ii. 37 General and Physical Chemistry. The Optical Behaviour of Water of Crystallisation. KATHE BRIEGER (Ani~. Physik 1918 [iv] 57 2 8 7 4 2 0 ) .-Experi- ments have been made to compare the reflection intensities of crystals of a number of hydrated salts in the infra-red region h=2+5-7*.0 l ~ with that of water. The reflection curve for water shows a well-marked maximum intensity a t h=3-07 p and very similar maxima are found for hydrated crystals (compare Schaef er and Schubert) n. 1916 ii 5X). Of the nine isomorphous alums examined eight show two maxima in the intensity curves in place of the single water maximum a t 3*07p their positions being near 3 . 0 2 ~ and 3-51 p but differing slightly in the different alums. The curve for czsium alum however shows a third maximum and in this respect is abnormal.The water maximum a t A= 6.22 p also becomes doubled in the case of the alums. As examples of uniaxial crystals the hexahydrated nickel sulphate and selenate (tetragonal trapezohedral) were examined. I n the case of these salts the water maximum is shifted only slightly. The intensity of the reflectioii depends on the position of the optic axis relative t o the reflecting surface the water of cryst allisation exhibiting dichroisin. Siilzilarly in the case of mono- clinic crystals of the magnesium sulphate and zinc-ammoniuni sulphate groups the reflection properties are closely related t o the optical properties the water being trichroic. It is concluded that the symmetry properties of the water in hydrated salts correspond with the symmetry of the cryst’als.The bearing of the results on Werner’s theory of the constitution of hydrated salts is discussed. Preliminary experiments with analcime indicate a shift of the water maximum towards the shorter wave-lengths whilst in every other case the shift is in the other direction. Further experiments may be expected to throw light on the constitution of the zeolites generally especially with regard to the state of combination of the water which bhey contain. Effect of the Electric Field on Spectrum Lines. VII. The Fowler Helium Series. J. STARK 0. HARDTKE and G . LIEBERT (9 mi. Phyxik 1918 [iv,] 56 569-576. Compare Fowler A. 1913 ii 811).-The effect of an electric field on the lines A3203 and A2733 of ths Fowler helium series has been studied.The effect is compared with that on the B a h e r hydrogen series and it is shown that both effects are very similar the lines in both cases being symmetrically resolved into their components. The intensity of the two components is eqiial. Effect of an Electric Field on Spectrum. Lines. VIII. New Principal Series of Helium Lines which Appear in the Electric Field. J. STARK ( A m . Physik 1918 [ivl 56 577-588. See preceding abstract) .-During an investigation of the influence of an electric field on the helium spectrum two new E. H. R. J. F. S. VOL. CXVI. ii. 2ii. 38 A3STRACTS O F CHEMICAL PAPEES. principal series of lines were ciiscovered. The lines of the first of these the He 1 principal diffused series are given by the formula v =h'~(u) - Lz(m,d). i n the electric field the lines of this series are resolved in such a way that the components do not lie sym- metricsliy but are displaced about 1 A.unit in the direction of shorter wave-lengths. The second new series of helium lines the He Z sharp series is represented by the formula v = Ez(h) - Lz(m,s) The lines of this series in an electric field were not resolved into several conipoxents but were displaced towards the longer wave- leiigths. The He+ ion possesses twelve series which can be arranged in two systems of six series. I n each system there are three principal series with the same end-number and three sub- sidiary series also with the same end-number. Effect of an Electric Field on the Ultra-violet Lines of Helium. G. L~EBERT (B?zn. Physzk 1918 Lit.] 56 539-609.See preceding abstracts) .-The ultra-violet lines of the principal series He I are displaced towards the red end of the spectrum under the influence of an electric field. Both the components undergo the same amount of displacement-. The lines of the second subsidiary series are displaced without resolution toward the red by an electric field. In this case the displacement is rather more than the proportional amount. The lines of the principal series He I1 are displaced toward shorter wave-lengths by an electric field. I n the case of the third subsidiary series of He I a displacement of the lines toward longer wavelengths is brought about by an electric field. J. F. S. A New Helium Series under the Influence of an Electric Field. G. LIEBEHT (Bniz. Physik 1918 Liv] 56 610-616).- Corresponding with the third Me I series a subsidiary He ZZ series becomes visible in a strong electric field whereas in a weak electric field i t has no noticeable intensity.The electric effect on every member of the principal series corresponds with the effect on the similar member of the subsidiary series. The lines of the sub- sidiary series like those of the principal series are displaced toward the shorter wave-lengths. Spectrum of the Chromosphere. An Eruptive Protuber- ance of the 6th July 191'7. P. CARRASCO (Ailtal. Pis. &aim. 1918 16 700-706) .-A comparison of the author's observations of the chromosphere spectrum with the results of Rowland and of Adams. The presence of the helium series 6678.1 4922.1 etc. is noted as a characteristic of the spectrum.The helium series 7665.6 4713.2 etc. was not observed. Optical Researches on the Constitution of Sulphurous Acid its Salts and Esters. KONR. SCHAEFER [with WILLY KOHLER] (Zeitsch. anorg. Chem. 1918 104 212-250).-Aqueous solutions of sulphur dioxide show an absorption band in the ultra- violet in the same position as the band given by sulphur dioxide gas although somewhat broader. Beer's law is not followed the J. F. s. J. F. S. W. S. M.QENPRAL AND PHYSICAL CHZMISTRY. ii. 39 baiid becoming much shallower wibh increasiug dilution. The absorbing substance in aqueous solutioiis is a hydrate of sulphur dioxide SO,. ..OH which appears t.0 be aiore active than sulphur dioxide itself. Since the normal sulphites and the sulphite esters Et*SO,.OEt and f30(O&t)2 are both transparent in the ultra- violet it is coriciuded that sulphurous acid itself is non-absorbing both i n the unionised and ionised condition.The diminished absorbing power of solutions of sulphur dioxide on dilution is attributed t o a change oi equilibriuin from left to right in the equation SO,. . .O H zz H2kW3. The aqueous solution contains very little suiphurous acid and it is considered that measurements of the ion concentration in sulphur dioxide solutions have given rise to false ideas regarding the degree of dissociation of sulphurous acid and consequently of the strength of the acid. At higher temperatures solutions of sulphur dioxide absorb more strongly owing to a shift in the equilibrium of the above equation towards the left.Similarly addition of sulphuric acid to an aqueous solution of the gas increases the depth of the band in the same manner butl only t o a certain limit. When the sulphuric acid exceeds 5X the absorption again decreases prob- ably owing to dehydration in the sense SO,.. .OH2 5- SO,+ H20 the free sulphur dioxide being a less strong absorber than its hydrate. I n pentane and chloroform solutions the absorption is similar to that of the free gas aud the solutions follow Beer’s law. In solutions of methyl and ethyl alcohols and ethyl ether the absorption is much increased and in the case of ethyl ether the band passes into general absorption in the ultra-violet. These solu- tions also follow Beer’s law and it is concluded that stable com- pounds of the type 0,s ...OHEt and 0,s ... OEtS must be present in such solutions. Normal sulphites show oiily end absorption in the extreme ultra- violet and when a solution oE a riornial sulphite solution is half neutralised with sulphuric acid the resulting metal hydrogen sulphite solution atl first shows only end absorption but after a few days particularly under the influence of light the sulphur dioxide band develops. When the metal hydrogen sulphite is formed how- ever by mixing solutions of sulphur dioxide and normal sulphite the band is present immediately. A solution of potassium pyrosulphite ‘ when freshly prepared shows only end absorption in the extreme ultra-violet but the band gradually develops. Similarly sodium methyl sulphite which in methyl alcohol shows only end absorp- tion in aqueous solution gradually develops the sulphur dioxide band more quickly in stronger solutions owing to hydrolysis into sodium hydrogen sulphite and methyl alcohol.The equilibrium conditions in aqueous sulphur dioxide solutions are represented by the scheme EO f SO,. . .OH p]‘ O H [ OSOH 0 -j’ [ OS0 0 - p [ OS0 01’’ t- IT‘.G. 40 ABSTRACTS 03’ 0HEIILIICA.L PAPERS. The hydrogen sulphite ion HS03/ is probably unstable and under- goes transformation into sulphurous acid and SO,” thus ZHSO,’ = SO,’ + H,SO,. The sulphurous acid then comes into equilibrium with sulphur dioxide hydrate in the sense €12S0 SO,. . .H,O. These changes would explain the gradual develop- ment of the absorption band in hydrogen sulphite solutions. The view is expressed that the sulphite ion [SO,]” has a symmetrical constitution and hence also the normal sulphites.This view is supported by the observation that selenious acid and the dialkyl selenites have similar absorption spectra t o the normal sulphites (compare A. 1917 ii 61 186; 1915 ii 389). The Ethylene Chrornophore. H. LEY (Ber. 1918 51 1808-1819. Compare Ley A. 1917 i 261; Kauffmann ebenda. 