgm and Hgm between wave-lengths 6149.37 and 2224.87 A. excited in an electrodeless discharge in mercury vapour by gradual increase of the exciting voltage. Classification of the lines into the respective orders is effected by reference to the variation of the lengths of the lines as the exciting voltage i s increased. J. S. G. T. The System of Quartets in the Scandium Spectrum and the Periodic System. M.A. CATALAN (Anal. Fb. Quim. 1923 21 464480; cf. A. 1923 ii 104).-The author discusses the application of Kossel and Sommerfeld's alternation law (" Wechsel- satz ") to the spectra of the neutral atoms of potassium calcium scandium titanium vanadium chromium manganese and iron. It is concluded that every element gives variws systems of series but either all odd or all even; one element never gives both odd and even series. A table is given showing the relation to the periodic classification of the various systems of series obtained. Whilst potassium gives doublets only calcium gives simplets and t.ripl& and scandium doublets and quarteta Larger numbers of systems are shown for element6 in the later groups for which however some systems still remain undiscovered. For example in the iron spectrum triplets and quintets are known whilst shplefs septets and nonets remain to be discovered G.W. R. The IiSeries of the Spectrum of Molybdenum. S. TANAKA and J. TSUTSUMI (Mem. CoZZ. Sci Kyoto 1923 7 1-5).- In continuation of previous work (A. 1922 ii 805) the authors have determined the wave-lengths of the four lines in theGENERAL AND PHYSICAL CHEWSTBY. ii. 79 R-series of the spectrum of molybdenum by observations of the second third fourth and fifth order spectra produced by a reflector of mica. The results are in close agreement with the values determined by h i d e and others. The Optical Spectrum of Hafnium. H. M. HANSEN and S. WERNER (Nature 1923 112 90O-W1).-A list is given of the strongest lines in the hafnium spectrum in the region between 7300 and 3500 A.the intensity for both arc and spark spectra being also noted. The authors find in the region of the spectrum for which Ewer and Haschek’s zirconium measurements are sufficiently exhaustive nearly all the strong hafnium lines here given among Exner and Hascheks zirconium spark lines as weak lines of intensity 1 or 2. The lines 6386 4093 and 3505 are for the time being regarded as the most persistent hafnium lines in this part of the spectrum. Ultra-red Natural Frequencies of Salts containing Sulphur. V. J. SIHVONEN (2. Physik 1923 20 272-279).-Employing the method used by Schaefer and Schubert (A. 1916 ii 506) the authors have determined the natural frequencies of the anions of various salts containing sulphur by observing the reflection of radiation included between Ah 5p and 2Op from the polished face of compressed pastilles of the respective salts.Reflection maxima indicating corresponding anionic natural frequencies were observed a t the following respective wave-lengths (measured in p) sodium sulphite 10.6 19.5; barium sulphite 10.9 19.7; sodium ethyl sulphonate 8-7 9.7 135 19.2 ; potassium ethanesulphonate 8.7 9.7 13.6 19-2; sodium thiosulphate 9.1 10.2 15.1 18.5 19.7; potassium thiosulphate 9.1 10.2 15.2 18.5 19.5; barium thio- sulphate 9.3 10.4 15.0 18.2; sodium dithionate 8.2 10-1 17.3 19.4 ; potassium dithionate 8-2 10.2 17-6 19.3 ; barium dithionate 8.2 10.2 18.0; potassium trithionate 8-3 10.0 15.1 16.5 19.3; potassium tetrathionate 8.6 10.1 16.3 19.4.Salts other than the sulphites and sulphates included in these results show two reflection maxima in the region 8p to 11p. The shorter of these wave- lengths is connected with the presence of the sulphate-ion the longer with the presence of the sulphite-ion. The occurrence of a reflection maximum at 8-7p (the sulphate maximum) in the case of the ethanesulphonates investigated is interpreted as indicating the presence of a sulphuryl group. It is suggested that the sulphonic group is in a state of labile equilibrium possessing alternately the frequencies associated respectively with the SO and SO configurations. J S. G. T. The Electrical Absorption and Dispersion Spectra of Methyl and Ethyl Alcohol in the Region M 30 to 90 Cm. G. POTAPENKO (2 Physik 1923 20 21-35).-The author has determined the respective values of the dielectric constants and absorption coofficients of methyl and ethyl alcohols for electric oscillations of wave-lengths ranging from 29.4 cm. to 89.7 cm.In the cam of methyl alcohol the value of the absorption coefficient increases 4 - 2 J. S. G. T. A. A. E,ii. 80 ABSTRACTS OF CHEXICAL PAPERS. rapidly as the wave-length employed decreases below about 60 cm. For short wave-lengths the value of the absorption coefficient of ethyl alcohol similarly increases rapidly and attains in the case of the shortest wave-lengths employed a very high value 0.843 which approximates to that characterising a metallic reflector. Under similar conditions the value of the dielectric constant of ethyl alcohol decreases rapidly a result which is interpreted as affording evidence of anomalous dispersion.No evidence of the existence of maxima of absorption characterising certain definite wave-lengths such as was deduced by Romanov from his observ- ations (A 1913 ii 182) was obtained. Values of the respective refractive indices calculated from the results are in close agreement with those determined by Drude von Bayer Wildermuth and Romanor. J. S. G. T. Ultra-violet Absorption Spectra of Unsaturated Com- pounds. Spectra of Vapours of Acraldehyde Crotonaldehyde and Glyoxal. A. LUTHY (2. physikal. Chem. 1923,107,285-304). -The absorption spectra of solutions of p-methyl-As-butylene ally1 alcohol acrylic acid acraldehyde crotonaldehyde and glyoxal have been measured in hexane and of the three last-named sub- stances also in ethyl alcohol. The absorption spectra of the last three substances have also been obtained with the vapours of these substances.p-Methyl-As-butylene shows two bands a flat one with a maximum at 2340-2350 8. and a second much stronger band which has its maximum beyond 1935 A . . Ally1 alcohol has a weak flat band in the region of 2340 8. possibly the maximum lies more towards the ultra-violet; the maximum of the second stronger band lies outside the region measured namely beyond 1935 8. Acrylic acid has a flat band with maximum a t 2409- 2415 8. and a second strong band with maximum a t 2050 A. Acraldehyde in hexane solution has a flat broad band with a maximum a t 3350 B.; further towards the red two small bands a t 3663 A.and 3514 A. are measured; it has a very strong band in the ultra-violet beyond 2080 8. In alcohol and water solution the maximum of the flat band iA displaced toward the violet whilst in ether solution the maximum is in the same position as in hexane solution. Crotonaldehyde has a flat band with maximum at 3290 A. in hexane which is displaced to 3205 8. in ethyl alcohol solution that is slightly more than the displacement of the corre- sponding band of acraldehyde. A second very intense band of crotonaldehyde lies in the extreme ultra-violet Glyoxal has no bands in the visible and long wave-length ultra-violet but the author has measured seventeen bands in the ultra-violet a t 4613 4483 4330 4199 4095 4013 3935 3863 3768 3670 3542 3487 3359 3301 3187 3154 and 3119 A.A very strong band occurs in the extreme ultra-violet at about 1960 A. and between this band and the last small band there is a flat broad band the maximum of which has not been measured The bands obtained with acraldehyde vapour agree very well with the bands of the hexane solution the bands of the vapour being slightly displaced towardsQENERAL AND PHYSICAL CHEMISTRY. ii. 81 the red. The heads of the vapour bands have an analogous structure the difEerence between the frequencies of the most intense lines being 1260. Crotonaldehyde has an entirely dserent spectrum from that of acraldehyde in the vapour condition In this case the bands are weak and not sharp but five regions which are much stronger than the remainder can be identified.The mean values are recorded. With glyoxal below 30 mm. pressure a group of very sharp absorption bands occurs in the visible region. At higher pressures a new absorption zone is found in the visible region. The second absorption region has the sisisisisisisisisisisisisisisisae position as that of acraldehyde. Glyoxal in the same way 88 acraldehyde shows strong absorption groups which coincide with the small bands in hexane solution. J. PICCARD and E. THOMAS (Heh. China. Ach 1923 6 1040-1043).-A.n attempt is made to correlate the colours of salts with the colour-producing properties of their const'ituent ions. Ions may be divided into three groups (1) coloured ions; compounds containing them are always coloured ; (2) ions possessing latent colour ; molecules formed from these tend to be coloured; (3) ions which have no colour latent or otherwise.Ions of the first group give a deeper colour with those of the second than with those of the third. The following classification is made. Group (1) MnOf Mn0411 CrOdI Cr20,11 PtCI,II AuCI,I CrII and Crm+aq Cu+aq Au Pt Fe Mn ; Group (2) S" 011 11 BrI ClI OH1 (phenols and derivatives certain enols) PeUI Gun C+ Bim Sb" Asrn HgI Hgn Pbn Cw AgI Tin Cdn; Group (3) FI SO4= ClO,= CH,*CO,I ZnI CsI RbI KI NaI LiI Can A P HI. Cadmium iodide is described as colourless but as both its 'ions have latent colour it might be expected to be coloured. Examination shows that a layer of the concentrated solution 10 cm. thick is distinctly yellow. Mercuric bromide and mercurous chloride also form distinctly yellow soh- tions. E.H. R. The Luminescence of Titanium Oxide. E. L. NICHOLS (Physica.2 Rev. 1923 [ii] 22 4 2 0 4 2 4 ; cf. Nichols and Wilber A. 1922 ii 105).