ii. 85 General and Physical Chemistry. N e w Observations R e l a t i v e t o the Zeeman Effect in the H y d r o g e n S p e c t r u m . F. CROZE (Compt. rend. 1912 155 lW7-1610. Compare A 1912 ii 613).-The author has extended his study of the Zeeman effect in the case of hydrogen to the 8-line and has made new measurements for this line and the a- /3- and yjines of the value of AX in a field of 24,122 gause. He maintains his previous view that these three rays furnish a pure triplet with normal deviation and replies to Paschen and Back (Ann. Physik. 19 12 [iv] 39 897) putting forward an explanation as to why their results differ from his. W. G. The Dispersion of C e r t a i n Metals in the Visible Spectrum. CONSTANTIN ZAKRZEWSKI (Bull. Acad Sci. C~acow 19 12 842-849).- A formula is deduced by means of which the optical constants of a metal can be calculated from the phase differences of the components parallel and perpendicular to the plane of incidence measured for two different angles of incidence This method of obtaining the constants does not like the usual method involve the preparation of highly polished metallic mirrors.The formula is applied t o dtlta obtained by the author for nickel and zinc and to Drude’s measurements for silver. €1. 11. D. E l e c t r i c Behaviour of C e r t a i n Vapours which E x h i b i t Absorption Bands. F. BURGER and JOHANNES KOENXGSBEHGER (Physikal. Zeitsch. 1913 13 1198-1199).-The nature of the process which gives rise t o banded emission and absorption is further discussed (compare A. 1912 ii 405).Bromine and nitrogen peroxide exhibit banded spectra and experiments with these substances are described which show t h a t the carriers responsible for the banded spectra in these two cases cannot be due either to electrons or free ions so far as the visible spectrum is concerned. An upper limiting value for the proportion of ionised molecules is calculated which also indicates that the bnnded absorption cannot be attributed to the occurrence of ionised molecules. The Absorption C u r v e of Colloidal S i l v e r Solutions. R. GANS (Phy~iknl. Zeitsch. 1912 13 1185-1186).-0n the assumption that the colloidal particles have the form of ellipsoids o€ rotation the author has calculated the form of the absorption curve for the region X = 4200 to X = 7500 for varying values of t h e axial ratios.As in the case of gold solutions (A. 1912 ii 508) the curve thus obtained may be utilised in the investigation of the geometrical form of the particles present in colloidal solutions of silver. H. A l . D. H. 111. D. The Influence of Acid Radicles on t h e Colour Intensity of Copper Salts. CHARLES SCOTT GARRETT (Zeitsch. Eleklrochenz. 1913 19 1-7).-The molecular extinction of a number of soluble VOL. civ. ii 7ii. 86 ABSTRACTS OF CHEMICAL PAPERS. copper salts was determined by means of a Hiifner spectrophotometer. The salts CuCI2 CuSr2 CU(KO~)~ Cu(C10,)2 CuCr,07 CnS04 and Cu(ClO,) were examined at concentrations from In- 0.01n. The experimental results are discusbed and interpreted on the basis of Baly's theory of the action of a solvent on a dissolved substance.It is shown (1) that when a concentrated solution is diluted a penetra- tion of the salt molecule system by the solvent molecule system occurs ; (2) the mean number of solvent molecules the fields of force of which penetrate the field of force of the dissolved molecule conditions the process of solution at a given concentration; (3) every vibration of the copper atom in the undibsociated molecnle which is active in the absorption of light is influenced by the free electric field of the typical atoms of t h e acid radicle; (4) the light absorption due to the copper is not specifically influenced by the presence of a coloured acid radicle a s is frequently the case with organic substances and (5) there is a strengthening of the light absorption when the free electric field of the acid radicle is negative.J. F. S. Absorption Spectra of Solutions as Affected by Tempera- ture and by Dilution. Quantitative Study of Absorption Spectra by meam of the Radiomicrometer. HARRY C. JONES and J. SAM GUY (Amer. Chenz. J. 1913 49 1-46).-1n a n earlier paper (A 1912 ii 70) a n account was given of a study of the effect of heat and of dilution on the absorption spectra of certain neodymium salts. The work on these salts has now been continued and the investigation has been extended t o Frafeodymium nitrate and chloride and to uranyl nitrate and sulphate. The radiomicrometer employed was a slightly modified form of that described previously. E. G. Quantitative Investigation of the Absorption of Ultra- viqlet Rays by the Fatty Acids and their Isomeric Esters.JEAN BIELECKI and VICTOR HENRI (Compt. rend. 1912 155 1617-1620. Compare A . 1912 ii 882).-The authors have made a more detailed comparison of the absorptive power in the case of three fatty acids and the esters isomeric with them and in the case of a number of groups of isomeric esters the absorption being measured in aqueous and alcoholic solutions and with rays of varying wave- lengths. Curves and tables are given and the results show that (1) the absorption spectrum of the various acids is different from that of the esters isomeric with them this difference being independent of the solvent; (2) the absorption of the ultraviolet rays is controlled by the molecular complexity and augments with increasing complexity.There is also a considerable variation in the absorption spectra of the isomeric esters dependent on the molecular arrangement as is shown by a study of the four esters butyl acetate prop91 propionate ethyl n-butyrate and methyl valerate. W. G. The Absorption Phenomena of Oxyhmnoglobin in the Grating Spectrum. OTTO SCHUMM (Zeitsch. pl~ycliol. C.henz. 191 3 83 1-24).-The spectrum obtained by a diffraction grating isGENERAL AND PEY SICAL CHEMISTRY. ii. 87 spoken of as the natural spectrum in contradistinction to that obtained by 5t prism. Careful measurements of tbe three bands (two in the visible one in the ultra-violet region) are given and the results figured in plates. Some variations occur even in the blood of the same species and it is not possible to distinguish the oxyhzmo- globin of different animals by any such spectrum differences.W. D. H. Long- and Short-waved Absorption and Fluorescence Bands of the Carbonyl Group. 39. GELBKE (Jcdwb. Radioaktiv. A’Zektvonik 1913 10 1-.34).-The ultra-violet absorption of a number of ketones has been examined the substances being dissolved for the most part in ethyl alcohol although in a few cases the substances were investigated i n the form of saturated vapour or dis- solved in water or n-hexane. From a comparison of the short- waved absorption bands of acetone and its substituted derivatives it is found that the acetone band which has its maximum a t A = 868pp is displaced in the direction of greater wave-lengths when one or more hydrogen atoms are substituted by alkyl halogen phenyl nitroso- or carbonyl groups.In the case of alkyl substitution the displacement of the maximum amounts t o 5-10pp whilst the effect of halogens the phenyl nitroso- and carbonyl groups is greater. The displacement of the band is accompanied by an increase in the absorption and also by R broadening of the band. From a comparison of acetone and cyclohexanone i t appears that the influence of r i n g formation is comparatively small. I n addition to the above-mentioned short-waved band acetono shows a band with a maximum a t about X=365pp. which extends from about A=327pp into the visible spectrum. A similar band is also found in the substituted derivatives of acetone and t h i s i t is which gives rise to the yellow colour exhibited by many such substances which contain the carbonyl group. From a comparison of the absorption spectra of dimethyl diketone and methyl ethyl diketone with that of acetone i t is found that the bhort- waved band in the a-diketones is shifted about 5-1Opp in the direction of longer wave-lengths and that the absorption is considerably inten- sified.I n regard to the longwaved band the shift exhibited by the two a-diketones amounts to about 30pp and the intensity is increased about forty times. In tho case of P-diketones the intensification of the long-waved absorption band is not nearly so great as that observed for the a-diketones. From a comparison of the spectra of methplheptenone mesityl oxide and phorone with that of acetone i t is fouud that both the short and long-waved bands are not appreciably shifted by a con- jugated ethylene linking although the intensity of the absorption ~ncreases very considerably in both cases.If the ethylene linking is not conjugated but separated from the carbonyl group by two methylene groups the absorption intensity of the short-waved band appears to be increased to a much smaller extent that when conjugation occiirs. The results of the spectrographic examination of the fluorEscence bands 7-2ii. 88 ABSTRACTS OF CHEMICAL PAPERS. of acetone dimethyl diketone and oxalosuccinonitrile have been pub- lished previously (A. 1912 ii 713) but a large number of other hub- stances have now been examined in regard to the emission of fluorescent light when subjectcd to the action of intense ultra-violet radiation These observations Lhow that the activity of the carbonyl group in this direction is quite general fluorescence being exhibited by both mono- and di-ketones and by both cyclic and acyclic compounds in which i t occurs.It is thus proved that the faculty of fluorescence is by no means confined to ring compounds. 8. M. D. Rotation Dispersion. ERNST DEUSSEN (J. pr. Chem. 1913 [ii] 87 96).-A correction. In reply to Tschugaev (this vol. ii 3) the author admits that the rotations for mercury light recorded in previous papers (A. 1912 ii 510 1020) refer to light of wavc-length 546pp and not 491 pp F. B. The Rotation Dispersion of Some Goloured Lactates. H. VOLK (Btr. 19 12,45 3744-3748).