391).-The absorption spectra of styrene stilbene and cinnamie acid and their derivatives have been studied. Bathochromic effects are produced by the introduction of a methyl methoxy- or ethoxy-group in the o-position in styrene whilst a hypsochromic effect results when the methyl or ethoxy-group is substituted in the a-position ; hypsochromic effects are also observed with 7-methylstilbene P-methylcinnamic acid P-ethoxycinnamic acid and a-methyl- cinnaxnic acid as compared with the parent substances.Attempts are also made to connect absorptiomekic anomalies with abnormal values f o r the molecular refraction. The esters of a- and P-methylcinnamic acid are less absorbent than the parent substance and have lower values €or the exaltation Za. &Methylstyrene ’has an absorption similar to that of styrene and its refraction anomaly is of the same order. With the hydroxy-derivatives a somewhat similar parallelism is observed the hypsochromic eff eet of the ethoxy-group in the a-position corresponding with a distinct diminution of the exaltation. Since in general an increase in refractive and disper- sive power corresponds with an increase in the unsaturated nature of the coImpound the explanation of the “ disturbing ” action of substitGents on the absorption spectra of substances by the assump tion of alteration in the degree of unsaturation appears t o receive f uather suppori;. The extinction-coefficients of chloroform solutions of I 3 5-tn- nitrobenzene and styrene a-ethoxystyrene B-ethoxystyrene and B-methoxystyrene respectively ’have also been measured but the ’ results scarcely lead t o a definite conclusion.Kauffrnann’s criticism of the author’s views (Zoc. c i t . ) is reviewed. E. H. R. H. W. Crystalloluminescence . 11. Triboluminescence and Crystalloluminescence. HARRY B. WEISER (J. Physical Chem. 1918 22 576-595. Compare A. 1918 ii 419).-For the further investigation of the relationship between crystalloluminescence and triboluminescence the phenomena have been studied with crystals of arsenic trioxide and the double sodiunilpotassium sulphate 2K2SO,,Na+3O,. In each case the colour of the light emitted by the crystals during rapid formation (crystalloluminescence) and t.hat>GENERAL AND PHYSICAL CHEMISTRY.5. 41 produced by rubbing or crushing the crystals (triboluminescence) have been determined by photographing through an appropriate series of light filters and it is shown that the light rays resulting from the two phenomena are identical in colour that from arsenic trioxide differing considerably however from that from the double salt. The colour of the luminescence is therefore a specific property of the compound. The phenomena are held to be chemical in origin.Crystalloluminescence which is only exhibited by a few inorganic compounds is due to the rapid re-formation of molecules broken up by the process of electrolytic dissociation. Triboluminescence is due to the re-formation of molecules broken LIP by t h s violent disruption of crystals. Whilst all crystalloluminescent substances are also tri- bolumir,escent the reverse is not true. I n certain newly-formed crystals a state of strain may exist and if outside force is applied t o such crystals before the internal forces have become adjusted violent disruption of the crystal may ensue with consequent molecular decomposition and triboluminescence. The intensity of the tri- boluminescence depends on the force needed to break u p the mole- cule and the magnitude of the coilversion factor for light in the luminescent reaction.E. H. R. Spectrum of the Ruby and Emera'd. IV. Spectrum Phenomena in the Chromium Compounds. JAMES MOIR (Trans. Roy. SOC. S. A fTz'ca 1918 7 129--130).-It has been shown previously (ibid. 2 321 ; 3,271,273 ; 4,339) that the almost unique spectra of ruby and emerald are due to chromium oxide which has been compelled to vibrate in an abnormal or constrained manner leading to the production of narrow absorption bands ; the constrain- ing substance in the case of the ruby is crystalline alumina and in the case of the emerald it is glucinnm silicate. The present com- munication describes attempts t o induce the constrained vibration of ordinary chromium oxide by artificial means.Suitable methods consist in dissolving chromic oxide in concentrated sulphuric acid and gradually addinq an organic substance such as starch until the orange colouF has changed to deep peen or more strikingly by replacing the sulp'huric acid by glacial phosphoric acid. It would appear possible to get a very close imitation of the emerald spectrum by the latter process; but the ruby spectrum on the other 'hand has not been reproduced except by means of alumina. Concentrated hydrochloric nitric perchloric. formic acet.ic and citric acids did not Geld very characteristic results although a faint hazy band a t A 6800 was seen in several of these solutions. R. W. A New Electrometer for the Measurement of Radioactivity. B. SZILARD (Anal. Fa's. Q z c ~ ~ . 1918 16 690-699).-A detailed description of a new irnmoved electrometer of the attracted alumin- ium needle type especially adapted for radioactive nleasiirements.W. S. M. Radioactivity of some Canadian Mineral Springs. J. SATTERLY and R. T. ELWORTIXY (Trans. Roy. Soc. Canada 1917-1918 [iii] 11 17-26).-The water of some sixty minerdii. 42 ABSTRACTS OF CHEMICAL PAPERS. springs has been esamiiied with respect t o its content of radium emanation and dissolved radium. Measurements of the rate of flow and temperature of the water are also given together with the nature of the strata. The emanation content varies from 11.2 x 10-12 curies per Iitre to 345 x 10-12 curies. The dissolved radium varies between the merest trace and 46 x 10-12 gram per litre although in most cases the amount is very small.A tabular comparison is given between the factors for these springs and those of American and European springs. Thermal springs are shown to be more active t.han the cold springs. J. F. S. The Radioactivity of Mud from Rockanje. E. H. BUCHNER (C’hem. 5VeekbZa$ 1918 15 243-246) .-Exniniiiations of samples of mud from the Waal lake near Rockanje (A.. 1913. ii 821) showed proportions of radium of the order of 1 x parts by weight. Subsequent; examination in t>he following year confirmed this figure. More recent work on samples of the same material by Folmer and Blaauw (A. 1918 ii 145) indicate a radium con- tent of the order generally found in common rocks and depositas namely 1 x 10-12. The author has examined some OE the samples tested by Folmer’ and Blaa17~ and confirms their results.The discov.dance between the later figures and the earlier result’s can- not be attributed t o differences in the methods of working o r to experimental error but is possibly due to deposition from the river water in isolated places of matter containing a high pi-oportion of some uranium mineral o r its degradation products. s. 1. L. The Radioactivity of the Mud from Rockanje. A. H. RLAAUW (Chem. Tl~eekblarl 1918 15. 361-364) ,-Biichner’s argu- ments (preceding abstract) t o account for the differences shown by the later analyses in the radioactivity of the Rockanje deposits are criticised and the asstumptions of patches of higher activity cruestioned. Other geological arquments suggested by Buchner are shown t o be incorrect.It is claimed that the later examination was complete. s. I. L. Ionisation and Resonance Potential €or Electrons in vapours of Magnesium and Thallium. P A m D. FOOTE miid FRED L. MOHLER (Phil. &fay. 1919 rvi1 37 33-50).