-Cathode rays but not ultra-violet light from an iron spark excite a very faint orange-red fluorescence in titanium oxide or after fusion of the surface layer a fine blue fluorescence with rapid fatigue. A thin layer of the oxide when heated gave a very feeble greyish-blue fluorescence up to 425" strong red from 425" to 677" and strong yellow from 677" to 1,000". When the oxide is heated directly with an oxygen-hydrogen flame two distinct phases are observed according as the oxygen or the hydrogen is in excess. Velocity of Photochemical Reactions under the Action of Light the Intensity of which is Periodic.P. LASAREV (Cmpt. rend. 1923,177,1436-1438).-Talbot's Law is shown to be readily deducible from a consideration of photochemical kinetics [cf. J . Russ. Phys. Chem. Xoc. (physical sect.) 1915 47 9581. J. 3'. S. Coloured Ions and the Colours of Salts. A. A. E. E. E. T.ii. 82 ABSTRACTS OF CHEMICAL PAPERS The Photolysis of Carbonic Acid E. BAUR and P. BUCHI (Hdv. Chim. Acta 1923 6 959-965).-Attempts were made to repetit the experiments of Baly Heilbron and Barker (T. 1921 119 1025-1035) who observed the reduction of carbon dioxide to formaldehyde through the action of light in presence of a sensi- tising dye such as Malachite-green. Formaldehyde was indeed found in the solutions after illumination but just as much or more when carbon dioxide was absent as when it was present.The production of formaldehyde was increased in presence of barium hydroxide which precluded the possibility of the presence of carbon dioxide in solution. It is concluded that the form- aldehyde is formed from the dye especially when this is present as- a colloidal solution of the leuco-base [ ? colour base] as it is in the barium hydroxide solutions Solutions of dyes (Rhodamine and Phosphine) in xylene containing lecithin or spermaceti were exposed to light ; also dyeings of Eosin Rhodamine Phosphine and Malachite-green on cotton and silk and resinates of the same dyes were illuminated under water both pure and saturated with carbon dioxide. In no case could formation of formaldehyde oxalic acid or formic acid be detected.Validity of the Photochemical Law of Equivalence in the Case of Emulsions of Silver Halides. J. EGGERT (Phgsikal. Z. 1923 24 488-490).-The experimental method described in a previous paper (ibid. 1921 22 673) has been considerably improved and the result definitely established that provided the time of exposure of the emulsion to the radiation is not excessive photochemical absorption of radiation by emulsions of silver chloride or silver bromide obeys the photochemical law of equiv- alence viz. one silver atom is affected per quantum of energy absorbed. For longer exposures the ratio of the number of silver atoms affected to the number deduced from quantum considerations diminishes as the time of exposure increases. The author considers that the latent image is constituted of metallic silver.J. S. G. T. The Photo-electric Effect of Caesium Vapour. J. KUNZ and E. H. WILLIAMS (Physical Rev. 1923 [ii] 22,456460).-Previous work (A. 1922 ii 809) has been repeated with greater precision by the use of a quartz tube with plane parallel end plates. The critical wave-length for photo-electric emission from caxium vapour was found to be 3180 8. (average) in agreement with the con- vergence wave-length 3184.28 k which is related to the ionisation potential Vi according to the equation Vi=hc/eXi. It follows that the separation of an electron from a cmium atom requires the same amount of work whether produced by an impinging electron or by absorption of light. A. A. E. Spectro-photo-electrical Effects in Argentite. P. H. GEIGER (Physical Rev.1923 [ii] 22 461469).-When argentite is con- nected in a closed circuit with n galvanometer and one surface of contact is illuminated an E.M.F. is produced which increases with the intensity up to about 0.013 volt for 600 candle-metres and over. E. H. R.CIENEBAL AXD PHYSICAL CHEMZ3TBY. ii. 83 Continued exposure results in marked fatigue. The effect which is not thermo-electric is shown also with proustite (Ag3AsS3) ppargyrite (Ag,SbS,) bournonite [3(Cu2,PbS)Sb,S,] molybdenite (MoS,) stephanite (Ag,SbS,) and acanthite (Ag,S) although to a less extent than with argentite (Ag,S); cuprite gave indefinite results whilst the effect was no€ observed in the case of galenite stibnite and polybasite. The electrical resistance of argentite was found to be only half as great for 20 as for 4 volts and is about 0-8 as great in the light as in the dark Slow recovery followed a condition of insensitiveness to light attained after some time.A. A. E. Rontgenographic Chemical Investigations. I. Rontgen Spectra and the Periodic System. H. STINTZING (2. physikul. Chem. 1923 107 154-162).-A theoretical paper in which it is shown that up to the presevt the study of Rontgen spectra has only led indirectly to the possibility of determining structure according to Bohr’s theory. The author puts forward a number of numerical relationships in the periodic system which may serve as an indication of the lines on which the direct investigation of the structure of atoms and atomic nuclei may be pursued. Itl is suggested that such an investigation could be carried out by means of y-rays.It is also pointed out that positive ray analysis may explain atomic structure and other chemical problems particularly the definite fixing of whole number atomic weights. Rontgenographic Chemical Investigations. 11. Technique of Rontgenology. H. STINTZINQ (2. physikal. Chem. 1923,107 163-180 ; cf. preceding abstract).-A description of the most suitable forms of apparatus and the most favourable conditions for the investigation of Rontgen spectra. A new arrangement of the electrodes is described for Rontgen analytical purposes. It is shown that if the discharge is divided between the discharge tube and a spark gap in parallel then the maintenance of a definite tube voltage as measured by the spark potential has little influence on the production and the intensity of a suitable Rijntgen radiation.Constitution of the X-Ray Spectra belonging to the LSeries of the Elements. H. HIRATA (Proc. Roy. Soc. 1924 [ A ] 105 40- 60).-Experimental data due to Siegbahn (A. 1919 ii 488) and to Siegbahn and Friman (A. 1916 ii 167 277 361) are used to calculate the respective values of the ratio of the wave number V t o Rydberg’s constant N in the L-series of the X-ray spectra of forty-eight elements included between zinc (atomic number 30) and uranium (atomic number 92). Values so determined are shown to confirm approximately certain deductions made by Sommerfeld from the extension of the theory of “ wasserstoffahn- lich ” spectra to the K and L series of X-ray spectra (Ann. Physik 1916 [ivl 51 133).The slight discrepancy is attributable to the number of electrons in the K - and L-orbits being so large that their effects on the motion of an electron in the same or an inner orbit J. F. S. J. 3’. S.ii. 84 ABSTRACTS OF UEEItfICAL PAPEBS. are appreciable. Moreover the radius of the innermost ( K ) orbit cannot be' considered negligible compared with the radii of the outer orbits. Sommerfeld's theory is extended by taking account of these correcting factors. A mathematical discussion involving quantum considerations of the initial and h a 1 stationary orbits corresponding with X-ray radiation the planes of the stationary orbits and the intensities of X-ray radiation leads to the conclusion that an electron moves from a certain outer orbit to the neighbouring inner one without changing its spatial quantum number but if there are intermediate orbits between the initial and h a 1 orbits the spatial quantum number changes every time the electron passes through one of the intermediate orbits.Reflection of the Characteristic Rontgen Radiation associated with the Chemical Elements composing a Crystal by the Elements. B. WALTER (Z. Phpik 1923,20,257-271).- Employing an ionisation chamber for the measurement of the respective intensities of radiation of various wave-lengths reflected by crystals of bromides iodides or containing caesium on which Rontgen radiation excited by means of a tungsten anticathode was incident Clark and Duane found that the intensity of reflected radiation characteristic of bromine iodine etc.was of the same order of magnitude as and even greater than the intensity of reflection of the incident radiation characteristic of the material of the anti- cathode (A. 1923 ii 468). Employing a photographic method of comparing the intensities of the reflected radiation of difTerent wave-lengths and using Rontgen radiation excited by means of a molybdenum anticathode the author has been unable to confirm this result. The intensity of the reflected radiation characteristic of the crystal element was in all cases immeasurably small. No satisfactory explanation of the apparent discrepancy shown by the results obtained by the respective methods is at present avail- able. It is pointed out that the results obtained by Clark and Duane indicate a very rapid decrease of intensity of the reflected radiation on the long wave-length side of the head of the absorption band characterking the crystal elements in question.This result is contrary to what would be anticipated from a knowledge of the distribution of intensity throughout the continuous spectrum afforded by a Rontgen tube. Particles of Long Range Emitted by the Active Deposits of Radium Thorium and Actinium. L. F. BATES and J. S . ROGERS (Proc. Roy. Soc. 1924 [A] 105,97-116).