-Abnormal rotation dispersion of coloured asymmetric organic compounds which has hitherto only been studied in the case of somewhat complicated substances (Gross- mann A 1900 ii 372 and Tscbugaev A 1909 ii 631) has now been observed with the copper nickel and cobalt salts of I-lactic acid.A maximum rotation is found in each case in the region of selective absorption. Increase of temperature diminishes the rotation but does not affect the character of the dispersion. For the preparation of these salts zinc ammonium I-lactate (Purdie T. 1893 63 1143) which has a normal but comparatively high rota- tion-dispersion is boiled with baryta water t o expel ammonia and precipitate the zinc and after removing the excess of barium hydroxide by means of carbon dioxide the filtrate is treated with the equivalent amount of the metallic sulphate and allosed to crystallise.Copper 1-lactate Cu(C,H,O,) 1 H,O a light blue crystalline powder has a maximum rotation and absorption in yellow light ; [MI calculated for lactic acid falls from 18.5 in the green to 15.6 in the dark blue. A'ickel 1-lactate Ni(C,H,O,),QH,O a light green powder is remai kable in that it is Zaevo-rotatory in all parts of the spectrum; the maximum rotation [MI = - 0-S and strongest absorption are in t h e yellow. CobuIt IJactate Co(C3H,03),,2H,0 a light red powder has maximum absorption for green light in which [M]=30; the rotation falls towards either end of the spectrum and changes sign between light blue and dark blue. J. C. W. Sunlight. PAUL C. FREER and HARRY D. GIBBS (J. Physical Chem. 1912 16 709-738).-The influence of sunlight on chemical and physiological processes is supposed by some to be greater in the tropics than in temperate climates.On the assumption that the difference lies in the intensity of the ultra-violet radiation an investigation of the photochemical action of tropical sunlight should yield some information as to its probable physiological action. The effect of sunlight in Manila in promoting the coloration ofaENERAL AND PHYSICAL CHEMISTRY. ii. 89 a number of benzene derivatives such as cresol and aniline has been described from time to time by Gibbs. Since these reactions have large temperature-coefficients and the temperatures of solutions exposed to the sun in Manila may attain 50° Gibbs’ experiments gave no certain indication t h a t sunlight is abnormally active in the tropics. Measurements of the total insolation in heat units made with the Angstrom pyroheliometer a t various latitudes indicate trhat the heat intensity varies but little from place to place.The intensity of violet and ultra-violet radiation is conveniently measured by the rate of decomposition of oxnlic acid in presence of uranyl acetate. This reaction has a very smnll temperature-coefficient and when used under standard conditions gives useful comparative results. I n Manila the average amount of oxalic acid decomposed in one hour during one year was 12*45% the maximum being 17.8% and the minimum 1.15%. The average values obtained by other observers were as follows Bltguio (Philippine Is.) 14.2%; Honolulu 13.9%; *Kuala Lumpur (Malay States) 15.29% ; Khartoum 17.8% (using quartz flask) ; Washington 11.96% ; Tucson (Arizona) 13.0% ; Munich 8.39%. When the sky is clear the activity is almost the same everywhere and the lower averages in the higher latitudes may be ascribed to the higher percentago of dull days.Oxyhemoglobin is converted by ultra-violet light into methsmo- globin which can be detected in the blood of rabbits which have died by exposure to the sun. Nevertheless experimonts in which rabbits monkeys and men of different shade were exposed to the sun seem to lead to the conclusion that the ill-effects are wholly due to the rise in blood-temperature and that the resistance of an individual depends largely on his opportunity for cooling by radiation etc. and the per- fection of his heat-regulating system #of sweat glands; thus black.grey and white rabbits monkeys and men showed an increasing resistivity. Monkeys exposed to the sun in Manila died in less than two hours the blood temperature rising almost loo but when the animals were kept cool by a fan although exposed to tho full radiation including the ultra-violet no ill-effects followed. The brain in these experiments was to sorne extent protected by the hair on the head. R. J. C. Photochemical Reduction of Copper Sulphate. CHARLES W. BENNETT (J. Physical Chem. 1912 16 782-785).-The author propounds the theory that the light which is absorbed by cupric sulphate solution tends t o decompose it so that a reducing agent might be found which although without action in the dark would precipitate copper when assisted by radiations of appropriate wave- length.Xxperiments are cited to show that light reduces ammoniacal copper sulphate t o cuprous oxide in presence of hydrazine hydrate and that the precipitation of copper phosphide from 5% cupric sulphate solution by an ethereal solution of phosphorus (Mrs. Fulhame 1794) is accelerated by the light from a carbon arc I n the latter case a quartz vessel was used. R. J. C .ii. 90 ABSTRACTS OF CHEMICAL PAPERS The First Stages in the Photolysis of Ethyl Alcohol Acetaldehyde and Acetic Acid. DANKEL BERTHELOT and HENRY GAUDECHON (Compt. rend. 1913 156 68-’71).-Of the three sub- stances studied. acetaldehyde is the only one which is act,ed on by sunlight (A > 0*3p) the alcohol and acid requiring radiations of shorter wave-length.There are four stages in the photolysis of acetaldehyde namely (1) gaseous decomposition into carbon monoxide and methane (compare A.? 1910 i 543). (2) Polymerisation and sub- sequent evoiution of hydrocarbons such as ethane etc. (compare A 1910 ii 814). (3) Formation of resinous products. (4) Conversion of the aldehyde into acetic acid. Only the first reaction is produced by initial ultra-violet rays the three others requiring the mean or extreme rays. The photolysis of ethyl alcohol consists fundamentally of the evolution of hydrogen and formation of acetaldehyde which is then further decomposed (compare A. 1911 ii 835). Acetic acid on decomposition by the rays from an Heraeus lamp gives a gaseous product containing carbon dioxide (44%) carbon monoxide (1 i%) and combustible gases (39%) (compare A.1910 ii 814). Water has no effect on the products of photolysis but lessens the rate of the reaction. W. G. Photolysis of Different Complex Sugars (Bioses and Trioses) by Ultra-violet Rays. DANIEL BERTHELOT and HENRY GAUDECHON (Compt. rend. 1912 155 1506-1509. Compare A 1912 ii 1120). - 9 study of the decompobition of the biosea maltose lactose trehalose and gentiobiose and the trioses raffinose melezitose and gentianose by ultra-violet rays of varying wave-length. I n the case of the trioses no first stage of formation of a monose and a biose could be detected. After the first stage in the decomposition the action was the same a s in the case of the monoses (Zoc. cit.). W. G. The Radio-elements the Periodic System and the Constitu- tion of the Atom.a. VAN DEN BROCK (Physikaz. Zeitsch. 1913 14 32-41).-This paper deals with recent work in the radioactive disintegration series and in the rare earth group with the intra- atomic charge and theories of the constitution of the atom and their connexion with the periodic lam in a manner unsuitable for abstraction. F. s. Evidence of Spontaneous Alterations of Concentration in Solutions and Gases. THEODOR SVEDBERG (Physikul. Zeitsch 1913 14 22-26. Compare A 1912. ii 905 906).-The question is discussed as t o whether the relative deviation of the number of a-particles produced in a volume element of a solution of a radio-active substance from the mean value should be theoretically greater than the corre- sponding relative deviation from the mean value in the case of the solid substance. It has been shown previously that the ratio of the mean deviations should be J2 1 and experimental observations indicate t h a t this is approximately the case.Further arguments areGENERAL AND PHYSICAL CHEMISTRY. ii. 91 now put forward in support of the correctness of the above theoretical conclusion. The question of the detection of spontaneous alterations in the concentration of solutions of other than radioactive substances is also examined with reference t o the negative results which have been obtained lip to the present. The author draws the conclusion t h a t such changes cannot be detected by the methods of examination which are a t present available. H. M. D. E x c i t a t i o n of y-Rays by a-Rays.JAMES CEADWICK (Phil. Mag. 1913 [vi] 25 193-197. Compare A. 1912 ii 1025).-Further eviderice is given that the impact of a-rays on metals of high atomic weight excites y-rays. A tube of such thin glass t h a t the a-rays escape containing radium emanation was surrounded with two con- centric tubes of metals of widely different atomic weights of thickness just sufficient to stop all the a-rays. Tho position of the two metals was transposed and the ionisation with the two arrangements accurately compared by a balance method. Always when the metal of high atomic weight was next the emanation tube and bombarded by a-rays the ionisation current was slightly greater than when the metal bombarded was of low atomic weight. The effect was much increased by filling the ionisation chamber with vapour such as carbon disulphide or methyl iodide.The ratio of the ionisation in these vapoura t o t h a t in a i r by X-rays of low penetrating power is much greater than for X-rays of high penetrating power and it was to be expected that the y-rays generated by a-rays would have a milch lower penetrating power than the primary y-rays and that the part of the effect due to them would be greater in a heavy vapour than in air. If a thickness of light material was placed round the tube sufficient to absorb the a-rays the effect disappeared. That i t was not due to very soft P-rags was proved by experimenting in a powerful magnetic field sutlicient to coil up the P-rdys and prevent them from reaching the metal. This did not cause any change in the effect With outer tube of aluminium the inner tube being of the followiog materials gold silver copper aluminium paper ; the relative ionisations in carbon disulphide vapour were respectively 104.8 102.5 301.2 100 and 99.7. This shows t h a t the excitrition of y-rays by a-rays increases with tho atomic weight of the material bombarded.A rough estimate of the coefficient of absorption of the excited y-rays was 8 (cm.