-The ion- isation and resonance potentials for electrons in vapours of mag- nesinm and thallium have been measured by the Tate modification of the Pranclr and Hertz method. The experimental values for magnesium vapour were 2-65 volts and 7.75 volts €or the resonance and ionisation potentials respectively. The values calculated from the cruantum relation hv=eP are 3-70 volts and 7.61 volts respectively. I n the case of thallium va~oui- the values 1-01 volts and 7.3 volt’s were obtained for the reconance and ionisation Dotentids resnectively.The theoretical valii e f o r the former quantity is 1.07 volts whilst that of the latter is unknown. The ionisation potential for magnesium is obtained from the limit ofGENERAL AND PHYSICAL CHEMISTRY. ii. 43 the combination series 1 *5S - rnp2 and the resonance potential is determineJ by the iirst h e in this series m = 2 . I n the case of thallium the resonance potential is determined by the shorter wave-length member of the first term of the principal series of doublets. No known series in thallium has a convergence fre- quency greater than 49,263. The observed ionisation potential for thallium suggests the presence of an undiscovered series of single lines converging a t v = 1 *5S = 57,000 to 60,000. The present work offers evidence that the single line spectra of magnesium and thallium are h=4571 and 11,513 A.rekpectively. The general behaviour of the metals as regards ionisation and resonance potentials appears to be identical for metals in the same group of the periodic table. I n the case of thallium the component of the doublet having the higher frequency determines the value of the energy quantum absorbed by the atom. This fact! suggests either the possibility of separate excitation of the components of a doublet by electronic impact or a behaviour of thallium vapour thus excited analogous to the emission of characteristic X-rays for which the A? group does not appear until the energy of the impact- ing electrons is greater than that corresponding with X y . The authors have obtained a value of Planck’s constant h by the method of ionisation and resonance potentials.Using thirteen determinations on seven different metals the final mean value h=6*55 x 10-27 erg. sec. was obtained which is in excellent agree- ment with recent deterninations by other methods. The question of photoelectric sensibility of metallic vapours has been briefly discussed and recent work on ionisation and resonance potentials has been considered critically. J. F. 8. The Emission of Fositive Electricity by Salts of the Alkalies and Alkaline Earths under the Influence of Canal Rays. W. VOLKER (Ann. Physilc 1918 [iv] 57 257-277).-1t is shown that when a suitably prepared thin layer of lithiulrr chloride or of calcium sulphate connected with a conducticg hystem is subjected to bombardment by canal rays the induced currents can only be explained on the assumption that when the cathode fall of potential exceeds a certain limiting value positive charges are emitted by the salt.The limiting value of the cathode fall in the case of lithirxm chloride is about 5130 volts and with calcium sulphate about 1500 volts. These values correspond with the limit- ing values found by Stark and Wendt. (A.> 1912 ii 720) above which these salts under the influence of canal rays. emit the characteristi-c metallic series lines. It is concluded that the carriers which emit the series lines which are only observed in the space immediately adjacent Do the salt are the positive ion. of the metallic elements themselves. E. IT. R. Measurement of Concentration of Hydrogen Ions and a New Form of Calomel Electrode.W STURM (Chem. ;I/iTeekhlad 1918 15 912-916).-7’he author has designed aii. 44 ABSTRACTS OF CHEMICAL PAPERS. hydrogen electrode which reaches equilibrium in sbou t five minutes. This consists of a flask fitted with a rubber stopper pierced with three holes. The electrode which passes through one of the holes consists of a glass tube open at its lower end and is surrounded by a cylinder of platinum gauze. An inner tube carries a copper wire connected to the platinum gauze by means of mercury and a platinum wire. Rydrogen enters through a tube in the second hole passes up the electrode and escapes through a side-tube. The third opening carries a thermometer. The arrangement has the advantage that! the fluid is throughout agitated in contact wit-h the hydrogen which results in a rapid equilibrium.The disadvantage however. is that a considerable volume of fluid is required. Oxygen is removed from the hydrogen by means of heated platinised asbestos; the presence of small quantities of oxygen must be stndiously avoided. A form of calomel electrode iq described in which mercury. mercury-calomel paste potassium chloride crystals and saturated potassium chloride solution are employed ; such an apparatus has been in use for eight months satisfactorily. S. T. T,. Electrolytic Conductivity in Non-aqueous Solutions. 11. Electric a1 Conductivity of p - Tolyltrimethylammonium Iodide in Water and several Organic Solvents. HENRY JERMAIN MAUDE CREIGHTON and D. RERBERT War ( J .I”ra~2kZin Znst. 1918 186 675-698. Compare ibid. 1916 182 745).- The electrical conductivity of p-tolyltrimethylammoniv.m iodide has been determined at 25O in water methyl alcohol ethyl alcohol propaldehyde anisaldehyde. benzaldehyde acetone Iormic acid acetic acid propionitrile. benzonitrile nitromethane nitrobenzene and epichlorohydrin. I n the case of some of the solvents measure- ments have also been made a t Oo and 1 8 O and the t,eniperature coefficient calculated. ~TolyltrimethVlamn1onitirn iodide like most. strong electrolytes does not follow Ost,wald’s dilution law in aqueous solution but+ is approximately in agreement with the Storch empirical dilution equation. The ionic conductivity of the ptolyltrimethylammonium ion a t 2 5 O in aqueous solution is 40.3.The equivalent conductivitv at infinite dilntion has been extra- polated for ?~-tolyltrimethylammoniui~~ iod.ide in all the solvents used and it is shown that. the values varg from dm =13 in anis- aldehyde to A m =I88 in acetone. Further the mapnitucle of this value bears a relationship t o the chemical constitiition of the solvent. which may he qenerally stated thus the equivalent con- ductivity of an electrolyte at infinite dililtion. 2nd conseanentlv the vetocitv of the ions is greater in an a h h a t i c solvent than in an aromatic solvent of corresnonding constitutioii. Thus A m f o nitrompthane= 115. for nitrohenzene 38 for mooionitrile 143 whilst for henzoiiitrile A x 4 4 . Tn a n homolocons wrieq the value for A w is preater the nearer- the solvent stands t o the heginninF of the series.For example A* for water=llfi. methvl ~lcohol 100 ethyl n!cnhol 48 formir acid 90. acetic acid 2 5 OfGENERAL AND PHYSICAL CHEMISTRY ii. 4.5 the various substituting groups the influence of the aldehyde group on the magnitude of A- is greatest whilst that of the carboxyl group is the least as is shown by the series acetaldehyde A =188 iiitroniethane 115 methyl alcohol 100 acetic acid 25. J. F. S. Electrolytic Deposition of Iron from Organic Solvents. E. H. *%RCHIBALD and L. A. PIGUET (Trans. Roy. s o c . Canada 1917-1918 [iii] P 1 107-112).-Solutions of ferric chloride in acetone ethyl alcohol and acetone and water and acetone have been electroiysed by various currents and a t various voltages at 250. The experiments were carried out' between platinum elec trodes but in some experiirieiits a silver cathode was employed.I n all cases the iron can be completely deposited; in the case of acetone and of acetone-water mixtures the iron is deposited free from carbon. At low voltages the iron is deposited in the metallic condition; as the voltage is increased i t comes down as a red deposit which is adherent and allows the iron t o be removed from the solution. From the acetone-alcohol solutions the iron although completely deposited is contaminated with carbon. Jbom acetone-water solutions of ferrous sulphate the iron can be completely deposited but the deposit? contains a considerable amount of carbon. Aldehyde is not formed during the electrolysis of any of the solutions. Electrolytic Precipitation of Zinc.D. MCINTOSH ( TTCL~ZS. Roy. SOC. Canada 1917-18 [iii] 11 113-119).