-The emission of long-range particles by the active deposits of radium thorium and actinium has been investigated by means of the scintillation method (A. 1921 ii 671). In the case of radium active deposit in addition to particles of range 9.3 cm. previously found by Rutherford particles of respective ranges 11-2 and 13.3 cm.were found. For every lo7 a-rays emitted by radium-(? the respective numbers of these particles present are 380 125 and 65. Thorium active deposit emits particles of ranges 16.0 and 18-4 cm. in addition to particles of ranges 8.6 and 11-5 cm. (A. 1916 ii 282). The J. S. G. T. J. S. G. T.relative numbers of these particles emitted were found to be 47 55 106 and 220 respectively. Actinium active deposit was found to emit particles of range 6.49 cm. previously recorded by Marsden and Perkins (A. 1914 ii 410) to the extent of 0.322% of the total number of a-rays emitted by actinium-C. Indications of the presence of particles of range greater than 6.5 cm. were obtained. Evidence as to whether these long-range particles were a-rays or H-particles is inconclusive.Schemes hitherto proposed for the modes of transformation of the C products of the radio- active substances must be considered incomplete. More especially for the atoms of radium4 and thorium-C explanation must be found for the presence of four particles of different ranges. J. S. G. T. The Absolute Energies of the Groups in Magnetic p-Ray Spectra. C. D. ELLIS and H. W. B. SKINNER (Proc. Roy. Xoc. 1924 [ A ] 105 6049).-Absolute determinations correct to 1 part in 500 have been made of the respective energies of the electrons associated with six of the principal homogeneous groups of @-rays emitted from radium-B. The values expressed in volt deduced from determinations of the curvature of the respective rays in a magnetic field are 0.3725 0.4983 1.529 2.067 2.638 and 3.379.J. S. G. T. Discoloration and Luminescence due to Becquerel Rays. I<. PRZIBRAM (2. Physik 1923 20 196-208).-Iiiterature dealing with the discoloration and luminescence accompanying the incidence of Becquerel rays on substances e.g. rock salt is reviewed more especially with reference to conclusions concerning the ,phenomena reached by Meyer and the author in a previous paper (A. 1922 ii 339). The mechanism of the phenomena is discussed from the point of view of the quantum theory. Ionisation Hydrogen and Oxygen. H. D. SMYTH (Proc. Roy. Soc. 1933 [ A ] 105 116-128).-1n continuation of previcus work (A. 1923 ii SOZ) the author has investigated the ionisation of hydrogen and oxygen by electron impact. The results obtained show that ionisation in hydrogen at about 16.5 volts is not accom- panied by dissociation as has been generally supposed.Atomic hydrogen-ions are first produced at a potential 4*25&0-06 volts above that a t which molecular ions first appear and the proportion of atomic ions remains very small up to 720 volts. I n hydrogen at sufficiently high pressure atomic ions predominate over mole- cular ions and appear a t approximately the same voltage. They originate owing to dissociation due to collisions of ionised molecules with other molecules. In oxygen molecular ions are produced a t about 15.5 volts whilst doubly and singly charged atomic ions appear a t about 7 and 7*5&0.2 volts higher. Results in the cme of oxygen can be interpreted by thermochemical equations similar to those previously employed in the case of nitrogen. W.&I. LATIMER ( J . Arner. Chew. $m.. 1923 45 2803-2808).-The heats of ionisation of the J. S. G. T. J. S. G. T. Ionisation of Salt Vapours. 4*ii. 86 ABSTRACTS OX CHEMICAL PBPEBS. hydrogen and alkali metal chlorides bromides and iodides hare been tabulated for the gaseous condition. The ionisation constant of sodium chloride a t its boiling point has been found to be 4 x From a consideration of the fields of force about the ions and the molecular diameters it seems that the heats of ionisation cannot be calculated by any method which treafs the ions as rigid structures of electrons about a positive nucleus. Radioactivity of the Alkalis. G. HOFFXANN (Physikd. Z . 1923 24 475476).-The radioactive character of salts of sodium potassium rubidium and caesium and the relative magnitudes of the small activities associated with these elements have been determined in two djfferent ways.The results are in satisfactory agreement with values of the activities given by Hahn and Rothenbach (A 1919 ii 312). Thermionic and Photo-electrical Properties of the Electro- positive Metals. A. F. A. YOUNG (Proc. Roy. Soc. 1923 [ A ] 104 61 1-639).-The fhermionic and photo-electrical properties of sodium and potassium have been investigated. It is found that sodium does not give a measurable thermionic emission that is greater than amperes for temperatures up to 390'. Potassium is found to give currents measurable in some cases by an electroscope a t temperatures down to the ordinary.The currents are held to be thermionic in origin and not due to chemical action. The values of b for potassium in the emission formula i=AlcT1/2e-b/T are found to be lower than those obtained with other metals and the values of A are also low. The change of the potassium from solid to liquid makes no dserence in the magnitude of the thermionic current or in the value of b. Deter- minations of the photo-electric work function t& from photo-electric measurements made a t the same time as the thermionic measure- ments and a t nearly the same temperature show that & is always greater than the thermionic work function Q calculated from the value of b. J. F. S. Dielectric Constants of Organic Liquids at the Boiling Point. F. V. GRIMM and W.A. PATRICK ( J . Amer. Chem. Soc. 1923 45 2794-2802) .-The dielectric constant of thirty-five organic liquids has been determined a t the boiling point. The following values are recorded benzene 2-17 ; toluene 2.17 ; m-xylene 2.15 ; chloroform 4.23 ; carbon tetrachloride 2.10 ; carbon disulphide 2-58 ; ethyl ether 4-11 ; these values are accurate to 0.01 unit ; p-cymene 2.27 ; ethyl bromide 8.81 ; ethylene bromide 4-09 ; isoamyl bromide 4.70 ; n-butyl iodide 4.62 ; methyl iodide 6.48 ; ethyl acetate 5-30 ; chlorobenzene 4.20 ; aniline 4-54 ; o-toluidine 4.00 ; butaldehyde 10.78 ; paracetaldehyde 6-29 ; anisaldehyde 10.38 ; acetone 17.68 ; methyl ethyl ketone 14.46 ; acetophenone 8.64 ; ethyl alcohol 17.30 ; n-propyl alcohol 11.83 ; n-butyl alcohol 8-19 ; isoamyl alcohol 5-82 ; o-nitrotoluene 11-82 ; pyridine 9-38 ; quinoline 5.05 ; these values are accurate to 0.02 unit ; acetonitrile 26.2 ; phenylacetonitrile 8-5 ; ethylene chlorohydrin 13.2 ; and J.F. 8. J. 8. G. T.GEl,NERAL AND PHYSICAL CHEMISTRY. ii. 87 nitromethane 27.75; these values are accurate to 0.1 unit. A method has been suggested for evaluating the volume occupied by the molecules in the liquid state. The above idea has been applied to Bakker’s equation for the latent heat of vaporisation. [Relation between] Cathode Drop of Potential Ionisation Potential and Atomic Weight. A. GUNTHER-SCHULZE (2. Physik 1923 20 153-158).-The author shows that the average normal cathode drop of potential V in the case of a discharge through any elementary gas of atomic weight.M is approximately related to the ionisation potential Vi of the gas by the equation Vn=(O*245211+4)V~or in which cc is a constant having the value unity for all monatomic gases whilst the appropriate value of a in the case of diatomic gases is to determined from consideration of the loss of energy due to electronic impacts and the efficiency of such impacts. It follows that the efficiency of the process of ionisation of the gas due to kations is to a first approximation inversely proportional to the square of the atomic weight of the gas. J. S. G. T. Influence of Neutral Salts on the Potential of the Hydrogen Electrode. J. PRZEBOROWSKI [with M. FLEISSNER and A. SABRODINA] (2. physikul. Chem. 1923 107,270-278).-The action of 0-1N and 0.01N solutions of potassium nitrate chloride and bromide sodium nitrate chloride and bromide and lithium chloride and bromide in hydrobromic acid solution on the potential of the hydrogen electrode has been investigated.It is found that the potential of the hydrogen electrode in 0-1N and 0.01N hydrobromic acid solutions increases in all cases. A shght increase of the potential is found for solutions containing the potassium salts at concentrations less than 1N. The influence of the neutral salts increases with increase in their concentrations. The influence of these salts increases with their solubility hygroscopic properties and power of forming crystal hydrates. The influence of the salts is in the following order KNO KCl KBr NaC1 NaBr LiCl LiBr lithium bromide having the greatest influence.Bromides have a greater influence than chlorides the kation being the same in both cases. With a constant concentration of neutral salt and a variable concentration of hydrobromic acid the fall in the E.M.F. is almost independent of the hydrobromic acid concentration for salts which form crystal hydrates. The corresponding difference of E.M.F. for salts which do not form crystal hydrates varies very much with the difEerent salts and still more the greater the con- centration of hydrobromjc acid. The action of neutral salts can be explained by the assumption that only non-hydrated hydrogen- ions have an influence on the potential; the ions formed from neutral salts dehydrate the hydrogen-ions and so increase the concentration of non-hydrated ions and in consequence the potential.The electrolytic solution tension of the hydrogen changes with the concentration of the neutral salt. In addition to the ions of neutral salts the undissociated molecules also exercise a dehydrating action on the hydrogen-ions. J. F. S. J. F. S. 4*-2ii. 88 ABSWCTS OX CHElKICAL PAPERS. Hydrogen Electrode in Alkalm ' e Solutions. A. H. W. ATEX (Trans. Amr. Electrochem. Soc. 1923 43 89-98).