-l of aluminium). I n con6rmation i t has been shown with A. S. Russell that ionium and polonium which give a-rays without P-rays give-y-rays in amount sutticient t o be studied in detail. F. s. Retardation of a-Particles by Metals. E. MARSDEN and H. RICHARDSON ( P I d . Mug. 1913 [vi] 25 184-193. Compare Taylor A. 1908 ii 783 ; 1909 ii 850)-The “air-equivalent ” of a metal foil is the distance in cm.by which the range OF an a-particle at 76 cm. and 15’ is reduced in passage through the foil and is for materials of atomic weight greater than that of air the less the less the velocity of the a-particle or the nearer it is to the end of its range. This variation has been studied for the a-rays OF radium-Cii. 92 ABSTRACTS OF CEEMICAL PAPERS by a scintillation method which offers advantages only the fastest a-particles that is those which have not been scattered being observed. A zinc sulphide screen is rigidly attached in the focus of the microscope and is carried by the rack axid pinion to any required distance from the source of a-rays. Between on a separate rack and pinion the foil is mounted and the air-equivalent of the foil studied a t various parts of the range of the a-particles. Tables are given for the variation of the air-equivalent which is considerable especially near the end of the path with the range and also a curve from which the air-equivalent of any thickness of foil (Au Ag and Al) a t any part of the range can be deduced.A brightening of the screen just outside the extreme range of the a-particles themselves which is unaffected by a transverse magnetic field was observed and is ascribed possibly to y-rays excited by a-rays (see preceding abstract) F. S. Absorption of /3-Rays. J. A. GRAY (Proc. Roy. Soc. 1912 A 87 487-489).-Two experiments on the absorption of the P-rays of radium-E by paper show that the rays become more and more absorb- able practically none passing 56 sheets weighing 8.77 mg.per cm.2 the percentage transmitted by 5 sheets decreasing from 50 to 10. These results are explained on the view that P-rays are both retarded and stopped by mattor and when the velocity of the surviving rays falls below the previous mean velocity the exponential law can no longer hold. F. S Similarity in Nature of X - and Primary 7-Rays. J. A. GHAY (Proc. Roy. Xoc. 1912 A 87 489-501).-The absorption of the primary y-rays of radium-E and of the y-rays excited in lead by the /$rays of radium-E is compared with the absorption of X-rays. The y-rays of radium-E are absorbed somewhat similarly to characteristic secondary X-rays excited in silver. There is no discontinuity with respect to absorption between X- and y-rays.Tin absorbs the y-rays of radium-E to an abnormal extent but for iron and gold compared with aluminium the absorption is similar to that of the silver An investigation of the radiation reflected from radiators of various materials showed that the y-rays of radium-E excite the characteristic X-radiation in silver tin barium neodymium and praseodymium This furnishes the most definite proof possible that X-rays and y-rays are of the same nature. Litstly the scattering of y-rays of radium-E and of X-rays has been compzred and it is shown that the phenomena is similar both in magnitude and character for both radiations. Expansion Apparatus for Making Visible the Paths of Ionising Particles in Gases and the Results Obtained with this Apparatus.CHARLES T. K. WILSON (Jcchrb. Radioaktiv. h'lektronik 1913 10 34-35. Compare A 1911 ii 565).-The expansion apparatus and method of illuminating the condensed water droplets described i n the previous paper have been improved in several X-rWp. F. S.GENERAL AND PHYSICAL CHEMISTRY. ii. 93 respects with the result that i t has been possible to obtain photo- graphs of the most rapidly moving P-rays which show the droplet8 due t o the individual ions. The greatly increased size of the cloud cbamber in the new apparatus not only enables the entire course of the longest-ranged a-particles to be registered but its depth is such that horizontal beams of X-rays may be examined without any danger of interference from the proximity of the roof or base of the chamber.The photographic records obtained iinder the new conditions are more- over much more distinct than those given by the earlier form of apparatus. I n regard to the a-particles the photographs show that these move along paths which are approximately rectilinear except for the last 2 mm. where there is a very marked change in direction. Apart from this sudden and pronounced change at the end of the range there is however in many cases a small and gradual change in direction over the greater part of the course and the photographs thus afford evidence of both single and compound scattering. The photographs of the P-rays indicate that the ions are formed in pairs over a considerable portion of the path but that a t certain pointe closely packed groups of twenty or thirty ions are produced.The occurrence of these groups of ions is regarded as evidence of the emission by the atoms of secondary corpuscles which possess sufficient energy to produce ionisation over a small range. I n opposition to the a-rays the @rays never exhibit sudden changes in direction although the slow change due to compound scattering is clearly shown pro- vided the velocity of the rays is not too large. The X-ray photographs show t h t t P-rays proceed from points in the paths of the X-rays beam. Apart from this production of /3-rays t h e X-rays appear to give rise t o no effect which is photographically regis- tered on the plates. Corresponding with the heterogeneity of the X-rays it is found that the paths of the individual P-particles vary considerably in length.The rays appear t o show both simple and compound scattering although the latter effect appears to be the much more frequent and characteristic type. The emission of the ,%rays appears to be uniformly distributed in all directions and no evidence of any dependence on the direction in which the X-ray beam is propagated has been obtained. I n the case of both the p-ray and X-ray photographs it has been possible to ascertain approximately by counting the droplets the number of ions which are produced per centimetre of path. H. M D. Ionisation Produced by p- and 7-Rays at High Pressures. D. C. H. FLORANCE (PhiE. Bug. 1913 [vi] 25 172-183).-The absorption and ionisation produced in air at high pressure with the P-rays of uranium-X and with the secondary /3-rays generated by the y-rays of radium-C have been studied by means of a special cylindrical ionisation vessel with pressures of air up to 80 atmospheres.Corn- plete saturation could not be obtained even with a field of 2800 volts per cm. the shape of the curves being similar to those obtained with a-rays. There is for constant voltage a constant ratio for the ioniss-ii. 94 ABSTRACTS OF CHEMICAL PAPERS. tion currents at any two pressures and although saturation is not attained the ionisation currents are proportional to the intensity of the radiation at different pressures. The experimental value found for p(cm.-1 of air a t atmospheric pressure) for the P-rays of uranium-X was for plates 7 cm. diameter and 1 cm. apart 0.04 but the true value for a parallel beam of P-rays was deduced t o be 0.007 in agree- ment with the value of p in solids.The values for the emergent and incident P-rays produced by y-rays of radium-C were 0.046 and 0.056 and they appear to be independent of the nature of the material in which they originate. With plates 1 cm. apart the part of the ionisation due to /3-rays originating in the gas is negligible st atmospheric pressure and 25% of the total ionisation at 80 atmospheres. F. S. Recombination of Ions Produced by Rontgen Rays S. J. PLIMPTON (Phil. Mug. 1913 [vi] 25 65-81 ; Anzer. J. Sci. 1913 [iv] 35 39-53).-The value of a the coefficient of recombination of the ions in the equation drL/dt = -. an2 mas determined by ionising the gas between two parallel plate electrodes by a single flash of X-rays restricted so as not to fall upon the plates allowing the ions produced to recombine in absence of any external field for small intervals of time regulated and determined mechanically and then by sudden application of a field driving the residual ions t o the electrodes and so evaluating their numbers. I n this way the effect of ionic diffusion can be greatly reduced corpuscular radiation from the metal plates eliminated uniformity of X-rays secured and the initial stages of recom- bination studied.The results revealed especially for carbon dioxide an initial value of a relatively large but decreasing and becoming constant after intervals usually of about one-third of a second. This is explained by initial recombination of the pair of ions formed before they have moved by diffusion out of their range of mutual influence and the high value in carbon dioxide and the slower rate at which the normal value is attained are ascribed to the slower diffusion in this gas than in air.If a field acts during the flash ceasing to act simul- taneously with the cessation of the flash the final value of a is not affected but the initial value is decreased. A table is given of the absolute values of a at various time-intervals from the cessation of ionisation for air carbon dioxide sulphur dioxide methyl iodide ethyl chloride ethyl bromide and ethyl ether at various pressures. For air the initial value of a/e when e is the atomic charge is 11,540 and the final value 3960. F. S. Further Experiments on the Mobility of the Positive Ions at Low Pressures.GEORGE W. TODD (Phil. Mag. 1913 [vi] 25 163-171. Compare A. 1912 ii 245 lOfiO).