-The electrolytic depoeition of zinc from solutions of the sulphate has been studied under various conditious with the object of ascertaining the most suitable conditions for obtaining good deposits. It is shown that the best results are obtained when the solution is free from colloids and when the amount of iron present> is low. The metals arsenic antimony copper cobalt nickel and all metals more electropositive than zinc must he absent. The solution particularly in tanks con- taining large amounts of acid should be cold. The zinc deposits in a semi-passive form ; but when it begins to dissolve solution. cannot be stopped in any simple way. The zinc concentration should be as high as possible (6-7"L) and n o attempt should be made t o electro- lyse solutions containing less than 1*5--2*0% of zinc.With the ordinary cascade system the maximum current density is 25-30 amperes per sq. foot. A large number of photographs of deposits obtained under varying coiiditions are reproduced in the paper. J. F. S. J. F. S. Critical Phenomena. WILLIAM R. FIELDINU (Chem. flews 1918 117 379-383).-The' relation between the critical pressure and critical temperature of an element can be expressed by the formula (T + 2363) / JPc = 70.9. This relation holds strictly for hydrogen and the members of the halogen group and less closely for nitrogen and xenon whilst in the case of oxygen the divergence is considerable. For series of inorganic compounds such as that of the 2"ii.46 ABSTRACTS OF CHEMICAL PAPEES. halogen acids a similar iormula holds (T - z) / JT= A where x and k are constants f o r any particular series. In inorganic compounds t h e critical pressure rises with the critical temperature in any series but in organic compounds the presstare falls as the temperature rises. For a number of organic compounds the value of T x d z has been calculated. In higher members of aliphatic series this quantity tends to become constant. Ia the aromatic series of hydrocarbons with the introduction of alkyl groups into the benzene ring it first decreases and as the complexity of the alkyl groups increases it becomes practically constant. Nultiplication of benzene’ rings in t h e molecule produces a considerable increase in the value of the quantity.E. H. R. Calorimetric Lag. WALTER P. WHITE ( J . Anzer. c’henz. SOC. 1918 40 1858-1872).-Lags may pertain t o the thermometer t o various portions of the calorimeter or to outlying bodies including air. I n this paper the lag of external bodies such as a thin metal shield surrounding the calorimeter is treated in a mathematical manner. The effects prove to be three one equivalent to a change in the heat capacity of the calorimeter which can be eliminated by direct calibration of the calorimeter ; another much smaller depending on the amount of thermal leakage which can be avoided by using the adiabatic method; and a third dependent on the jacket temperature which disappears if this is constant. On account of the lag effect the effective heat capacity of a shield midway between calorimeter and jacket is only one-fourth its actual capacity and with due regard to the possibility of change such shields may often be used to reduce thermal leakage.Used as a cover such a shield has a specially small error and offers a parti- cularly easy method of dealing with evaporation. (Compare A. 1918 ii 149.) E. H. R. The Conditions of Calorimetric Precision. WALTER P. Wmm ( J . ,4mer. Chem. Soc. 1918 40 1872-1886).-1n a Cali- brated calorimeter most of the errors arise in thermometric measurEments and most of these come in the ‘‘ cooling correction,” the determination of the effect of thermal leakage. The leakage effect is equal to K+,T where T is the time # is the difference between calorimeter and environment temperature and H the leakage modulus of the calorimeter.By diminishing Ii as by rneane of a vacuum the effect of errors in # is diminished. These errors arise from the difliculty in getting the jacket and calorimeter temperatures uniform. The advantage of diminishing # t h e thermal head is largely illusory. since the main error t h a t in deter- mining X by means of the cooling rate is little affected. Lags which vary with K and are independent of T and + can be made t o cause little or no error. The rate of stirring should be as uni- form as possible since the heat produced varies as the cube of the speed. The tvpe of stirrer and design of the calorimeter with reference to stirring may be important. E. H. R.GENERAL AND PHYSICAL CHEMISTRY ii.47 Calorimetric Methods and Devices. WALTER 9. WHITE ( J . Amer. Chem. Soc. 1918 40 1887-1898).-The general rules for calorimetric precision (preceding abstract) are applied to jacket covers and stirrers and to such special devices as vacuum-jacketed vessels the adiabatic method aneroid or dry calorimeters double or differential calorimeters and measured-shield calorimeters and the advantages of the different forms of calorimeter are compared. The measured-shield calorimeter is a new device having between the calorimeter and the jacket a thin metal shield connected to the wall o f the jacket by therino-elements which are used to measure the thermal head. E. €3. R. Sensitive Bath Thermostat. A. HORMAN MHAW (Tmzs. Boy. SOC. Camda 1917-18 [iii] 11 129-135).-A thermostat on a rather large scale is described which differs but little from the well- kfiown types.A nots on the regulation of temperature is appended t o the paper. An Accurate Method for Measuring the Density of Gases. 0. MAASS and J. RUSSELL (2. Arne?.. Chem. Soc. 1918 40 1847-1852).--The method described is applicable to' tlie deter - mination of the density of those gases which can be condensed by liquid air or some other freezing agent. The purified dry gas is condensed in a suitable vessel and thence transferred to a large exhausted glass flask (abouti 22 litres capacity) the volume of which is accurately known. The flask has r2 coniiection t o a manometer and the pressure is read a t Oo the flask being immersed in ice. By a suitable arrangement of tubes the gas in the flask can be transferred to an exhausted glass bulb immersed in liquid air.When the gas has liquefied the bulb is sealed off and weighed and again weighed after it has been opened and the liquid allowed to1 evaporate. The pressure of the gas remaining in the flask a i d connecting tubes is read after the bulb1 has been detached to enable the volume of gas taken into tile bulb to be calculated. The accu- racy of the method is limited by the accuracy with which the height of the mercury in the manometer can he read. The probable error with the method used is 0*0504; with tlie aid of a cathetometer it could be reduced t o 0.01%. The following densities were determined acetylene 1.1695 ; methyl ether 2.1103 ; hydrogen bromide 3.6397. The Change of Density of Liquid Mixtures with In- creasing Temperature.mT. HERZ (Zeitsch. anorg. Chem. 1918 104 251-252).-The formula expressing the relationship between density and temperature 1 / d =a - h log (0 - t ) where dt is the density at temperature t 8 is the critical temperature and n and 71 are1 constants has been found t o hold for liquid mix- tures as well as for pure liquids. Different binary mixtures of benzene toluene aniline and nitrobenzene were used for the experiments aC temperatures between 25O and 905 and the differ- ences between observed and calculated densities were very small. J. F. S. E. H. R. 2*-2K 48 ABSTRACTS OF CHEMICAL PAPERS. The critical temperature 6 of the mixture was calculated from those of the constituents 8 and 8 by means of the formula B = no + (100 -n)O2/ 100 where and (100 - n ) are the percentages of each constituent.E. H. R. Molecular Attraction and Attraction of Mass and some New Gas Equations. JAMES KAM (PhiE. Mag. 1919 [vi] 37 65-97) .-A theoretical paper in which gas equations are deduced from considerations of the actual and theoretical co-volumes. A general equation is deduced which considers the effect of tempera- ture on the relation of the co-volumes to the volume. Further reduced isothermals and reduced border-curves are considered. It is shown that the cohesive forces causing the “ inward pressure ” appear to follow a law similar to the inverse square law of mass attraction. From the deviations from the gas laws a value for the tensile strength of iron is deduced which is of the same order as the experimental value.It is further shown that whatever the ultimate nature of the cohesive forces may be they seem to be proportional to the square of the molecular weight and to obey the inverse square law-a Iww consequently similar to Newton’s law of attraction of mass. J. F. S. Internal Molecular Forces of Solid Substances and their Relations with the Elastic Properties. STEFANO PAGLIANI (Nuovo Cim. 1918 [vi] 15 i 103-129).