-The potential of the hydrogen electrode has been determined in solutions of sodium hydroxide and in solutions of sodium hydroxide containing sodium chloride. It is shown that when a hydrogen electrode saturated with hydrogen is in equilibrium with 0.1N hydrochloric acid it is in the same state of equilibrium with 1.ON hydrochloric acid and vice versa.This however is not the case when a solution of an alkali hydroxide is used. When a hydrogen electrode hi equilibrium with 1*ON sodium hydroxide is placed in 0-1N sodium hydroxide a very considerable time elapses before it reaches a new equilibrium. The same phenomenon is observed in a more marked degree when the electrode is changed from 0*1N sodium hydroxide to 0.1N hydrochloric acid or the reverse. The explan- ation given is that the electrode must absorb sodium-ions or give them out as the case may be in order to reach an equilibrium with the final solution. J. F. S. M. KNOBEL P. CAPLAK and M. EISEMAN (I'rans. Amer. Electrochem. Xoc. 1923 43 55-74).-The value of the hydrogen overvoltage has been determined with cathodes composed of twenty-two different materials that of chlorine bromine and iodine at three different anodes and that of oxygen a t nine difTerent anodes.The deter- minations have been carried out at various current densities from 1-0 milliampere to 1.5 amperes per sq. cm. All measurements were made a t 25.0"&0.2". An investigation of the method of measuring overvoltage has led to the conclusion that the use of a small glass tip less than 1 mm. diam. pressed against the active electrode surface while the current is passing will give correct results . J. F. S. S. J. BIRCHER and W. D. HARKINS ( J . Amer. Chem. Xoc. 1923 46 2890-2898).-The effect of pressure on the potential of the hydrogen electrode and on the potential and overvoltage of a mercury cathode has been ascer- tained a t 0" for pressures from 757 mm.to 9 rnm. The results show that the effect of pressure on the potential of .a cathode a t which hydrogen is being liberated is in general very small at pres- sures between 760 mm. and 11 mm. The slight changes which occur are due to bubble expansion and consequent shielding of the cathode and to increased stirring caused by more rapid bubble liberation at low pressure. Overvoltage defined with reference to a reversible hydrogen electrode increases as the pressure is decreased. This increase is paralleled entirely by a decrease in potent.ia1 of the hydrogen electrode. Between 760 mm. and 11 mm. the range of variation of the overvoltage is about 50 millivolts. The present data are of importance in connexion with the theory of overvoltage due to MacInnes and Adler (A.1919 ii 131). Here the over- voltage E is given by E=(3RT/2p) y where p is the pressure in the bubbles r their radius and y the surface tension of the liquid. Since the bubble radius. is practically constant with variation of Effect of Current Density on Overvoltage. Effect of Pressure on Overvoltage.GENERAL AXD PHYSICAL CEEBIISTRY. ii. 89 pressure the overvoltage should vary inversely as the pressure and thus increase very rapidly as low pressures are approached. The present work indicates that the overvoltage increases with decrease of pressure only to the extent that the hydrogen electrode potential decreases that is a8 the logarithm of the pressure. On the other hand the overvoltage of a number of inactive metals is found to have the same temperature coefficient and to decrease 2 millivolts per degree in 0.1N sulphuric acid.This seems to point to the surface tension of the liquid as an important factor in overvoltage. This fact indicates that MacInnes and Adler's hypothesis is a partial rather than a complete hypothesis of over- voltage. In the experiments care was taken to avoid stirring the liquid round the cathode so the smallness of the increase of over- voltage with decrease of pressure cannot be due to the effect of stirring. The overvoltage a t a mercury meniscus in dilute sulphuric acid is found to increase as the large bubble a t the top of the meniscus becomes larger and to decrease suddenly when the bubble escapes. J. F. S. Free Energy of Dilution and the Activity of the Ions of Hydrogen Iodide in Aqueous Solution.J. N. PEARCE and A. R. FORTSCH ( J . Arner. Chem. Soc. 1923,45,2852-2857).-The E.9i.F. of cells of the type H,IH(c),AgIIAg has been measured at 25" 30" and 35" for concentrations of hydriodic acid between 0.24661 and 0-005M. The decrease of free energy and heat content attend- ing the cell reaction has been calculated for 25". The decrease of free energy accompanying the transfer of one gram-molecule of hydrogen iodide from the various concentrations to a concentration exactly 0.00561 has been calculated. From these values the geo- metric mean activity coefficients of the ions of hydrogen iodide have been calculated and it is found that for concentrations up to 0.00561 these coefficienfs are practically equal to the corresponding coefficients for hydrogen chloride at the same concentrations.Ob- viously therefore the activities of the iodide- and chloride-ions are equal when in equivalent concentrations of their salts. Activities of Zinc Cadmium Tin Lead and Bismuth in their Binary Liquid Mixtures. N. W. TAYLOR (J. Arner. Chem. Xoc. 1923 45 2865-2890).-Accurate determinatious of the activities of liquid zinc cadmium tin lead and bismuth in the binary alloy systems zinc-tin zinc-cadmium cadmium-bismuth cadmium-lead and cadmium-tin have been made at temperatures from 400" to 600" from E.M.F. measurements of cells of the type MIelectrolyteIAlloy. The electrolyte wm the entectic mixture of lithium and potassium chloride with a small amount of zinc or cadmium chloride and potassium hydroxide.With the single exception of the system cadmiurn-bismuth which gave. a very complicated type of activity curve all the alloy systems investigated showed escaping tendencies or activities greater than that required by Raoult's law. I n regard to divergencies from this ideal solution law and also in regard to the heats of mixing the results cjf the J. F. S.ii. 90 ABSTRACTS OF CHEMICAL PAPERS. present work furnish strong evidence for the validity of the internal- pressure theory as applied to liquid metal systems. Velocity of Hydrogen-ions in Gels under the Influence of a Current. N. ISGARISCHEV and A. POB~RANZEV (2. Elektrochem. 1923 29 581-586) .-The migration velocity of the hydrogen-ion in gelatin gels has been measured in the presence of difEerent electro- lytes including the chlorides of ammonium potassium sodium barium calcium magnesium lithium manganese and cadmium and the sulphates of most of the above-named metals and also of zinc and rubidium under various experimental condit.ions.A chamcteristic slowness of the hydrogen-ion is observed on passing from one potential to another which is explained by the fact that the gel has a semi-rigid structure in which the ions are included and have thereby lost to some extent their free condition and power of progressive movement. The influence of the various kations in the gel on the movement of the hydrogen-ions is in direct relationship with the atomic volumes of the kations. Mzgnetic Properties of some Paramagnetic Double Sulphates at Low Temperatures.L. C. JACKSON and H. KAMERLINGH ONNES (Proc. Roy. Soc. 1923 [A] 104 671-676; cf. A 1923 ii 609).-1n7 continuation of previous work on the magnetic properties of compounds of iron nickel and cobalt the magnetic susceptibility of cobalt potassium sulphate cobalt rubidium sulphate and manganese ammonium sulphate has been determined a t temperatures from 14.54" Abs. to 291.0" Abs. The results show together with those given in the previous paper (Zoc. cit.) that all the cobalt compounds obey Weiss's law x'm(T+A)=C down to about 70" Abs. where x'm is the molecular susceptibility but deviate from the law below 70" Abs. in such a manner that the susceptibility at the lowest temperature is greater than the value calculated by Weiss's law. The magneton numbers of the cobalt compounds are not equal but increase as the molecular weight of the com- pound increases Manganese ammonium sulphate follows the simple Curie law xT=C over the whole range of temperature investigated.Comparing this result with those obtained for hydrated and anhydrous manganese sulphate it is found that tfhe manganese compounds shorn a simple behaviour as regards the dependence of the susceptibility on temperature the greater the magnetic dilution of the substance. The complications found with cobalt nickel and iron compounds do not occur The Crystallisation of certain Salts in the Magnetic Field. G. ROASIO (Rev. gdol. 1923 4 297).