-The ex- periments have been repeated with a larger apparatus in order to obtain information at lower pressures at which it was formerly suggested sub-molecular particles might constitute the positive ions. The new experiments show nothing like the enormous increases in the mobilities previously observed but indicate that below theGENERAL AND PHYSICAL CHEMISTKY. ii. 95 pressure of 1 mm. the mobilities vary with the period of alternation of the field when the period is high (above 150-). F. S. Combination of Hydrogen and Oxygen in the Presence of Heated Platinum and Carbon. J. K. THOMPBON (Physrkccl. Zeitsch. 1913 14 1 l-l5).-Experiments have been made with the object of ascertaining whether there is any connexion between the temperature a t which hydrogen and oxygen begin to combine in contact with heated platinum or carbon and that a t which these begin to emit negatively charged particles.The platinum spiral or carbon filament was contained in a cylindrical glass tube which could either be exhausted or connected with a supply of the gas mixture. The temperature of the spiral could be gradually raised by means of a regulated current from an accumulator battery and a voltmeter placed across the ends of the wire enabled the resistance to be measured. An aluminium cylinder surrounding the heated wire was connected with one pair of quadrants of a Dolozalek electrometer and permitted the measure- ment of the electrical effect a t gradually increasing temperatures.From observations made when the tube was highly exhausted and when hydrogen and oxygen were admitted at different pressures meaaured by the manometer it has been found that in the case of platinum and of carbon the temperature a t which the emission of negative particles is first observed is within the limits of experimental error identical with the temperdture a t which combination occurs between the hydrogen and oxygen. I n all cases the explosive com- bination of the gases was found to be accompanied by large ionisation currents but these could not be measured. The actual temperatures a t which the two effects occur were not measured the temperature being measured in terms of the resistance of the wire as ascertained from the readings of the voltmeter and ammeter.The influence of the gas on the observed resistance was allowed for by means of a series of readings with the apparatus when the heated wire was in contact with air a t different pressures. Preliminary observations shorn also that combination of hydrogen and oxygen occurs if the platinum wire is subjected to the action of B beam ot X-rays. The combination occurs more readily if the percentage of hydrogen is less than 10. Investigation of the Spectrum of Ionium. ALEXANDER S. RUSSELL and R. ROSSI (Proc. Boy. Soc. 1912 A 87 478-484).- The ionium preparation employed (Boltwood A 1911 ii 359) from measurements of the number of a-particles emitted should contain 16% of ionium if the period of that substance is 100,000 years the remainder being thorium oxide.After chemical purification of the ionium-thorium preparation by various methods its arc spectrum between the wave-lengths 3800 to 5000 was photographed in juxta- position with that of a pure thorium oxide preparation in the first ordef by a Rowland’s grating of 21.5 feet radius giving a dispersion of 2.6 Angstrijm units per mm. The spectrum was identical with that of pure thorium save for the presence of five of the stronger lines of H. M. D.ii. 96 ABSTRACTS OF CHEMICAL PAPERS. scandium the amount of which in the purified preparation was esti- mated as three or four parts per thousand. It was found that an admixture in thorium oxide of 1% of cerium oxide could easily and 1% of uranium oxide could just be detected spectroscopically.The obvious conclusion is that ionium cannot be present in the preparation to the extent of more than 1 or 2% and that its period of average life cannot be more than 12,000 years. This result is consistent with the estimate given by Soddy of at least 100,000 years only if at least one unknown product of long life intervenes between uranium and radium. On the other hand the period deduced from the range of the a-rays 200,000 years is not consistent with the result obtained unless the a-rays of ionium are exceptional. It is possible that ionium has no arc spectrum in the region investigated although this is improbable or that ionium and thorium have identical spectra as they have identical chemical properties. In a note the paper on the same subject by Exner and Haschek (Sitxungsber.K. Akad. Wiss Wzen 1912 121 iia 1075) with the same result is referred to the ionium in the Austrian preparation being estimated as two-thirds of that in the one employed The Solid Radioactive Disintegration Products Suspended in the Atmosphere. K. W. FRITZ KOHLRAUSCH (Physikal. Zeitsch. 191 2 13 1193).-Certain corrections are discussed in the methods employed in the measurement of the solid radioactive products in the atmo- sphere and new expressions more in accord with recent work are deduced. For the determination of the amount of radium-A in the air by passing the air a t a suitable velocity for a known time through a tubular condenser with inner electrode negative and measuring the ionisation current a t the end due to the active deposit on the inner electrode tables are given showing the ratio of the effect for any time of aspiration to that for an indefinitely long time.I n the second place the effect of the thorium products in the air is taken into consideration. F. S. Electro-chemistry of Extremely Dilute Solutions in P a r t i - c u l a r of Radioactive Substances. KARL F. HERZFELD (Physikal. Zeitsch. 1913 14 29-32).-1f a metal plate is immersed in a solution of one of its salts which contains a mixture of two radioactive substances in a definite ratio then according to the experiments of Hevesy (A. 1912 ii 414) the ratio in which $he radioactive metals separate on the metal plate is dependent solely on the potential difference a t the contact surface of the metal and its salt solution.This observation is contrary to what would be expected on the basis of Nernst’s formula for the equilibrium a t the surface. The discrepancy may be explained if i t is assumed that the quantity of deposited radioactive metal in the equilibrium condition is not sufficient to form a complete moiscular layer. If P’ denotes the solution pressure of the radioactive metal p the osmotic pressure of the amount initially present in the solution and ‘IT the pressure correspond- ing with the metal deposited in the equilibrium condition then the cathode potential E will be related to these quantities by the equation F. S.GENERAL AND PHYSICAL CHEMISTRY. ii. 97 E’= RTlogP’r/( - T). The application of this equilibrium equation to two radioactive substances present in the solntion leads to results in agreement with the observed facts.I n regard to the variation of the ratio of the deposited metals with the magnitude of the potential at the metal surface the formula gives a curve which is also of the same type as that representing the experimental data. H. M. D. Electrical Excitation in the Splashing of Liquids (Ballo- Electricity). CHRISTIAN CHRISTIAKSEN (Ann. Yhysik 19 1.3 [iv] 40 107-137).-An account is given of experiments in which the electrical phenomena associated with the incidence of liquid spray on a solid surface have been investigated. By means of a current of air under constant pressure the liquid under examination is made to issue from a small aperture in the form of a fine spray which impinges on a plate or on the walls of a tube of the solid substance t o be investigated.I n most of the experiments a platinum disk or tube mas employed and this was connected by a wire with the liquid in the supply reservoir and also with one pair of quadrants of an electrometer the opposite pair being eai thed. The magnitude of the electrical effect obtained when distilled water is sprayed mas found to vary very considerably with the sample of water employed. I n searching for the cause of these irregularities it has been found that contact of the water with any substance such as glass paper wool silk o h etc. which has been exposed to the air for a considerable time gives rise to a greatJy increased electrometer deflection. The influence of electrolytes on the ballo-electric effect is relatively very small and the much greater influence of contact with the above-mentioned substances is supposed to be due to the condensa- tion of radioactive substances on the air-exposed surfaces.The deflections obtained with aqueous solutions of acids bases and salts and the influence of dilution have been examined with the result that the chemical nature of the electrolyte appears to be of little importance in connexion with the phenomenon. According to the magnitude of the electrical effect it is possible to divide aqueous solutions into two groups. When solutions belonging to these diflerent groups are mixed i t is found that the ballo-electric activity of the mixed solution is very much greater than that of the com- ponents. As an example it may be mentioned that if equal volumes of molar solutions of ethyl alcohol and potassium chloride are mixed together the effects obtained with the mixed solution was 7.0 whilst that given by the alcohol solution was 0.9 and by the potassium chloride 0.1.Several such mixtures have been investigated in such a way as to obtain comparative numbers for related substances belonging to the two above-mentioned ballo-electric groups. Other experiments show that the effect varies with the nature of the solid surface on zhich the liquid spray impinges. With a distilled water spray platinum glass and ivory become positively charged whilst paratlin becomes negative. I n the case of wax and shellac the solid is at first negatively charged but after prolonged action the sign of the charge becomes reversed.The most constant results are obtained with a platinum surface. H. M. I).ii. 98 ABSTRACTS OF CHEMICAL PAPERS. Dielectric C o n s t a n t s of Dissolved S a l t s . 