-The existence of three forms of internal molecular forces in solids is considered and their values calculated. The first P termed intermolecular force of coherence enters as a factor in the internal work into which thermal or mechanical energy can be transformed without change in the volume of the substance. The second T termed internal pressure is analogous in character t o external pressure and enters as a factor in the internal work in which thermal energy is used with variation of the volume of the substance.The third called intermolecular force of change of phase enters as a factor in the internal work accomplished during changes of phase-in the fusion of solids; this force which acts a t the temperature of fusion where Lindemann considers that the amplitudes of the atomic oscillations about their equilibrium position become of the same order of mag- nitude as the mean distance between the atoms appears t o be of the same nature as P which acts with definite values a t different temperatures. The magnitude of the force F exhibits relationships with the values of some of the magnitudes considered in the study and appli- cations of the elastic properties of solids.The simplest of these relations with the moduli of elasticity od tension E of rigidity n and of flow N are as follows E=1.70 F or approximately E = 5 F / 3 n=F/?r and N=2F/7r. The relatively great value of the ratio between E and F depends an the fact that the load E which should double the init.ia1 length of the solid is only an ideal load unattainable in practice. It is hence more accurate to replace E in technical forrnulz by F this being a real magnitude the valueGENERAL AND PHYSICAL- CHEMISTRY. ii. 49 of which a t different temperatures may be calculated by means of physical magnitudes exactly measurable ; a number of expressions are given in which this substitution is made.The modulus of elasticity for each species of defo’rmation is found to be directly proportional to the inolecular force of coherence. The Adsorption of Arsenious Acid by Ferric Hydroxide. MAITLAND C. BOSWELL and J. V. DICKSON ( J . Amer. Chem. Soc. 1918 40 1793-1801).-The adsorption of arsenious acid from solution by ferric hydroxide wliicli had been prepared several months previously has been studied both in presence and absence of sodium hydroxide. The adsorption of the arsenious acid by the ferric hydroxide is diminished by the preeeiice of sodium hydroxide whilst the adsorption of the sodium hydroxide is itself increased by the presence of the arsenious acid. The age oE the ferric hydr- oxide and the addition of sodium chloride appear riot to have any effect on the quantitative data recorded.The results follow only approximately the ‘ I adsorption law,” E=PA4T’ where I? is the concentration of the adsorbed substance in the adsorbing pliase ,4 its concentration in the solution at equilibrium and /3 and p are constants. The equation can be written log E = p log A -I- B . Hence if the logarithms of the con- centrations in the adsorbent and the solution be plotted a straight line should be obtained the slope of which gives the value of p whilst the intercept gives the value of B =log P. The curves thus obtained from the results of the authors’ experiments approximate to straight lines but all show a distinct concavity t o the x or logA axis. The adsorption phenomenon cannot. be exactly represented by the above equation. The adsorption of arsenious acid by ferric hydroxide was also studied by Mecklenburg (A.1913 ii 676) who concluded that the adsorption curves for any particular gel p r e pared under varying conditions are so related t o a unit curve that the ratio of any ordinate on one curve to the corresponding ordin- ate on the unit curve is a constant. The authors interpret this to mean that the exponent p in the equation E=FfA4p is constant for any particular gel -whilst /3 is a variable depending on the condi- tions of preparation of the gel. The values of p calculated from t>lie authors’ and Mecklenburg’8 results agree when arsenious acid is used alone but in presence of sodium hydroxide! its value increases. E. H. R. T. H P. Ferment Action. I1 a Adsorption of Amino-acids and Poly- peptides by Animal Charcoal.Relation of the Observed Appearances to the Cleavage of Polypeptides by Yeast Juice. EMIL ABDERHALDEN and ANDOR FODOR (Fermentfomch. 1917 2 74-102; from Chem. Zentr 1918 ii 738. Compare A 1917 i 306).-Animal charcoal adsorbs scarcely a trace of glycine but it adsorbs a little alanine and a comparatively large quantity of leucine. Similar variations are observed in the case of the poly- peptides even isomeric substances being adsorbed t o very different extents by the charcoal. T”h0 amount adsorbed a t various dilutionsii. 50 BBSTBACTS OF CIiElCIfCAL PdPERS. varies according t o the laws of adsorption just as has been observed in tlie case; of enzymes. The adsorptive power trf charcoal differs from thati of the enzymes however in being uninfluenced by changes in the concentration of the hydrogeu and hydroxyl ions in the solution.When mixtures of amino-acids or polypeptides are treated with animal charcoal the adsorption of the substances containing t110 more complex molecules is relatively increased whilstr tihati of smaller molecular substances is decreased. H. W. B. Ferment Action. III. Adsorption of Amino-acids and Polypeptides and also of various Carbohydrates by Animal Charcoal. Emd ABDERHALDEN and ANDOR FODOR (Ferwzeitt- forsch. 1911 2 151-L66; from Chern. Zenffr. 1918 ii 739. Com- pare preceding abshract) .-Z-Arabinom lamdose and inositol are less readily adsorbed by charcoal than dextrose sucrose and other disaccharides and trisaccharides Tkre presence of carbohydrates diminishes the adsorptive power of animal charcoal towards pdy- peptides and conversely the adsorption of carbohydrates is adversely influenced by polypeptides.H. W. B. Retardation by Sugars of Diffusion of Acids in Gels. EVARTS A. GRAEAX and HELEN TEEDWAY GRABAX ( J . Anzer. Chenz. Soc. 1918 40 1900--1917).--The rates of diffusion of a number of inorganic and organic acids into gelatin sottitions containing vary- ing amounts of dextrose smxose and lactose have been measured. All the sugars have a niarked retarding influence on the diffusion of any acid the disaccharides having a considerably greater eEecl than the monosaccharide. In agreement with the general diffusion law the ratio of the distance of diffusion to the square root of the time is found to ba a constant in gelatin bath with and without sugar.The retardation is not proportional tol the concentration of the sugar but is relatively greater for smaller concentrations of sugar. I f the ratio d l Jf is equal t o KO without sugar and E with sugar %he relative retardation is given by (I~o-E')/.E'~~ or A/Ko. The relation between concentration-of sugar and retardalion can be expressed approximately by the equation (A. I go)" =CW where c is the concentration and a and .n are tonst.ants. I n 10% geIatiii nitric acid behaves abiiorrnalfy in that the ratio d i i s not constant but increases rapidly a t first and after remaining constant for some time again decreasw. Hydrochloric acid shows the1 same behaviour in 3% gelatin. Sodium chloride aha reta.rds the diff mion of acids into gelatin but to a less extent than the sugars when equimofecular solutions are compared. This is the reverse of their effect in reducing the acid sqcslling of gefatin in which sodium chloride is more effective than sugar The physiologkal bearing of the results obtained is briefly discussed particular1-y with reference to the protective actiun of dextrose in cases of chloro€orm poisoning the effects ofGENERAL AND rHYSICAL CHEMISTRY.5. 51 which are ascribed to t'he presence of free hydrochloric acid in the substance. The mechanism of the retardation is also discussed. E. H. R. Chemical Affinity in Crystals and the Velocity of Crystallisation. hl. PADOA (Atti €2. Accad. Lzncez 1918 [v] 27 ii 59-64; Gazzetta 1918 48 ii 139-147).