-Experiments with salts of iron nickel and cobalt show that the magnetic field exercises an influeiice on (1) the orientation of the crystals the angles between axes and lines of force varying with the substance but being con- stant for a given compound; and (2) the growth of the crystals which become elongated in a definite direction the angles remaining constant but again varying with the substance.J. F. S. J. F. S. J. F. S. CHEMICAL ABSTRACTS.GENERAL AND PHYSICAL CHEMISTBY. ii. 91 Specific Heat and Heat of Mixing in the Neighbourhood of the State of Critical Miscibility. N. PERRAEIS (Compt. rend. 1924 178 83-86; cf. A. 1923 ii 836).-The specific heats and heats of mixing of the following pairs have been measured (1) di- phenyl ether and .ethyl alcohol ; (2) benzene and ethyl alcohol ; (3) benzene and n-butyl alcohol ; and (4) o-cresol and ethyl alcohol. It is shown that the found specific heat of a mixture is greater than that calculated from the simple law of mixtures this fact necessit- ating a correction in the calculation of the heats of mixing.Curves representing the above constants are approximately rectilinear for a longer or shorter portion of their course according as the mixture under examination is near to or remote from a state of non- miscibility. Thus the length of the rectilinear portions of the curves decreases in the order (l) (Z) (3) (4) this being also the order of increasing miscibility. Thermodynamics of the Ionisation of Monatomic Gases R. BECKER (Physikal. Z. 1923 24 485486).-Cu and C the respective concentrations of the atoms and electrons in a thermally-ionised monatomic gas at temperature To are related by the equation C$/Cu=e-Q~I~'T(KT)3~ 2 .( 2 ~ p ) ~ / 2 / p ~ h ~ in which Q0 denotes the heat of ionisation p the mass of the electron p is a stmatbistical factor and the other symbols have the customary significance. It follows on the basis of the quantum theory that if ccU(v) denote the power of absorption of the gas for radiation of frequency Y = ( & + E ) / ~ where e is the kinetic energy associated with an electron in a a-orbit then Q ( E ) / ~ C ~ ( Y ) = ( E + & ~ ~ ~ ~ ~ / C ~ C ~ qrT( E ) denoting the statistically calculated electron-absorbmg cross sectional area of the ion available for trapping an electron E in a a-orbit!. This equation applied to absorption measurements relating to the principal series of absorption bands should afford quantit,ati.c-e information relating to the trapping of free electrons by an ion.E. E. T. J. S. G. T. Calorimetric Determinations. V. The History of the Thermochemical Standard Substances. P. E. VERKADE (Chern. Weekblad 1024 21 13-20).-The considerations leading up t o the adoption of benzoic acid as a standard by the Inter- national Union at its third conference a t Lyons in 1922 are described in some detail and the desirability of deciding on some other The Heat of Formation of Bleaching Powder and the Thermochemical Explanation of its Decomposition and Formation. SHUICHIRO OCHI ( J . Chem. Ind. Jupan 1923 26 961-978; cf. ibid. 1923 26 1 ; A. 1923 ii 564 573).-Assuming that the formula of bleaching powder is OCl*CaCl,I3,0 the author has made a determination of the heat of formation and then tried to solve thermochemically the problem of the decomposition and the formation of the compound by applying the value obtained.As the reaction OC1*CaC1+H,O2=CaCl2+ 02+H20 proceeds material as a second standard is pointed out. s. I. L.quantitatively and is not affected by the impurities contained in bleaching powder the heat of formation was measured with a simple calorimeter using hydrogen peroxide OCI*CaCl=CaCI,+ O+lS,W cal. and [Ca,O,Cl&=177,260 cal. The method em- ploying the bomb calorimeter and that of decomposing with hydro- chloric acid did not give accurate results. The averages were 176.1 Cal. from the former and 179-7 Cal. from the latter. The heat of decomposition of OCl-CaCl was calculated as follows OCl*CaCl=CaC1,+0+ 13,040 cal. OCl*CaC1,H,O=CaC~,H20+O+ 14,440 cal.=CaCl,+H,O(liquid)+O+10,740 cal. =CaC&+H,O (vapour)+O+l50 cal. OCl*CaCl,aq. =CaCl,,aq. +0+20,620 cal. OCl-CaCl=CaO +C&-25,360 cal. OCl*CaCl,H,O =Ca( OH) +C12 - 12,120 cal. and OCl*CaCl,aq. =Ca(OH),+aq.+C4-20,650 cal. All oxygen-forming decompositions are exothermic reactions whiIst the chlorine-forming ones are endothermic. At the ordinary temperature chlorine may not therefore be evolved as indicated above but is produced by the decomposition of C1,0 which may be easily formed from moist bleaching powder. The chlorate which occurs without exception is considered to be the product of the oxidation of the other molecule of OCl*CaCl or Ca(OCl) caused by the eliminated oxygen in the presence of water. The partial pressure of decomposing oxygen from OCl*CaCl is calculated to be 2-86 x 1013 atm.by Nernst's heat theorem. The heat of reaction between calcium hydroxide and chlorine is represented its follows Ca( OH),+Cl,= OCl*CaCZ,H,O + 12,120 cal. and Ca( OH),+ aq. + C12=OCkCaCl,aq. +20,650 cal. The partial pressure of chlorine reacting on calcium hydroxide is calculated by Nernst's heat theorem 0.004 atm. a t O" 0.016 a t 15" 0.49 a t 30° 0.138 a t 45" 0.355 a t 60° 1.000 a t 78" 3.1 at loo" and 26-7 a t 150". The author has explained the preparation of bleaching powder from the thermochemical point of view. K. K. B. IUIN (2. physikizl. Chem. 1923 107 145-153).-A theoretical paper in which an equation in the form of that of a unimolecular reaction is derived for the kinetics of sorption; this equation is given the form dc fdt=kcM .e-kt where c is the sorption capacity and in general is a function of the concentration p or the temperature T and E is a constant which is a function of the temperature. An equation for the sorption isothermal in the form c 1 -e-yp) is deduced in which o is a constant. It is shown that under certaJn conditions the sorption capacity as placed in the above equation is not always constant and the correction in the kinetic equation introduced on this account increases the agreement between theory and experi- ment. In consequence the number of sorption processes expressed by the equation c=c (l-e-*t) is increased. The dependence of the sorption capacity on the temperature is given by the equation c ==coe-*dT where 6 is a constant.This equation agrees well with the experimental data. The influence of temperature on the initial sorption velocity wi has been determined and it is shown that wi can increase with the temperature whereas cK always decreases General Theory of the Phenomenon of Sorption.GENERAL AXD PHYSICAL C!HEMISTRY. ii. 93 with increase in temperature. This explains the crossing of the isotherms of the sorption kinetics for different temperatures. J. F. S. Investigations on the Adsorption of Iodine by Various Substances. A. LOTTERMOSER (Rolloid Z. 1923 33 271-274). -The adsorption of iodine from solutions in potassium iodide and carbon tetrachloride by starch basic lanthanum acetate lecithin-albumin calcium hydroxide cholic acid and calcium oxide has been investigated It is shown in the case of basic lanthanum acetate that when the basic acetate is prepared a t 18" iodine is adsorbed to form a brown precipitate but when the precipitate is kept for twelve hours or warmed at 40-50° it then adsorbs iodine to form a blue precipitate.Heating a t 70" gives a precipitate which adsorbs iodine with the formation of a dirty green precipitate whilst if the precipitate is boiled it then adsorbs iodine to form a yellow adsorption compound. The adsorption curves in the various cases are entirely different. The longer the basic acetate is kept the less iodine is adsorbed. This is due to the fact that the structure of the basic acetate changes on keeping and in four weeks it becomes crystalline and then ceases to adsorb iodine.Calcium hydroxide adsorbs iodine from a solution in carbon tetrachlolTide with the formation of a brown colour ; in this case it is held that the iodine is chemically combined with the calcium hydroxide. Calcium oxide adsorbs iodine from solutions in carbon tetrachloride according to an adsorption isotherm and calcium carbonate does not adsorb iodine a t all. Cholic acid adsorbs iodine from solutions in potassium iodide when it crystallises from such solutions but crystalline cholic acid when placed in a solution of iodine in potassium iodide does not adsorb iodine. J. F. S. Adsorption Compounds. IV. R. HALLER (Kolloid Z. 1923 33 306-309; cf. A. 1919 ii 198; 1921 ii 21).-!Che dyeing of cotton and wool by indigotin Alizarin VI Indanthren-blue-RS and Para-red sol both alone and in the presence of sodium hydrogen sulphate haa been investigated and both the dye solution and the dyed fibres have been examined microscopically.In the case of indigotin with cotton a greyish-blue colour is produced which is not changed much by sodium hydrogen sulphate and practically the whole of the dye can be removed by washing with water. The indigotin particles are very loosely held by the fibres. With wool the case is Merent ; here the fibres become dyed more rapidly and more intensely and very little of the colour is lost on prolonged mashing. The presence of sodium hydrogen sulphate gives a deeper colour and the dye is homogeneously disseminated through the fibre. With Alizarin VI suspensions cotton on boiling with the solution forms a weak adsorption compound which breaks down on washing repeatedly with water.In the presence of sodium hydrogen sul- phate there is a strong adsorption and the cotton is dyed brownish- yellow ; the adsorption compound thus formed is more stable towards water a d on repeated washing the cotton remains pale yellow. With wool the adsorption is much stronger than with cotton theii. 