11. PAUL WALDEN (Bull. Acccd. Xt. Yitersbowg 1912 [vi] 1055-10SG).-Iuvestigations similar to those previously made with chloroform solutions (compare A. 1912 ii 421) have been carried out with solutions in dichloro- methane (dielectric constant 8.3) and in ethyl formate (dielectric constant 8 9 ) which differ from chloroform in giving measurable ionic dissociation of the salts employed (Zoc. cit.). Further ethyl formate is a solvent of an entirely different type so that its use opens up the possibility of studying the chemical rBle of the atoms constituting the solvent.As with chloroform so also with dichloromethane the dissolution of the salts is accompanied by a considerable increase in the dielectric constant of the solvent. Further the dielectric constants of the salts calculated by means of Philip’s Bouty’s and Silberstein’s formula are mostly very bigh the greatest values being obtained with the first of these fcrmuh. I n dichloromethane the value of the dielectric constant of any salt is much higher than in chloroform This con- stant also increases with the complexity of the salt and according to Philip’s formula has the values NHEt,,HCl 58 ; NEt,,HCl 390 and NEt,Cl 1980 ; the numbers for the corresponding nitrates are 71 425 and 1365 respectively. For one and the same salt a t equal dilutions the molecular con- ductivity is from ten t o thirty times as great in dichloromethane as in chloroform and whilst in the latter solvent it diminishes rapidly and continuously as the dilution increases in the former a minimum is observed at a certain dilution.The conclusion is drawn that the ions formed from the salt are able to raise the dielectric constant of the solution and alco the dissociating power of the solvent. Ebullioscopic measurements show that the salts have different molecular magnitudes in these two solvents those being for N Et,CI M i n chloroform and M in dichloromethane and for NEt,,IICI 21 and M - M respectively. The increased values ot the dielectric constants in dichloromethane may bs due partly to this depolymerisation and partly t o 8 specific catalytic influence of the solvent.The chemical process of salt-formation by the conjunction of two neutral constituents (NEt + EtBr etc.) is expressed physically in a reconstruction of the dielectric properties of these con~titueiit s in such a may that the new product shows a dielectric constnrjt which is about six t h i r t e e ~ or eighteen times t h a t calculated additively for secondary amine tertiary amine and tetralkylammonium salts respectively. Calculation by means of the Clausias-Rlosotti equation x = ( K - 1)/ ( K + 2) where x represents the fraction of the total volume actually occupied by the molecules shows that those salts are the best electrolytes for which x approaches the value 1. Not only is the general behaviour of ethyl formate as regards the alteration of the dielectric constants of dissolved alkylamino-salts similar to t h a t of dicbloromethane but the actual values of these constants are very nearly equal with the two solvents.Sodium iodide and lithium bromide are characterised by extremely high dielectricGENERAL AND PHYSICAL CHEMISTRY. ii. 99 constants in ethyl forniate the values obt,ained being 5595-6820 and 4760-5300 respectively according to Philip's formula ; calcium iodide shows a much smaller value (1540) and mercuric chloride one still snialler (490); thus the dielectric constant of these salts diminishes with their capacity t o dissociate. The results obtained with solutions of salts in the stxongly-ionking solvents acetone acety lacetone acetonitrile and propionitrile confirm the conclusion that if the dielectric constant of the medium remains constant the formation of ions increases the ionising tendency and the degree of electrolytic dissociation of the salt.Certain anomalies exhibited by solutions are discussed in connexion with the increases produced in the dielectric constants and hence in the ionising power of solvents by the dissolution of electrolytes. It is pointed out that this phenomenon must be taken into account in considering deviations from t h e law of mass action (Ostwald's dilution law) the r61e of neutral salts the disagreement of salts dissolved in media havicg feeble ionising actions with the requirements of the theory of Arrheniuy etc. since by it as well as by the other well- known factors the degree of electrolytic dissociation is appreciably altered.T. H. P. Behaviour of Carbon on Electrical Pulverisation. CARL TKOMAE (Zeitsclh. C I L ~ ~ . I2zd. Kolloicle 19 12 11 268-269).-1f a continuous current of 3-4 amperes a t 220 volts is passed between retort-graphite electrodes immersed in pure water the liquid remains quite transparent but if a very small quantity of sodium hydroxide is added a dark brown colloidal solution is obtained. The solution is stable and may be boiled withput change but the addition of hydrochloric acid or sodium chloride gives rise to a black precipitate. Similar resuIts were obtained with arc-lamp-carbon electrodes. H. $1. D. Theory of the Electrical Synthesis of Colloids. CARL BENEDICKs (zeitsch. Chem. Ind. Kolloide 1912 11 263-268).-A criticism of the views expressed by Kutscherov (A 1912 ii 1148) in connexion with the interpretation of the phenomena of electrical pulverisation.I n opposition to the theory that the process is partly electrochemical in nature the author maintains that the observations can be adequately explained on the assumption t h a t the action is entirely thermal in character. H. 31. D. The Mobility of the Bromide Ion in Water. CABL DRUCKER M. TARLE and L. GOMEZ. (Zeilsch. Elektrochsm. 1913 19 8-1 l).-The transport numbers of tribenzylmethylammonium bromide were deter- mined at 20' by the usual method employing a zinc anode and a platinum cathode. As a mean of many experiments the value n = 0.775 was obtained. To calculate the mobility of the bromine ion from this figure the conductivity values at 18' were used (Drucker A.1912 ii 732). The tempersture-coefficient of the conductivity was determined and found to be 0.022 between l S o and 25' and from these data the mobility of the bromine ion is 65.5. This value is muchii. IOO ABSTRACTS OF CHEMICAL PAPERS. lower than that obtained by Kohlrausch (67*0) and affects the con- ductivity at infinite dilution of potassium bromide making i t 129.8 instead of 132.6 as extrapolated by Kohlrausch. Possible sources of error in the Kohlrausch value are discussed. It is shown on calcu- lating the degree of dissociation of potassium bromide from the new value that it obeys the Ostwald dilution law up to 0.01 ?a. The authors assert that very probably all strong electrolytes obey the dilution law up to this concentration. J.F. S. Investigation of the Chemical Reactions Taking Place at the Cathode and Anode During the Electrolysis of Simple Salt Solutions. J. HAMILTON PATERSON (Proc. Univ. Du~ham Phil. Soc. 1912 4 187-220).-The author describes a small electrolytic cell which allo6a of the microscopic examination of the deposits formed on the electrodes during electrolysis Experiments made with simple salts (chlorides nitratep or sulphates) of a number of metals have shown that it is possible to obtain from cobalt nickel iron zinc cadmium manganese and possibly lead salts either the pure metal or the hydroxide deposited on the cathode. In the case of cobalt the film of hydroxide has a deep blue colour but the moment the current stops i t turns green; the composition of the green com- pound corresponds with the formula Co(OH),,H,O.The conditions of concentration and current density affecting the production of these substances has been noted and it has been found that the dilution necessary to produce the hydroxide increases and the limit of current strength decreases as we proceed along the list of metals in the order given. The hydroxide of the metal is not produced as a secondary reaction of the metal deposited on the cathode but is a precipitate from the electrolyte in the cathode area. F. B. The Constitution of Water and the Thermal Variation of its Magnetisation. AUGUSTE PICCARD (Compt. rend. 1912 155 1497-1499. Compare this vol. ii 17)-From a study of the change in the coefficient of magnetisation of water on solidification at 0" and a comparison of the curves representing respectively the variation i n density and magnetisation coeficient with respect to temperature the author draws the conclusion that in water at the ordinary tempera- ture there a r e two different substances in equilibrium.W. G. New Thermo-regulator for Instantaneous Adjustment to any Given Temperature. R. FANDER (Chem. Zeit. 1913 37 40j.-The regulator bears much resemblance to that of Reichardt but differs in having the adjusting screw on the lower half of the capillary tube the upper half of the tube being graduated from 30°to 1.00'. The regulator is set by immersing it in a water-bath at any tem- perature and adjusting the mercury column so that the readings on t h e capillary tube correspond with those of a thermometer in the bath €or about a minute.The mercury is then raised or lowered to the graduations corresponding with the desired temperature and the regulator is ready for use. H. B. H.GENERAL AND PHYSICAL CHEMISTRY. ii. PO1 The Expansion Coefficient of Graphite. ARTHUR L. DAY and ROBERT B. SOSMAN (J. Ind. Eng. Chern. 