-The suggestion is made that in rnost cases crystallisation represents a transposi- tion effected by means of valencies among the more or less free molecules of the gaseous or liquid phase so t h a t velocity of crystal- lisation may be regarded as velocity of reaction.The author has measured the velocities of crystallisation of a number of super- cooled compounds with the object of ascertaining (I) if with isomorphous compounds of similar structure the velocities of crystallisation are equal or of the same order of magnitude and (2) i f in a series of coapounds of perfectly analogons chemical coiistihtions the velocities of crystallisa tion are comparable and ceteris p r i b z t s depndent on the mo!ecular weights. Similar values are obtained f o r m-bromo- (600) and w -chloro-nitrobenzene (882) and also for p-clichloro- (4800) p-dibromo- (7000) and piodo-benzene (6300) the low value f o r p-chloroiodobenzeiie (1 153) being apparently dependent 011 its markedly difierent viscosity.The introduction of methyl into the benzene ring greatly retards the cry-stallisation ; naphthalene and its anzlogues show far lower values than henzene the isomeric anthrwene and phenanthrene 'having identical velocities. Sirrilaritv also exists between the velocities f o r isomorphous inorganic compounds. Of two compounds of similar structure the one ssturatpd and the other unsaturated the latter should nresent the ?rester velocitv of crystallisatior since the latent affinities in the dolihIe linkirirrs should favour the prwecs of crv4?llisstinn.This conclticion is comp!etely confirmed by tilie result8s oFtqiver7 for a niimher of v2ir of isomorphous orcanic compounds cne member of esch psir h i n g satizrated or a t least less u n s a t ~ a t e d than the other mewber. The general conclusicns drawn are (1) the velocitv of crvst.11- lisation is a constitutive propert-v and (2) the bonds hetween the atoms in crystalline networks are of the same nature as rliemical valencies. T. H. P. Point of Transformation betweeil Reversible Modifica- tions. C. VIOLA (Atti B. ilccad. Lii.tccz 1918 [v] 27 ii 107-112).-The author applies the law of maximum work and Curie's law t o the consideration of the possible co-existence a t 9 temperature different from i l l of two reversible crystslline modifica- tions which occur in contact in stable equilibrium a t temperature 2' and a t a definite pressure.It is shown that either where a process is effected reversibly and crystallisstion takes place and should take place according to Curie's law or where between two modifications in contact a minimal surface tension is established a point ofii. 5.2 ABSTRACTS OF CHEMICAL PAPERS. transformation must always exist there ; vice versa the surface tension is a minimum a t a transformation point between two modi- fications. T. H. P. The Artificial Coloration of Liquid Crystals. PAUL GAUBERT (Compt. rend. 1918 167 1073-1075).-Indophenol may be satisfactorily used for colouring liquid crystals and also solid crystals of a large iiumbei- of organic compounds. The presence of the indophenol lowers the solidification point of the molten substances and in the case of substances giving several biref ringent liquid phases all the melting points or transformation points are lowered.There exists for the ir:dophenol a coefficient of so1ubilit”y particular to each phase this coefficient diminishing with the molecular vollume. The optically positive liquid crystals obtained by combining cholesterol with glycollic acid glycerol etc. are i i ~ t coloured by indophenol although the isotropic liqvid dissolves i t . W G. Theory of the Velocity of Coagulation. 11. FREUNDLICH (fi-olloid Zeitsch. 1918 23 163-173).-A theoretical paper in which it is shown that the velocity of slow coagulation of hydro- phobe sols increases very much with increasing concentration of the coagulating electrolyte whilst for rapidly coagulating sols the velocity of coagulation is constant and independenC of the nature and concentration of the electrolyte.Srnoluchomsk explains the rapid coagulation by assuming that the collisions between the partlicles are noa-elastic and lead to the coagulation of all the particles which cqme within the sphere of influence. The rapid increase in velocity of the slow codgulation is explained by the assumption that in the concentration present the charges on the particles produce a repulsion and only those paxticles are coagu- fated the kinetic energy of which is greater than a certain criticaf value. This critical value is greater the greater the charge on the particle. The number of these particles increases as the electrolyte concentration increases.A f orinula giving the relationship between electrolyte concentration and velocity of coagulation is given and this is in keeping with observed results. J. F. S. The Brownian Movement and the Coagulation of Colloidal Solutions. H. T.V. WOUDS’L‘RA (Chem WeekbZnd 1918 15 679-680) .-The author claims priority over Kruyt and van der Spek (A. 1917 ii 563) in directing attention to the func- tion of the Brownian niovement of the colloid particles in the coagulation of a colloidal solution. w. s. M. The Brownian Movement in Relation to the Mechanism of Flocculation. D. J. HISSINK (Chem. TVeekbZad 1919 16 20-21).-The author disputes the contention of Henri and Ostwald (see Grundriss der allgemeinen Chemie,” 1917 544) that the Brownian movements are diminished by addition of a coagu-GENERAL AND PHYSICAL CHEMISTRY. ii.53 latiing electrolyte before coagulation and conkends that by micro- scopic observation i t can easily be seen that the coagulation itself is the cause of the diminution of the movement which is therefore a consequence of the coagulation and does not precede it. s. I. L. The Lowering of the Degree of Dissociation. K. N. DE HAAS (Chem. Weefiblad 1918 15 1352-1355).-An elementary mathematical consideration of the fundamental equation of mass action with reference t o the shifting of the equilibrium consequent on the addition of oiie of the products of dissociation or ionisation t o the system. W. S. M. Heterogeneous Equilibria between Aqueous and Metallic Solutions The Interaction of Mixed Salt Solutions and Liquid Amalgams.V. A Study of the Ionisation Relations of Potassium and Strontium Chlorides in Mixtures. G. McP. SMITH and EDWIN ARTHUR REES ( J . dnzer. Chem. Soc. 1918 40 1802-1847).-To study t.he equilibrium between mixed strontium and potassium amalgams and the mixed aqueous solu- tions of the chlorides of these metals a definite quantity of either strontium amalgam or potassium amalgam of known concentration was shaken up with successive portions of a known aqueous solu- tion of the mixed chlorides until equilibrium was reached. The quantities of strontium and potassium in the amalgam were then determined by analysis. Experiments were made on the effect of varying (1) the amalgam concentration up to about 0.6 Inilli- equivalent of potassium or strontium per 10 grams of mercury (2) the total salt concentration in the aqueous phase from 0’05&\- to 3Ar ( 3 ) the ratio of the two salts in the aqueous phase a t different total salt concentrations and (4) the temperature from 1 5 O to 40°.To interpret the data obtained the value of the ex- pression (Sr,,)(K-ealt)2/ (R,,)2(Sr-salt8) = C was calculated in which (SrKg) and (KH2) are the respective atom-fractions in the mercurial phase and (K-salt) and (Sr-salt) are the mol-fractions in the aqueous phase (compare A. 1917 ii 247). The value of the ‘’ equilibrium expression ” C increases proportionately with the total conceiitration in the mercurial phase up to 0.3 milli- equivalent per 10 grams of mercury a t temperatures from 15-45O.V7hen the amalgam concentration is kept constant and the con- centration of the aqueous phase increased the proportions of the two salts being kept constant the value of C first falls rapidly as the concentration rises to about 0-4fl but with stronger so’1utions i t approaches a constant value. Assuming that in the most dilute solutions examined 0*05N an equimolecular mixture of the two salts is completely ionised it is shown that as the concentration in the aqueous phase is in- creased but the salt‘ ratio kept constant the proportion of K+ ions t o Sr++ ions increases I-apidly; also whea the ratio ofii. 54 ABSTRACTS O F CHEMICAL PAPERS. potassium chloride to strontium chloride is increased the propor- tion of K+ ions increases at’ a greater rate.The results obtained at different temperatures have been used to calculat*e the heab of reaction accompanying the changes between 1 5 O and 40°. With rising temperature the reaction is exothermic and the heat of reaction is found to decrease with increasing total salt concentration. As regards the dissociation of potassium and strontium anialgams it is shown that the ordinary mass-action expression holds good for mixtures of t,he two up to a total concentration of about 2 milli-equivalents per 100 grams of mercury. There is evidence of the formation in the aqueous salt mixture of molecular complexes and complex ions the results being in harmony with ViTerfzer’s views. E. H. R. Influence of Substitution of the Components of Binary Equilibria in Solution XIII.The Solution Equilibria of the Three Isomeric Phenylenediamines with Phenols and the Dinitro-derivatives of Benzene respectively. ROBERT KREMANN and WOLFGANG STROESCHNEIDER (Monatsh. 1918 39 505-570. Compare this vol. ii 15).