94 ABSTRACTS OP CHEMICA4L PAPERS. fibres becoming violet coloured but in the presence of sodium hydrogen sulphate chrome-yellow. Both withstand wmhing with water and microscopic examination shows that the fibres are homo- geneoilsly coloured. The results obtained with Indanthren-blue- RX are similar to those obtained with alizarin. The experiments show that wool is a more powerful adsorbent than cotton.Wool forms stable derivatives with chemically inactive substances which can only be termed adsorption compounds. The results raise a doubt as to the correctness of the generally held view that the dyeing of wool is a chemical process. H. G. DEMING and B. C. HENDRICKS ( J . Amer. Chem. SOC. 1923 45 2857-2864).- An apparatus is described which enables the diffusion of gases through metals to be measured a t temperatures up to 1,000" under perfectly definite conditions. The specific rate of diffusion of hydrogen t'hrough several common metals has been measured and the following values expressed in mg. per hour per sq. cm. area per mm. thickness have been obtained aluminium no diffusion detectable a t 555" ; zinc 0.0012 at 375" ; lead 0.001 a t 265" ; copper 0.011 a t 500" ; copper 0-028 a t 770" ; nickel 0.012 at 500" and 0-100 at 750".In addition the specific rate of diffusion of hydrogen has been measured for copper and nickel over a considerable range of temperature. The following results are recorded copper 500" 0.027; 550" 0-032; 645" 0.042; 687" 0-056; 770° 0.072. These values are C.C. of hydrogen per hour diffusing through 1 cm.2 of copper 0-391 mm. thick. Nickel 403" 0.004; 450" 0.007; 527" 0.020; 580" 0.029; 625" 0.056; 695" 0.114; 745" 0.153. In this case the nickel plate was 0-653 rnm thick. Solubility. IX. Metallic Solutions. J. H. HILDEBRAND T. H. HOGNESS and N. W. TAYLOR ( J . Amer. Chem. SOC. 1923 45,2828-2836 ; cf. A. 1923 ii 315).-A theoretical paper in which the authors discuss the various methods of estimating the relative internal pressures of metals. Tables are given of the values of this quantity deduced from expansion and compressibility surface tension and heat of vaporisation.Data are given showing the correlation between these tables and the behaviour of binary metallic solutions. J. F. S. Individual Thermodynamic Properties of Ions. J. N. BRONSTED ( J . Amer. Chem. AYOC. 1923 45 2898-2910).-The principle of the specific interaction of ions is presented in the form of a simple equation and a diagram. The individualities of the fhermodynamic properties of salts vary linearly with their concen- tration when the total concentration is kept constant. On the basis of this law several of the results obtained by means of the principle of the specific interaction may be derived.Thermodynamic and experimental evidence to prove the invalidity of the principle of the independent activity coefficients has been adduced. Solubility measurements embracing a series of cobaltic ammine salts in solutions of sodium sulphate and chloride have been carried out a t Z O O and J. F. 5. Diffusion of Hydrogen through Metals. J. F. S.GENERAL AND PHYSICAL CHEMISTRY. ii. 95 the results are found to be in full agreement with the principle of specific int,eraction. J. 3'. S. Dependence of the Mobility of Univalent Ions on the Temperature. P. WALDEN and HERM. ULICH (2. physikal. Chem. 1923,107,219-234).-A theoretical paper in which i t is shown that the mobility of some ions changes proportionally with the fluidity of the solution and also follows Stokes's law.The deviations of the other ions from Stokes's law are represented by simple empirical relationships. The regularities found are in opposition to the deductions from Born's hydration hypothesis. It is shown to be probable that the meaning of the regularities found may be obtained from Cunningham's formula (Proc. Roy. Xoc. 1910 [ A ] 83 357). Many facts are shown to be best explained by the existence of a firmly held water sheath on the ions. The Crystal Structure of Metals Mixed Crystals and Metallic Compounds. K. BECKER (2. Metullk. 1923 15 303- 305).-A review of the present knowledge of. the crystal structure of metals is given together with a table showing the systems in which thirty-nine metals crystallise as shown by the Debye-Schemer method of X-ray analysis.Fifteen of these metals crystallise in a face-centred cubic lattice nine in a space-centred cubic lattice nine in the hexagonal two in the tetragonal two in the rhombohedral and two in a diamond lattice In addition five elements have a second crystalline modi6cation. Values for the density of the elements calculated from the dimensions and arrangement of the lattice are always about 2% higher than those actually found by experiment. Metallic mixed crystals with a static arrangement of the atoms give X-ray photographs in which the lattice constants if both metals crystallise in the same system are an approximate linear function of those of the constituents. If the lattices are clissimilar t.he atoms of the metal present in the smaller amount arrange themselves in the lattice of the other metal.No relation appears to exist between the symmetry of crystals of intermetallic compounds and that of their components ; thus Cu,Zn and A13Mg4 have a regular face-centred lattice whereas zinc and magnesium crystallise in the hexagonal system; &,A1 and CUM are tetragonal or hexagonal and NiAl has a simple cubic lattice whilst copper nickel and aluminium all have a cubic face-centred lattice (cf. A. 1923 ii 519). H. =GI (Kolloid Z. 1923 33 284-286).-h an earlier paper a description of the production of rhythmic crystals of the racemic form of the menthyl ester of benzyl- acetoacetic acid was given (A. 1920 i 748) ; further details for the production of rh-ythmic crystal figures are now given.It is shown that t'he ester employed must melt a t 48-52' and that it is best. purified by distillation in steam since crystallisation from acetic acid and alcohol produces a separation of the isomerides. Further a freshly prepared benzene solution should be used for when kept a iuutarotation occurs as shown by the following values of the rotation J. F. S. A. R. P. Rhythmic Crystallisation.ii. 96 ABSTRACTS OF CHEMICAL PAPEBS. measured at intervals of eight days - gelo -gago -l0-lo. The solution having the last-named value gives no ring-formed crystal structure. J. F. S. Explanation of the Colloidal Appearance of the Proteins. J. LOEB (Rev. Gen. Cdlokh 1923,1 3-17).-Portion of a lecture delivered a t the Institute of Medicine in Chicago November 24th 1922 in honour of Pasteur. The lecture deals with the pro- perties of the proteins and shows that the chemistry of these sub- stances is not Merent from that of crystalloid substances and that they can combine stoicheiometrically with acids and bases to form protein salts which in solution dissociate electrolytically.The very large ions and molecules of the proteins are unable to diffuse freely across gels and membranes which are easily permeable to the smaller crystalloid ions. These facts lead to an unequal partition of diffus- able crystalloid ions between a solution of a protein and an external aqueous solution or between a protein gel and an aqueous solution. In this partition the total concentration of crysta.lloid ions is alu7ays greater in the protein solution or gel than in the surrounding aqueous solution.This is the cause of the colloidal appearance of solutions and gels of proteins. Measurement of the membrane potential shows that the excess concentration of crystalloid ions in the solution of protein over that outside this solution can be determined by means of Donnan's theory of membrane equilibria and that all the effects of electrol_ytes on osmotic pressure swelling and viscosity of proteins can be calculated with satisfactory exactitude by Donnan's equili- brium equation. Colloidal chemistry therefore appears to be only an imperfectly observed condition of equilibrium of the classical chemistry a t least so far as it deals with the proteins. The mistake is due to two causes first the omission of colloid chemists to measure the hydrogen-ion concentration of their solutions which is the chief variable factor in these cases and second the failure to measure and take into consideration the membrane potential of the solutions and protein gels which furnishes the proof that the theory of membrane equilibria must be used to explain the colloidal appearance of proteins.J. F. S. Variable and Invariable Properties of Dispersion. TV. OSTWALD (Kolloid Z. 1923 33 300-306).-A theoretical paper in which it is shown that the independence of for example the osmotic pressure and the mean kinetic energy of a particle on its degree of dispersion is rather an intuitively derived fact than a rigidly derived consequence of the molecular kinetic theory. On the other hand the thermodynamic or the capillary physical theory of disperse systems gives a strict explanation of this remarkable independence.The osmotic pressure mean kinetic energy density refractive power and cataphoretic velocity belong to the group of properties which are proportional to the absolute surface of the particlea and also to the square of the specific surface. If the measure of these properties is given by W then W=R920(o/v)* or oln . (o/v)a=K where o is the surface v the volume n the number of monobperse particles and K a very general topographical con-QENEBAL AND PHYSICAL CHEMISTBY. ii. 97 stant which depends entirely on the shape of the particles ; for cubes K = 216 and for spheres K = 113. This constant in prticular is independent of the degree of dispersion.The experimentally observed independence of the above-named properties of the degree of dispersion is due therefore to the fact that these properties depend in a double sense on the degree of dispersion. These two functions are opposed to one another and in the h a 1 result eliminate each other. When one or other of the two functions varies so that the above-named condition is not fumed exceptions to .the secondary independence appear. The properties which are most strongly dependent on the degree of dispersion are those which are governed by only one of the functions Determination of Size and Distribution of Size of Particles by Centrifugal Methods. T. SVEDBERG and J. B. NICHOLS ( J . Amer. Chem. Soc. 1923 45 2910-2917).