1912 4 490-492).- Contains sketches of the apparatuq employed by the authors for measuring the expansion coefficient of graphites whilst the resiilts obtained are demonstrated in tables and curves. From the wide divergence in expansion coefficients obtained the concluqion is drawn contrary t o the views of Le Chnteliur and Wologcline ( A . 1908 ii 177) aud in harmouy with tho.;e of Arsem (A 1912 ii 5 0 ) that there must be some fundamental difference between the vhrious forms of graphite.F. M. G. M. Volume Changes of Amalgams. J. W URSCHMIDT ( B v . Deut. physzkal. Ges. 19 12 14 1065-1087).-The influvnce of temperature on the thermal expansion of a number of amalgams has been examined with the aid of a special type of dilatorneter. The curves obtzined by plotting tho coefficient of expansion as n function of the temperature exhibit very sharp msxima at definite temperatures. I n the caye of the tin lead and cadmium amalgams which according to Puschin (A 1903 ii 212) consist of solid solutions these teriiperntures are identical with the melting points of the alloys. On the other hand the zinc amalgams which consist of mechanical mixtures of the components exhibit the maxiruum expansion a t temperatures bet ween 50' and 70° which are very much lower than the melting points of the alloys.I n regard to the nature of the expansion temperature curve sodium amalgam resembles the corresponding tin lead and cadmium alloys but differs from these in t h a t the volume changes with time when the temperature is kept constant. H. M. D. Thermal Conductivity of the Metals and its Relation to Their Other Properties. STEFANO PAGLIANI (i'vuovo C m . 191 2 [vi] 4 ii 48'2-497).-'l!he author gives a table showing the coetfi- cients of thermal conductivity the specitic heats and the densities of a number of metale the material being derived from published reTults of various workers. By-a comparisou of other published data he shows that the metals can be arranged in a series in which increased con- ductivity and reflecting power are associated with decrease of emis4ve power and that these properties also fall within the regulmty of the periodic law.R. V. S. Theory of Specific Heats. MAX BORN and TH. VON K ~ R M ~ N (Physikul. Zeitsch. 19 13 14 15-19).-Reference is made to a previ\)us payer (ibid. 1912 13 297) in which the discrepancy between the observed specific heats of solid substances at low temperatures and the values calculated from Einstein's formula has been eliminated to a large extent by means of a modified specific heat formula. The t8heory on which this is based is compared with the theory advanced by Debye (A. 1912,ii 1134) and it is shown tnat the approximation formula proposed by Debye for the calculation of the atomic heats of solid non-atomic substances can be deduced from the geueral formula given by the authors.In the case of anisotropic substances the approxima- VOL. CIV. ii. 8ii. 102 ABSTRACTS OF CHEMICAL PAPERS. tion formula is inapplicable and the general formula referred to must. be employed in the calculation of the specific heat,s. H. M. D. Specific Heat of Certain Elements at Low Temperatures. TADEUSZ ESTREICHER and M. STANIEWSKI (Bull. h a d . Sci. Cy*acow 1912 S34-S41. Compare A. 1911 ii 16).-The authors have deter- mined the mean specific heats of bromine iodine sodium potassium and manganese over the temperature intervals -190' to 1s" and - 18" t o 18'. The mixture method was employed with water or toluene as calorimetric liquid. The specific heat of toluene a t the ordinary t,emperature was found to be 0.4015 f 0.0003 cal.From the observations the following average specific heat values are deduced bromine 0.07016 ( - 1'30' t o - 1s') ; iodine 0.0454 ( - 191' t o - SOo) 0.04852 ( - SO" to 1'7') ; sodium 0.2478 ( - 191" to - SO') 0.2714 (-SO' t o 17'); potassium 0.1551 (-191' t o -SO') 0,1677 ( - SO' to 18"); manganese 0.08203 (-18s' to - 79*2O) 0*10906 ( - 79.2" to 15'). These values agree in general very well with the numbers obtained by Koref (A. 1911 ii 964). H. M. D. Measurement of the Specific Heat of Solid Substances at High Temperatures. MARCELLO VON PIRANI (Ber. Deut. physika2. Ges. 1912 14 1037-1054).-Three methods are described by means of which the true specific heats of different metals have been determined at high temperatures.I n the first methcd the apparatus consists of a n insulated electrically heated tube furnace the temperature of which is adjusted to that at which the specific heat measurement is to be made. Inside this is a much smaller tube furnace the temperature of which is measured by means of a thermo-couple. If a quantity of electrical energy measured by a watt-meter is passed through the coils of the inner tube during time t its temperature will be raised by a small amount AO which is measured by the couple. Under exactly eimiiar conditions the energy required to produce the same rise of temperature when the inner tube is filled with the sub- stance under examination is determined. From the data thus obtained the specific heat of the substance may be calculated The two other methods are somewhat similar and depend on the fact t h a t when a conducting substance is heated electrically to a bigh temperature in a vacuum the electrical energy communicated is lost almost entirely in the form of radiant energy.If the supply of electrical energy is increased for a short interval of time the tempera- ture mill increase to an extent determined by the specific heat of the substance and by the increased amount of radiation. Since the latter case can be determined by experiments in the stationary con- dition it is obvious that the measurement of the extra electrical energy and of the consequent rise in temperature will afford a means of obtain- ing the specific heats of the metals at very high temperatures.The specific heats of tantalum and tuugsten have been measured in this way at temperatures between about 350" and 1400'. Data are also recorded for iron between 110' and 9'70". The curve showing the relationship between the specific heat of iron and the temperature isGENERAL AND PHYSICAL CHEhIlSTRY. ii. 103 very irregular showing a very decided maximum a t about 7 Oo and a sharp minimum at about 790'. No evidence of this minimum has been obtained in previous high- temperature measurements. H. M. D. Measurement of Specific Heats of Solid Substances at High Temperatures. ALFRED MAGNUS (Physikd. Zeitsch. 19 13 14 5-lO).-The measurements were made with a form of calorimeter similar to t h a t used by Nernst Koref and Lindemann (A. 1910 ii 263) but in order t o diminish the influence of the heat losses incurred in the introduction of the heated substances into the calorimeter its capacity was increased about fifty times.The block of copper weighing 22 kilograms was provided with a conical shaped central cavity for the reception of the heated substance which was contained in A conical silver tube fitted with a stopper. The rise in temperature of the copper calorimeter was measured by means of a battery of 50 iron constantan thermocouples half the junctions being situated in deep holes in the block of copper and the other half in the water-bath surrounding the D e w n vessel containing the copper block. Data representing the mean specific heats between the room tempera- ture and looo 270° 550° and 750' have been obtained the substances examined being quartz amorphous silica magnesium oxide calcium oxide zinc oxide lead oxide sodium and potassium chlorides sodium and potassium bromides and calcium carbonate.The molecular heats apperr t o be in satisfactory agreement with those arrived at by recent workers at lower temperatures continuous curves being obtained when the available molecular heats are plotted a s a function of the temperature. The values for sodium chloride potassium chloride and potassium bromide are also in agreement with those calculated from t h e equation Cp = C + ACp2T in which A = 8.7 2.7 and 3.3 x for the three salts respectively. H. A l . D. Investigations on Specific Heat. VI. Calculation of Atomic Heats. WALTHER NERNST and F. A. LINDEMANN (Sitzungsber.K. Akad. Wiss. Berlin 1912 1160-1171. Compare A. 1911 ii 464 964 1059).-On the assumption t h a t a vibrating atom gives rise to a continuous spectrum corresponding with a continuous series of component vibrations Debye (A. 1912 ii 1134) has obtained a modified Einstein formula for the specific heat of a monatomic element. This has been applied in the calculation of the atomic heats of aluminium copper silver diamond potassium and sodium chloride at different temperatures and the results are compared with the experimental data. This comparison shows t h a t the new formula agrees better with experiment than either of the older formulae of Einstein and of Nernst and Lindemann. The available specific heat data indicate that two groups of elements may be distinguished.In t h e one group the variation of the specific heat with the temperature can be satisfactorily represented by means of Debye's formula whilst the substances in the other group are characterised by a n appreciably less rapid fall in the specific heat with falling temperature than is required by the formula. Crystalline 8-2ii. 104 ABSTRACTS OF CHEMICAL PAPERS. elements belonging to the fir& group are probably monatomic whilst those of the second group are probabIy polyatomic. The fact that potassium and sodium chloride behave like monatomic substances is attributed t o the approximate equality of the vibration frequencies of the component atoms. H. M. D. Investigations on Speciflc Heat. VII. Calculation of Chemical Affinities. WALTHER NERNST (Sitzungsber.R. Akud. IT'iss. Berlin 19 12 11 72-1 176).