-Melting-point and cooling curves have been obtained for the six binary systems made up of the three phenylenediamines with either a-naphthol or P-naphthol f o r the four binary systems composed of a- or &naphthol with a- or P-naphthylamine for the system 8-naphthol and ptduidine for the nine binary systems composed of one of the three diphenylamines with one of the three dihydroxybenzenes for the three binary systems composed of 2 4-dinitrotoluene with each of the phenylenediamines and for the nine binary systems composed of one of the three phenylenediamines wit.h one of the dinitrobenzenes.The results are given in curves and in voluminous tables. It is shown that 1,-phenylenediamine forms compounds with two molecules of a-naphthol (m. p. l l O o ) and with two mde- cules of P-naphthol (m. p. 150-5O). One molecule of m-phenylene- diamine forms compounds with two molecules of P-naphthol (m. p. 1 1 4 O ) and with one molecule of a-naphthol (m. p. 359. One mole- cule of o-phenylenediamine f o m s compounds with one molecule of a-naphthol (m. p. 60°) and with one molecule of P-naphthol (m. p. 8 6 O ) respectively. a-Naphthol forrrs two compounds with a-naphthylamine with one and four molecules of the latter to one of the former colmnound. A compound is formed contailling two molecules of P-naphthol to three molecules of a-naphthvlamine.b u t no compounds are formed between a-naphthol and 8-naphthvl- amine a simnlie eutectic is produced by these substances a t 47O and contains 56.5% of the last-named component. An equimole- cular comnonn d ir formed between B-nanhthol and ptoluidine (m. p. 87-Fi0). Equimolecular compounds are forwed with m-ohenvlenedianiine and catechol frn . p. 64*5O) m,-nhenylene- diamine and reForcinn1 (m. p. ‘?go). v-phenvlenediamine and quinol (ni. 1’. 1 2 7 O ) . with two molecules of catechol and one mofe-GENERAL AND PXYSICAL CHEMISTRY. ji. 65 cule of p-phenyleuediamine with three iiiolecules of cat echo1 and two molecules of p-phenylenediamine and with three molecules of quinol with one molecule of p-phenyfenediamine.The details of the system o-dinitrobenzene-p-phenylenediamine have not been worked out because of a secondary chemical reaction which inter- feres with the measurements. Compottnds are formed between m-dinitrobenzene and the three pheiiylenediarnines of the follow- ing composition (i) three molecules of the nitro-compound with two molecules of o-phenylenediamine (ii) two molecules of the nitrecompound with one molecule of o-phenylenediarnine and (iii) m e molecule of the nitro-compound with two molecules of tn-phenylenediamine ; no compounds are formed between pphenylenediamine and m-dinitrobenzene. I n the case of pdinitrobenzene and the three phenylenediamines compounds are not formed with 0- and m-phenylenediamine but with p-phenylene- diamine a compound of two molecules of t-he amine and one mole- cule of the nitro-compound is formed.Full details of the numerous eutectics are given in the paper. The Dilution Limits of Inflammability of Gaseous Mixtures. 111. The Lower Limits of some Mixed Inflammable Gases with Air. ITT. The Upper Limits of some Gases Singly and Mixed in Air. HUBERT FRANK COWARD CITARLES WILLIAM CARPENTER and WILLIAM PAYMAN (T. J. F. 8 . 1919 115 27-36). The Propagation of Flame through Tubes of Small Diameter. 11 WILT,IAN PAYMAN and RICHARD VERNON WHEELER (T. 1919 115 36-45). The Inflammation of Mixtures of Ethane and Air in a Closed Vessel. The Effects of Turbulence. RICHARD VERNON WHEELER (T. 1919 115 81-94). The Ignition of Explosive Gases by Electric Sparks. JOHN DAVID MORGAN (T. 1919 115 94-104).The Range of Existence of Substances Kinetic Analysis and the Estimation of Vapour Pressures from Reaction Velocities. M. TRALTTZ (Zeitach. rmory. Chem. 191 8 104 169-210) .-The author distinguishes between the range of stability and the range of existence of any substance. The stability of a substance is dependent on the chemical factors which bring about its formation and is measured by the energy of the reaction whilst its range of existence is determined only by the heat of activation needed to bring about its decomposit'ion. The upper limit of existence of a compound is determined by the heat of activation of that binary decomposition which requires the least energy. From the point of view of the author's theory of reaction velocity (A. 1917 ii 23; 1918 ii 151) a large number of reactions areii.66 ABSTRACTS OF CHEMICAL PAPERS. discussed. The formation of nitrosyl chloride or of nitrosyl bromide is apparently a reaction of the third order but in all probability is made up of superimposed reactions of lower orders and formulz are developed by means of which the higher order reaction can be expressed in terms of reactions of lower orders. On the assumption that nitrosyl chloride formation proceeds in the two stages NO +- (21% NOCI NOC1 + NO = 2NOC1 a method of kinetic analysis is developed which enables tAhe concentration and heat of formation of the intermediate product t o be calculated. The heat of fomiation of NOC1 is 2500 cal. The reaction between hydrogen and oxygen in porcelain vessels was supposed by Bodenstein to he of the t-hird order but the velocity constants do not distinguish between a third- and a second- order reaction.Moreover probable values €or the impact consta,nt. are obtained only i f it is treated as a second-order reaction. The course of the reaction is probably B + 0 = H202 (measurable) ; H,O + H = 2H,O (imiiieasurably fast). The probable value of the heat of activation for N,+0 is 46,374 cal. and the mean diameter of oxygen and hydrogen molecules 1-62 x 10-8 cni. The formation of nickel carbonyl from nickel and carbon mon- oxide which is apparently a third-order reaction probably pro- ceeds in the three stages . Ni + CO = NiCO (momenta,ry equil- ibrium) NiCO + CO = Ni(CO) (measurable) and Ni(CO),+ 2CO= Ni(CO) (immeasurably fast). It is shown how from the velocity of the reaction the vapour pressure of nickel can be calculated and also its probable boiling point 3358O A.T.The true order of a reaction for example in the case of nickel carbonyl formation can be determined by consideration of the absolute reaction velocity and its temperature coefficient when the mean molecular diameter is assumed. The velocity of surface reactions is fully discussed and in this connexion the decomposition of ammonia phosphine arsine and antimony hydride are considered. These reactions are assumed to be of the firstl order corresponding with XH -+- XH+H (measurable). The heat of activation in this series decreases with increasing molecular weight. From the heat of activation the maximum wave-length of the actinic rays which will bring about the decomposition is calculated from the formula qo= (2.843 x l V ) / h .It is concluded that ammonia phosphine and arsiii'e should show absorption in the ultra-violet antimony hydride in the visible spectrum. For ammonia the maximum wavelength is shown to be 219*6,up and in confirmation of this it is shown experiment'ally that light of 214-203 pp is photochemically active towards ammonia. Other reactions discussed are the f omation and decomposition of sulphuryl chloride the decomposition of sulphur trioxide and hydrogen selenide the formation of hydrogen sulphide and hydrogen selenide from their elements and the slow combustion of hydrogen iodide of phosphorus sulphur and carbon. From RhearlGENERAL AND PHYSICAL CHEMISTRY. ii.57 and Wheelers experiments on the combustion of carbon t o carbon monoxide (T. 1912 101 846) assuming that the vapour of carbon is monatomic the boiling point of carbon is calculated to be 8542O and its vapmr pressure in mm. of mercury a t llOcio 4-05 x the corresponding concentration being 4.74 x mols. per litre. E. H. R. Consecutive Reactions. IV Relationships of the Constants in the Acid Hydrolysis of Esters of Oxalic and Malonic Acids. ANTON SKRABAL and (Frl.) DANICA MRAZEK (Nonatsh. 1918 39 495-503. Compare A. 1917 ii 250)- The kinetics of the acid hydrolysis of methyl oxalate ethyl oxalate methyl malonate and ethyl malonate have been studied a t 25’. The hydrolysis kas effected by 0.W-hydrochloric acid and since as has been previously shown the determination of the change in acidity did not yield a satisfactory constant the amount of normal ester remaining unchanged was determined iodometrically .In this way it is shown that the hydrolysis takes place according to the equations for first-order reactions. The ratio of the reaction constants of the two consecutive reactions is 2 1. Methyl oxalate is hydrolysed approximately twice as rapidly as ethyl oxalate but in the case of the malonic esters the rate of hydrolysis is approxim- ately the same in the two cases. J. F. S. The Conception of the Chemical Element as Enlarged by the Study of Radioactive Change. FREDERICK SODDY (T. 1919 115 1-26). Atomic Weights in 1917. E. MOLES ,(Anal. Pis. Quinz. 