-Stokesys law has been modified to give an exact formula for determining the radius of a particle sedimenting under centrifugal force.This formula has the form r = d9qloge (x+a)/a!d2(4-dl)w21 where T is the radius a the distance from the axis of rotation to the meniscus of the sol in a centrifuge tube x the distance which the boundary has moved in time t,. r] the viscosity of the liquid and dp and dl are respectively the densities of the particle and the dispersion medium. A special type of centrifuge is described which permits a sol to be observed or photographed while it is being precipitated. This method depends on the projection of a uniform beam of light up through a tube con- taining the material each time the tube passes over a certain point. The rate of movement of the particles in the tube may then be observed.To illustrate the method for a fairly uniformly sized colloid results for two different gold sols clay barium sulphate and arsenious sulphide are given. Another method is discussed for determining the distribution of size of particles depending on the variation of concentration with distance from the axis ot rotation in a disperse system subjected to centrifugal force'. P. FINKLE H. D. DRAPER and J. H. HILDEBRAND ( J . Amer. Chem. Soc. 1923 45 2780-2788). -A paper in which the present position of the theory of emulsification is outlined. The types and relative stabilities of emulsions formed by the aid of various soaps have been explained by the aid of the theory of orientation of the soap molecules in the interface. The curvature of the film of soap adsorbed at the interface is more convex towards water yielding more stable emulsions of the oil-enclosed type the larger the metal atom in the. soap.As the number of hydrocarbon chains attached to a single metallic atom increases the curvature is reversed becoming strongly convex towards the oil phase with soaps of the tervalent metals aluminium and iron which yield the most stable emulsions of water in oil. Experimental data on the type and relative stabili- ties of emulsions of various liquids with water show that the transi- tion from the most stable oil-enclosed to the most stable water- J. F. S. ~ ~ _ _ _ ~- J. F. S. Theory of Emulsification.ii. 95 ABSTRACTS OF CHEMICA& PAPERS. enclosed emulsion for both stearates and oleates follows the order caesium potassium sodium calcium silver magnesium zinc aluminium iron This order is in accordance with the valencies and atomic diameters of the metals as interpreted by the orientation hypothesis.It is suggested that the type of emulsion produced by a solid powder is determined by the angle of contact of the inter- face with the solid. In order for the powder to remain in the inter- face the angle must be finite and unless the angle is go" the interface will be on one side or the other of the points of contact of the particles and its tension will cause the film to be concave on t,hat side. J. F. S. Preliminary Attempt to Measure Gravimetrically the Distance Effect of Chemical Affinity. T. W. RICHARDS and W. T. RICHARDS (Proc. Nut. Acad.Xci. 1923 9 379-383).- After a brief discussion of attempts which have been made to ascer- tain the law connecting the force of cohesion and distance a number of experiments designed to measure the force of chemical affinity a t short distances are described. The experiments consisted in suspending horizontally a plate of aluminium 6 cm. sq. to one arm of a delicate balance and after counterpoising it bringing materials which have considerable affinity for it underneath it and measur- ing the pull by the change of apparent weight,. The materials used were oxides of silver copper iron zinc and magnesium sulphur iodine and bromine the two last-named substances being covered with mica. I n the case of the oxides and sulphur the distance between the aluminium and the oxide was about 0.001 m.with iodine 0.01 mm. and with bromine 0.02 mm. The experiments shorn that in no case is there an attractive force amounting to 0.1 mg. Hence it is concluded that the force of chemical afkity must decrease very rapidly as the distance between the attracting atoms increases. J. F. S. Extremely Dry Liquids. G. N. LEWIS ( J . Amer. Che,m. Soc. 1923 45 2836-2840).-A theoretical pa er in which it is shown tive desiccation of liquids which seems to be consistent with thermo- dynamics rests on the assumption that water is a catalyst for pro- cesses between various molecular states and that its removal merely inhibits such processes. If this explanation is correct the process of drying only " freezes " an existing equilibrium and (at constant temperature) cannot alter the static properties of a liquid.This leads among other things to the prediction that liquids will be found which exhibit abnormally low as well as abnormally high boiling points. J. F. S. that the only plausible explanation of the e B ects produced by exhaus- Theory of Chemical Reactivity. F. 0. RICE ( J . arner. Chem. Sm. 1923 45 2808-2820).-A theoretical paper in which an ex- planation for the high temperature coefficient of chemical reactions is offered; it is assumed that the law of mass action in its classical form is true and that its apparent failure is due t o usingC ENERAL AXD PIiYSICdL CEENISTRY. ii. 99 stoicheiometric equations which do not represent even approxi- mately the reactions taking place in solutions. When the ordinary stoicheiometric equations are replaced by equations representing more nearly what occurs it has been shown that certain molecules are connected with an equilibrium constant in such a way that their concentrations vary with the temperature.These are referred to as active molecules and it is due to these that “ slow ” reactions have a high temperature coefficient. The hypothesis predicts that chemical reactions will fall into comparatively few classes each class having a characteristic temperature coefficient. In this hypo- thesis the assumption is made that the non-hydrated hydrogen- and hydroxyl-ions are the catalytically active particles and this leads to the conclusion that stoicheiometrically neutral water is distinctly alkaline catalytically and it is not until the hydrogen-ion concentration has a pE value about 5 that the concentrations of the non-hydrated ions become equal and the catalytic activity is a t a minimum.J. F. S An Extension of the Equation for the Velocity Constant of a Unimolecular Reaction. H. J. PRINS (Chem. WeehNud 1923 20 686-689).-Starting from the assumption that in the ideal gas the relation between the atomic energy and the combining energy is the same for the greater proportion of the individual molecules but in a few individuals may vary from the average value and combining the mathematjical probability of any given configuration so deduced with the relation between the stability and entropy of the system it is possible to arrive a t Boltzmann’s equation X=klog W+A where X is the entropy and W the probability of the given configura- ation at a given instant; then from the assumption that the small proportion of activated molecules take up radiant energy from the medium as expressed in the photochemical law of Einstein it is possible to form a conception to account for the relation v=vN for the velocity of dissociation of simple molecules in the gaseous The Rate of Hydrolysis of Methyl Acetoacetate.G. LJUNGGREN (Ber. 1923 56 [B] 2469-2471).-The hydrolysis of methyl acetoacetate by sodium hydroxide has been re-examined since the velocity constants observed by Goldschmidt and Oslan (A. 1900 i 132,373) and Goldschmidt and Scholz (A. 1907 ii 244) diminish rapidly for a reason which these authors have not been able to explain. A source of error in Goldschmidt’s experiments is caused by the ketonic hydrolysis of methyl acetoacetate which pro- ceeds fairly rapidly in acid solution and hence commences rapidly when the alkaline solution is introduced into an excess of acid in order to stop alkaline hydrolysis.The main cause of error how- ever lies in the unsuitability of phenolphthalein for the titration in the presence of the markedly acidic methyl acetoacetate. The previously observed irregularities disappear when phenolphthalein is replaced by Bromothymol-blue. The velocity constant is 0.01211 this value being lower than that observed by Goldschmidt. phase. s. I. L. H. W.ii. 100 ABST&AC!FS OF CHEIk5CA.L PAPERS. The Formation of Aniline Black a BimolecuIar Reaction. J. PICCARD and F. DE MONTMOUIN (Helv.Chirn. Act% 1923 6 1021-1029).