-The effect of the substitution of Debye's formula for the specific heat in place of that used by Nernst and Lindemann is considered in reference to the application of the author's heat theorem to the calculation of chemical affinities. Tables are given which indicate the magnitude of the correction factors for a series of vi bratiori frequencies. H. M. D. The Polymerisation of Substances at Low Temperatures. JACQUES DUCLAUX (Compt. Tend. 1912 155 1509-1511. Compare this vol. ii 18).-A mathematical paper in which the author shows that the value of -- where dQ is the quantity of heat necessary at each temperature Y' to produce the modification studied To the absolute zero and 5'; is the absolute boiling point of the substance is a constant for all substances and equal to 42 or twice the value of the constant in the Pictet-Trouton law obtained by dividing the mole- cular beat of vaporisation by the temperature of vaporisation.The values are given as calculated for ten substances the specific heat of which has been completely studied. E d W. G. The Liquid Condition (Correction). WILLTAM C. McC. LEWIS (Z~itsch. phyeikal. Chem. 1913 81 626-628. Compare A. 1911 ii 855).-The author shows that the expression L= - Ta/& has only an emrirical meaning and he develops from thermodynamical principles the expression I = - 2h//3 in which I is the latent heat of expansion that is the amount of heat which must be added to a system t o cause an increase in volume of one unit whilst tho temperatureis kept constant a is the coefficient of expansion and p the compressibility.The expression only holds for homogeneous systems that is either entirely liquid or entirely gaseous. The author has calculated the values of 2 for a number of liquids and finds that its value is of the same order as the latent heat of vaporisation from which he draws the conclusion that the region of molecular attraction must be of the same order as the mean distance of the molecules from one another that is about cm. J. F. S. Simple Method for Determining the Melting Point of the Less Fusible Substances. XRN. HAVAS (Chem. Zeit. 1912 36 1438).-The author recommends a modification of the mercury method in which the substance of which the melting point is to be determined is placed in small quantity on the surface of mercury,GENERAL AND PHYSICAL CHEMISTRY.ii. 105 which is gradually heated. The mercury is replaced by an alloy of 2 parts of t i n and 1 of lead (ordinary soft solder) and melting points up t o 450" can readily be determined. T. S. P. Melting Point of Fire Bricks. C. W. KANOLT (Tech. Papers Bur. Stands. 1912 No. 10 I-l'I).-The melting point was taken as the lowest temperature at which a small piece of the brick could be distinctly seen to flow. An electric vacuum furnace was used and the melting points of fifty-four samples of fire brick including fire-clay bauxite silica magnesia and chromite bricks were determined as well as t h e melting points of the following materials important in the manufac- ture of fire brick kaolin 1740'; pure alumina 201bO; bauxite 1820" ; bauxite clay 1795 ; chromite 2180O.T. S. P. Pressure-temperature Diagrams for p - Azoxyanisole and a-Naphthylamine. NICOLAI A. PIJSCHIN and I. V. GREBENSCHTSCHIKOV (J. Rugs. Phys. Chem. Soc. 19 12 44 17%- 1736).-These diagrams have been investigated by the manometric method. TJp to pressures of 1088 kilograms per sq. cm. the melting point of p-azoxyanisole is a linear function of the pressure and is expressed by the formula tp = 135*9O+ 0.03949~. Hulett whose measurements were confined to the region 1-300 atmospheres found the value 0,0486 for the pressure-coefficient of the melting point (A. 1899 ii 468). this differing considerably from the authors' value 0.U395. A possible bending of the melting-point curve a t higher pressures such as occurs with the great majority of other substances has n o t been investigated.The transition point from the crystalline state t o the liquid-crystalline condition varies linearly with the pressure as far as 3000 kilograms per q. cm. and is expressed by tp = 11 7.3" + 0.02560p (Hulett Zoc. cit. foutid dT21dp=0*0320). Increase of the pressure from 2000 to 2645 kilograms per sq. em. is accompanied by increasing divergence of the transition point from the rectilinear curve this behaviour being similar to that of the melting points of substances crystallising with diminution of volume. The considerable difference between the pressuve-coefficients of the melting and transition points of p-azoxyanisole and the rectilinear character of the curve connecting melting point and pressure up to 1088 kilograms per sq.em. are regarded as contradictory to O~twald's assertion (Lehrbuch der allyemeinan Chemie 2nd Edition 11 2 392) tbat with liquid-crystalline substances small values tor the change of volume on melting and for the heat of crystallisation readily become zero a t pressures slightly in excess of t h a t of the atmosphere; or i n other words that with liquid crystals the critical point for liquid-to-crystals can be realised experimeutally. For pressures up t o 3000 kilograms per sq. cm. the melting point of a-naphthylamine is nearly a linear function of the pressure and is given by tp= 48-5O-t 0~01723p-0~000000555p2. The value 0.0168 for dT/tlp from 1 t o 500 kilograms per sq. cm.differs considerably from the value 0*0%10 found by Hulett (Zoc. cit.). The authors' results do not agree with those of Damien (Compt. Tend. 1891 112 785) which would indicate a maximum for the temperature-pressureii. 106 ABSTRACTS OF CHEMICAL PAPERS. curve of a-naphthylamine a t a pressure of 83 atmospheres The equation given above shows that a maximum may be expected at about 182’ (15500 kilograms per sq. cm.) so that a t higher tempera- tures and pressures crystalline a-naphtbylamine would be incapable of exietence. T. H. P. Thermal Analysis. T. VAN DER LINDEN (J. Chim. phys. 1912,10 454-466).-The paper gives an account of the theory of the method largely used by Holleman and his pupils of estimating the com- position of mixtures from freezing-point data alone.I n a homogeneous liquid with n components,. which form no com- pounds or mixed crystals a t atmospheric pressure the totlal weight of the system being 100 the degrees of freedom = n of which (n - 1) are variable concentrations and the remaining one is the temperature. On cooling the system one of the 12 constituents crystallises out at the first freezing point and as there are now two phases the degrees of freedom are reduced to (n - 1). The first solidifying temperature is therefore fixed by the proportions of the (n - 1) constituents remain- ing in the melt. Similarly the second solidifying point is a function of the ( n - 2 ) constituents remaining and so o n ; thus the (n-1)th freezing point is independent of the proportions of the (n - 2) con- stituents already solidified but dependent on the fatio a of the two constituents remaining in the melt that is the (9% - 1)th freezing point = f ( a ) .This function is represented by a curve similar to the ordinary binary freezing-point curve given by the two constituents in question. Hence the observation of the (n- 1)th freezing point enables the ratio cb of the remaining constituents to be estimated. Any pair of constituents can be arranged to crystallise out last by the addition of suitable amounts of certain of them and thus by the observation of the (n - 1)th freezing points of (n-1) solutions the composition of the original mixture of n substances is arrived at. Binary mixtures (9% = 2) are therefore analysed by one freezing-point measurement. Ternary mixtures require two estimations of second freezing points quaternary mixtures three estimations of third freez- ing points and so on.The observation of third and higher freezing points is very difficult. owing to the large amount of solid matter already present. With ternary mixtures it is necessary to obtain two complete second freezing-point curves with mixtures of known composition. Each of these curves ropresents the binary freezing- point diagram of t w o constituents saturated with the third. For instance when the freezing-point diagram of a ternary mixture has the “ideal ” form the three binary curves are superposable. The influeme of a quantity of 0 on the solidifying point of P being the same as the influence of an equimolecular quantity of M is also the same as that of an equivalent quantity of a mixture of 0 and icf mixed.In this ideal case the first freezing points of two solutions depositing P and 0 respectively say are sufficient to establish the compositions of t h e two solutions. This is tantamount to reducing a ternary mixture to a binary one. The method most used is to combine the two principles and establish I n some cases simplified methods are possible.GENERAL AND PHYSlCAL CHEMISTRY. ii. 107 the composition of the solution by one first and one second freezing point measured successively. Similar simplifications are sometimes possible in the case of quaternary mixtures particularly where some of the four constituents are isomeric substauces. R. J. C. Relations of Isomorphism in Organometallic Compounds.111. Derivatives of Bivalent Metalloids. PAUL PASCAL (Bull. SOC. chim. 1912 [iv] 11 1030-1037. Compare A. 1913 i 739).