1918 16 625-653).-A review of the work on the determination of atomic weights published during 1917.Complex Ions. 1. N. KOLTHOFF (Chenz W-eekblud 1918 15 1636-1644).-The definitions of complex ions given by de Haas (ibid. 1917 14 752; 1918 15 1352) and Abegg and Bodltinder (A. 1899 ii 542) are criticised. The latter state “One of the ion-forming components of a complex compound is built up of a single ion with an electrically neutral molecule.” de Haas in his later paper defines an inorganic complex ion as “ a metal-contain- ing ion built up of one or more molecules and one or more ions.” The author does not regard the presence of a neutral molecule as essential. Complex ions arise from the tendency of ions to can- bine with other molecules or ions which tendency is the greater the lower is the “ electro-affinity ” of the ion; the nature of the complex is dependent also on the properties of the molecules or other ions which enter into it.It follows that for any salt form- ation of complex ions becomes important only when neutral mole- cules are present to enter into combination with the ions that is when the degree of dissociation is small. Thus salts of strongly electropositive and electronegative ions or groups and salts of oneii 68 ABSTRACTS OE CHEMTCAL PAPERS. stroug and one weak radicle being highly dissociated in solution show small tendency to complex formation; salts oi one strong and one weak radicle or element may however form complex ions by combination of the ions of the weaker element or radicle with each other. The mercuric salts afford iiurrierous examples of the formation of complex ions; dissociation of mercuric chloride forms the ions l-€gCl* Eg” and Cll and combination of the C1’ with the undis- sociated molecule gives rise to the complexes HgC1,’ and Hgc1,I’.Of these the groupc €IgCl’ has the highest concentration namely about 10,000 times as great as the concentration of the Hg” ion. A consideration of the possibilities of combination or dissociation of a group AB’ where A is a bivalent positive and B a univalent negative ionogeri will show that the neutral molecule A4W can be regarded as an intermediate pro[iuct- between the complexes AB’ and RB31. In the first the electro-affinity of both ions is great and the tendency to complex f ormatioil small increasing as the electro-affinity falls. I n the case of the chlorides of the alkali metals in solution in presence of an alkaline earth chloride for example the lower dis- sociation of the alkaline earth chloride causes the presence of its neutral molecules which form complexes with the chlorine ion of the alkali metal chloride of the type RCl,’; in concentrated solu- tions the product of the concentration of this complex and that of the ion of the alkali metal may be so great as t o cause separa- tion of double salts.I n the second class the electrocaffinity of both ions is small; itq is clear that there will be a gradual transition between the two classes. Silver iodide and merctiric cyanide form examples of slightly soluble salts in this class; here the solubility is great.ly increased by presence of eithen ion owing t o the possi- bility of complex formation which this introduces.Cadmium iodide affords a good case of a soluble salt of this class. Consideration is given to complex ions containing hydrogen. Fo’r the case of a weak &basic acid H,A two dissociation con- stants corresponding with the ions HA’ and A” have to be con- sidered the second being usually very small. It is shown that the low dissociation of HA’ as compared for example with RA‘ where R is an alkali metal is in contrast with the case of the halogen acids and the alkali haloids the former having here the higher dissociation. The following definition is finally put forward “ Complex ions break up t o a certain extent into Fimple ions generally with form- ation of neutral molecules.” s. I. L. The cases of iormation of complexes fall into two classes.The Valency-hypothesis of J. Stark. W. JACOBS (Cliem. TVeekbZad 1918 15 1566-1571) .-The theory of partial valencies has been employed to explain molecular aggregation in the fluid and crystalline states and to account for the various states of aggregation in which elements and compounds exist at ordinaryGENERAL AND PHYSICAL CHEMISTRY. ii. 59 temperatures and &he changes in the states of aggregation whch they undergo with change of temperature. l'he lines of jiorce of a valency-electron can be distributed partly to atoms of the same molecule and part'ly to other molecules ; consequently intramole- cular linking and inhermolecular iinkings are essenually due to the same forces and all substances fall wiGhin a series in which the relative strength of the two classes of linhng gradually changes.If the valency-electron is adjacent to a positive surface within its own molecule intermolecular attraction will be weak and will be easily overcome by a small inflow of energy from without which will leave the intramolecular system unchanged; at the other extreme the strength uf the two kinds of linkings approaches equality the molecule behaves as it were like a single atom and intermolecular aggregation is pronounced. I n the last case any disturbance of the intermolecuiar linkings must affect strongly the internal arrangement of the single molecules and it may happen that a particular atomic system can only exist within very iiarrow limits of physical properties. Stark has directed attention to the distinction between the mean- ing of the term molecule in the chemical and in the physical sense; in the former molecules are distinguished as having identical arrangements of atoms in the latter as being particles free to move in space in a definite manner determined by thermal condi- tions.The two definitions agree f o r ni,lecules in the gaseous cow dition since here no intermolecular linkings exist. I n the crystal- line condition a chemical molecule is not free t o move independ- ently of its neighbours and the physical definition fails. I n the liquid condition molecular aggregation occurs but generally varies continually in strength and character ; physical molecules exist but are generally not identical with chemical molecules being usually built up of several chemical molecules and varying con- tinually in mass.I n the case of solution the molecules of solute by the attraction of their partial valencies form ' I solution mole- cules" with molecules of the solvent each of which is a free- moving physical molecule and changes readily with slight changes of temperature. Where intramolecular attraction is weak and the electric valency- field is diffused outside the molecule as in the case of metals and electrolytes the power of crystallisation is high and in the liquid condition the substance has low vapour pressure; where the intra- molecular linkings are very st-rong as with the inert gases the substance cannot be liquefied or solidified at ordinary temperatures. The gradation of physical properties of the hydrocarbons similarly is explained by the gradual decrease of strength of the intra- molecular bonds as the number of Fimilarly linked atoms within the molecule increases. s. I. L. A Device for Introducing a Vapour into a Gas. E. H. ZEITFUCHS (J. Amer. Chenz. Soc. 1918 40 1899).-To introduce a known quantity of vapour into a stream sf gas the followingii. 60 ABSTRACTS OF CHEMICAL PAPERS. device is used. On to the end of a burette is sealed a glass capillary. This capillary is enclosed by t*he constricted end of a wide tube closed a t the top just below the cock of the burette by a piece of rubber tubing which keeps it in position. The capillary touches in the wide tube a roll of asbestos fibres wound round with resistance wire by which it. is heated electrically. A measured quantity of liquid can be r u n from the burette on to the asbestos fibre where it is vaporised and takeri up by a stream of gas enter- ing the constricted portion of the tubs by a side-tube and passing down through the asbestos roll. Any know771 quantit,y of vapour can thus be iutxoduced into the gas. E. H. R. Stopcock for Dropping Liquids arranged €or Equalising the Pressure Above and Below the Outlet in the Stopcock. HARRY L. FISHER ( J . I d . Eng. Ghenz. 1918 10 1014-1015).- An annular groove is provided in the key of the stopcock and a central tlube is fitted in the two aims. The groove is in connexion with the outer tube so that when the tap is turned off there is still communication between the atmospheres in the vessels above and below the stopcock. The liquid flows from the upper to the lower vessel through the central tube and the bore of the stop- cock. w. P. s.