-The oxidation of aniline by potassium dichromate and sulphuric acid to Aniline-black has been found to be a reaction of the second order (bimolecular). The experimental method used was the comparison of the rates of reaction a t different concentrations of aniline these being measured by the time taken for the solution to attain a given degree of opacity. It might be supposed that there may be a rapid unimolecular reaction a t first as suggested by Gold- schmidt (A. 1920 i 226) followed by a measurable bimolecular reaction but this suggestion is disproved by the recovery of un- changed aniline from solutions in which oxidation has been cut short. The reaction velocity increases with increasing acidity but the order of the reaction is not affected.The question is complicated by the discovery that the oxidation of aniline to Aniline-black is autocatalytic. If the velocity at any moment is proportional to the concentration of aniline present and to the amount already oxidised which is also proportional to the initial amount of aniline then the observed reaction velocity is a t least proportional to the square of the aniline concentration. The reaction is then unimolecular in Goldschmidt’s sense but bimolecular from the point of view of reaction velocity. It is impossible to draw conclusions regarding the intermediati products fiom the reaction velocity. E. H. R. Second Report of the Committee on Contact Catalysis. W. D. BANCROFT ( J . Physical Chem. 1023,27 Sol).-A review of the experimental work done during the last two years.Catalytic Metals. J. PICCARD and E. THOMAS (Helv. Chim. Acta 1923 6 1044-1045) .-The precipitate obtained by reducing copper sulphate with a chromous salt contains when freshly pre- pared 98.9% of copper. It is therefore essent4ially metallic copper not cuprous oxide as stated in Abegg’s handbook. A very active form of silver is obtained by reducing a solution of silver perchlorate with a solution of chromous acetate in perchloric acid. It forms a white powder d 1.7 containing 99.6% Ag and is twice as active as “ molecular ’’ silver prepared by the action of zinc on silver chloride. The activities were compared in the catalysis of the decomposition of chromous chloride by the reaction 2CrC12+2HC1 = 2CrCl,+ H,. A catalytically active colloidal palladium solution is obtained by dissolving palladium hydroxide in glacial acetic acid and reducing with hydrogen It is valuable for promoting reduc- tions with hydrogen.Hydrogenation of Fats. I. Influence of Various Sub- stances on the Nickel Catalyst. G. KITA and T. MAZUME (Mem. CoZZ. Sci. Kyoto 1923 3 81-94).-The influence of the addition of aluminium oxide magnesium phosphate magnesium oxide calcium borate calcium phosphate sodium carbonate stearic and palmitic acids and moisture respectively to the nickel cakalyst was studied as regards the volume of hydrogen absorbed and the E. H. R.GENERAL AHD PHYSICAL CHEMISTRY ii. 101 rate a t which this absorption took place. The effect of these addi- tions was found to depend on the conditions under which the mixing was carried out.In most cases optimal proportions of the mixture exist which give the maximum rate of absorption of hydrogen this optimum mixture depending on the conditions under which the mixing was effected. In most cases a much greater improvement in the activity of the catalyst was obtained if the exciting substance was added to the nickel oxide before its reduc- tion this being especially the case with aluminium and magnesium oxides and calcium borate. In the case of calcium phosphate an exceedingly active catalyst was obtained by adding 12% of calcium phosphate to nickel carbonate and reducing the mixture. The addition of stearic and palmitic acids to the oil to be hydrogenated was found to have a beneficial effect on the rate of hydrogenation especially the former acid in quantities up to 10% of the oil.Sodium carbonate was found to have a retarding effect on the catalyst in all the proportions studied. The oil used in all the above experiments was soja-bean oil Tehed with sodium hydroxide and hydrogenation was carried out at atmospheric pressure and a t 170". The presence of moisture in the catalyst whilst not specially affecting the activity of a powerful catalyst was found to have a serious disturbing influence on a weak catalyst. This feature and the existence of optimum quantities of exciting substances suggests an analogy with the action of enzymes. H. C. R. Heterogeneous Catalysis. L. GURWITSCH (2. physikd. C'henz. 1923 107 235-248).-The author describes a number of cases of heterogeneous catalysis from which he draws the conclusion that such catalysis is best explained by adsorption and the formation of adsorption compounds.There is no need to assume the formation of definite intermediate compounds. Among the examples quoted are the polymerisation of pinene in the presence of partly dehydrated floridin (Florida earth). In this reaction a considerable amount of beat is evolved and the products consist of camphenes and polyter- penes. The amount of polyterpenes produced is less the lower the temperature a t which the reaction take place. The amount of pinene changed is greater the smaller the water content of the floridin down to 6.19y0 of water a t which point 99% of the pinene is changed and of this 4.9% is converted into polyterpenes.The greater the water content of the floridin up to 30% the greater is the percentage of the changed pinene converted into polyterpenes. The mechanism of the reaction is therefore the formation of poly- terpenes as primary product due to the adsorption of pinene by the floridin; the polyterpenes are then broken down by the heat of the primary reaction with the formation of isomeric monoterpenes. The water of hydration of the floridin takes a part in the reaction and the active force which brings about the polymerisation is the force of attraction between the pinene and the strongly dehydrated floridin. Pa,rtly dehydrated " grown " alumina prepared according to Wislicenus's metthod catalyses the reaction between carbon disulphide and water giving hydrogen sulphide and carbon dioxide,ii.102 .ABSTRACTS OF CHEMICbl; PAPERS. and in the same way ethyl chloride and water give ethyl alcohol and aluminium chloride. Precipitated alumina whether moist or partly dehydrated or moist "grown " alumina have no such action. The reaction is explained by the adsorption of the carbon disulphide or ethyl chloride by the alumina which is thereby brought into intimate contact with the water and so into reaction. The practice of coating glass vessels with paraffin wax to reduce the decomposi- tion of hydrogen peroxide is quoted and in this connexion it is deduced theoretically that since high molecular paraffins have a greater power of adsorption than lower molecular paraffins the former will catalyse the decomposition of hydrogen peroxide more than the latter and also organic compounds containing oxygen having a greater physico-chemical energy have a greater adsorptive power than the paraffins consequently they catalyse the decomposition of hydrogen peroxide more than the paraffins.It follows therefore that adsorptive power and catalytic power run parallel. These conclusions are confirmed experimentally for it is shown that the catalytic power with respect to the decomposition of hydrogen peroxide increases in the order paraffin m p. 52" palmitic acid paraffin m. p. 88" cerotic acid. Other cases which also show the importance of adsorption in heterogeneous catalysis are quoted. J. F. S. Possible Reconciliation of the Octet and Positive-Negative Theories of Chemical Combination. W. A. NOYES (J. Amer. Chem. Soc. 1923 45 2959-2961).-A discussion of what happens to the electrons on the separation of two atoms from one another. The two alternative views namely that one electron remains attached to each atom in which case both atoms would be electro- neutral and that both electrons remain with one atom making it electro-negative whilst the other is electro-positive are considered. It is shown that an atom does not always separate from a compound in the same electrical condition as it enters it. H. G. GRIMM and K. F. HERZFELD (Physikul. Z. 1923 24 486-488).- The authors discuss the question of how loosely a valency electron mist be associated with the nucleus in order that it may become detached in a chemical reaction. The heat of formation of a metallic compound is regarded as numerically equivalent to the algebraic sum of the heat of sublimation of the metal the heat of ionisation of the vapour and the heat of dissociation of the kation together with the energy due to electronic affinities and that associated with the ionic lattice. The respective heats of formation of the fluorides of the elements in the third period of the periodic classiiication are deduced in this manner and attention is directed to an analogy be- tween the values so calculated and the respective heats of ionisation of the corresponding elements. Deductions are then made as to the possibility or otherwise of the production of the respective fluorides in which one two three or four fluorine atoms are associated with a single positive atom as the result of chemical reaction. J. F. S. Chemical Valency from the Point of View of Energy. J. S. G. T.INORGANIC CHEMISTRY. ii. 103 The Construction of Simple Micro-balances. H. 6. DENHAM ( J . T a t . Inst. 1924 15 T. 10-13).-A short-beam quartz fibre micro-balance is described; it is intended for use with a graticuled reading telescope and a graduated scale or by it mirror galvanometer method and has a sensitivity of 2 x lo4 mg. with maximum load about 1 mg. A short review is given of the subject of micro- balances generally. J. C. W.