-The author has determined the freezing-point curves of binary mixtures of phenyl ether pfienyl sulphide phenyl selenide and phenyl telluride. These substances we1 e each purified by fractional crystallisntion and were found to have the following melting points 26" - 21.5' 2*5' and 4.2' respectively. Phenyl sulphide and phenyl selenide give a continuous series of mixed crystals the freezing-point curve passing through a minimum at -26*7" corresponding with the mixture containing 5% of the selsnide. Similar results are obtained with mixtures of the selenide and the telluride the minimum point being - 4*2" corresponding with the mixture containing S0.4% of the selenide and with mixtures of the sulphide and telluride the mixture containing 83.42% of the sulphide having tbe minimum freezing point - 30.7'.Phenyl ether and phenyl sulphide do not form a continuous series of mixed crystals the freezing-point curve exhibiting a eutectic point at - 27*7O corresponding with a mixture cont.3ining 13.3% of the ether. The two series of mixed crystals in equilibrium a t the eutectic point contain respectively 4.5 and 53% of phenyl ether. The above results agree with the classification of sulphur selenium and tellurium apart from oxygen. T. s. P. Internal P r e s s u r e and L a t e n t Heat of Liquids. WILLIAM C. McC. LEWIS (Pld. Mug. 1913 [vi] 25,61-65. Compare A. 1912 ii 136).-For the internal pressure of a liquid the author has obtained previously the expression K - 1 = I'dK/dl' where R is the internal pressure and I the latent heat of vaporisation OF 1 C.C.According to Davies (A. 1912 ii 903) the temperature-coefficient oE the internal pressure dKjKdT is equal to the coefficient of expan- sion a and by substituting in the above equation i t is found that K = Z / ( l + uT) which is identical with one of the results 'obtained by Davies. If I is considered to represent the latent heat of expansion of the liquid (the heat absorption corresponding with isotherinal expansion equal to 1 c.c.) it is deduced that 1 = - Ta//3 where a is the coefficient of thermal expansion and the compressibility of the liquid. For non-associated liquids the latent heat of expansion calculated from this equation is of the same order of magnitude as the latent heat of vaporisation.For associated liquids the differences are much greater. Values representing the internal pressure are calculated for a number of liquids from the equation K=Z/(l + a T ) on the assumption that I is the latent heat of expansion. The numbers so obtained are approximately double those recorded previously. H. 11. D.ii. 108 ABSTRACTS OF CHEMICAL PAPERS. Guldberg’s Law and the Corresponding States. ANATOLE LEDUC (Compt. rend. 1913. 156 65-66).-8 criticism of Boutaric’s deductions (this vol. ii 21) in which the author suggests a modification of Guldberg’s lam to read the boiling points of substances under equally reduced pressures are corresponding temperatures. Composition and Pressure of the Vapour of Solutions.31. S. VHEWKI ( J . IZuss. Phys. Chem. Soc. 1912 44 1739-1747).- The re-ults previously obtained (A. 1912 ii 1139) with aqueous solutions of methyl ethyl and proppl alcohols containing from 20-30% to 100% of alcohol are in complete accord with the relation e*tablibhed by Duhem and Mnrgules between the partial pressures of the vaponrs arid the compohitions of the solutions d In p,/d Zn x = d 2rL p 2 / d En( 1 -x). The applicttbility of this relation t o the solutions coritnir~irig less than 20-30% of alcohol remains undecided owing to t h e lack of det,ailed experimental data. The author discusses the numerical results from which he draws the following conclusions (1) The displacement produced i n the equi- librium between solutions of two liquids and their saturated vapours with constant volume of the system by change of temperature is siillject to Le Chatelier and van’t HOB’S law which may be formulated :IS f’ollows I n all cases of equilibrium between phirses of variable compouitiou heating of the system under constant volume leads in the phase formed with absorption of heat to increase of that component ttle transformation of which absorbs the greatest quantity of heat.(2) On the basis of Kirchhoff’s theory the change in the composition of tht vapour under the influence of temperature m4y be regarded as the result of the corn bined action of two Factors the physico-mechanical arid the phybico-chemical. I n every sysrem foi med with either development or absorption of heat there exists a region of parallel anti one of opposed action of the two factors.The limits of these regions ate the solutions corresponding with the maximum or minimum heat of formation. (3) The direction in which the change in composi- tion of the vapour proceeds under the influence of temperature remains the same for all concentrations of the solution if in the region where the actions of the two factors are opposed the magnitude of the physico-rYieehanica1 factor always predominates. (4) On change of texriperatu~ e of solutions formed without heat-effect the variation in the coniposition of the vapour takes place under the influence of a sirlgle physico-met hanical factor and follows the same law as does change in the cumposition of the vapour of mechanical mixtures ; (5) with change of temperature of a solution which possesses a vapour- prSesbu1.e curve showing a maximum the composition of the vapour and thxt of the mixture with constact boiling point change in one and the S;IWO direct:on.(6) If however the vapour-pressure curve shows :L r n i ~ i m ~ r i i the compositioris of the vapour and that of the mixture W. G. with constant boiiing point charrge in opposite directions. T. H. P. Association of Ethyl Ether and Chloroform in the Gaseous State. k’ni~~nrcrr DOLEZALEK and ALFHED SCHULZE ( B e y . Deut.yhys?kd. GBM. 1912 14 1~9l-lW6).-.tt is shown that the mixing of theGENERAL AND PHYSICAL CHEMISTRY ii. 109 vapours of ethyl ether and chloroform at 80” and atmospheric pressure is accompanied by contraction. When the volumes of the two vapours are equal the data obtained correspond with a reduction of pressure of 4.8 mm.of mercury when the volume is kept constant. I t is shown that the observed contraction is not due t o deviations of the vapours from the gas laws and the change in volume is consequently attributed t o the formation of a compound. The magnitude of the volume change indicates that an equimolar mixture of ethyl ether and chloro- form at SOo and atmospheric pressure contains 0.64 mol. % of the compound. By applying the lam of mass action it is seen t h a t the molecular proportion of the compound in thevapour mixture will increase almost; in the same ratio as the total pressure. Similarly it is shown t h a t the saturated vapour in equilibrium with an equiruolar liquid mixture of the two substances at 80” should contain 1.4 mol.% of the compound. Since the liquid phase contains 14.6 1x01s. % of the compound it follows from the vapour pressure law for binary mixtures that the vapour pressure of the compound at 80” should be about 0.26 atmosphere. H. M. D. Isotherms of Diatomic Gases and of their Binary Mixtures. XI. Determinations with the Volumenometer of the Com- pressibility of G a s e s under Small Pressures and at Low Temperatures. W. J. DE HAAS (Proc. K. Akad. Vetensch. Amsterdum 1912 15 299-306. Compare A 1912 ii 1138)-The question of the distribution of pressure between the volumenometer and the piezometer used in the mclawrements of the compressibility of gases at low temperatures is discussed and a formula is deduced for calculating the change in the difference of pressure between the two communicatitlg vessels as a function of the time.This formula involves the dimensions of the apparatris and the distribution of temperature along the connecting capillary. The curve corresponding with the formula is in sat ihfactory agreement with the actual observations of the change in the pressure difference with time. H. M. D. The C r i t i c a l Point. ETTORE CABDOSO (J. Chinz. pAys. 1912 10 4’70-496. Compare A. 1911 ii Sb4).-1n order t o evaluate the compression coeficients of gases to 1 part in 10,000 for atomic-weight caltwlations the critical tamperir tures and pressures must be known wihhin the limits 0.1’ and 0.1 atm. respectively. The temperature and pressure ought t o be observed a t the same time with the same specimen of gas so that any error due to impurity may affect both values equally. Although various investigators have given critical pressures to within 0.01 atm.tbe author considers that they were not justified since in general a variation of O * O 0 l o in the temperature corresponds with 0.01 atm. in the pressure. At the same time i t is admitted t h a t observations of pressure or temperature made separately a s by the Natterer tube may be more accurate than the author’s as far as the actual specimen handled is concerned. The paper dexribes the methods employed at Geneva in critical constant investigations. The gases were purified by fractional dis- tillation under vwuum (compare Briaer and Cardoso A. 1909 ii,ii.110 ABSTRACTS OF CHEMICAL PAPERS. 124) alternating with passage through suitable absorbents maintained slightly above the boiling .qoint of the gas. A sample of gas was not considered sufficiently purified until it could be liquefied at pressures constant to within 0.05 atm. a t two different temperatures one of which was near the critical point. A manometer of the barometer type was used consisting of a 1 mm. capillary tube with suitable bulbs in the upper and lower parts surrounded with a glycerol-water jacket maintained at a constant temperature. It was filled with atmospheric nitrogen the corn- pressibility of which is known froin Amagat's accurate data. The manometer was read to 0.1 mm. corresponding with not more than 0.02 atm. pressure. Owing to the uncertainty of the capiliary cor- rection in a 1 mm.tube cloher reading would have been useless and the use of a finer capillary would have introduced unknown errors due to diffraction and the retention of small gas bubbles below the meniscus. The different manometers agreed to within 0.1 a t m . at least and the accuracy of the methods was confirmed by the fact t h a t concordant critical data were obtained with different mano- meters and different samples of gas often at considerable intervals of time. The working tube was provided with a small piece of soft iron wire constituting the armature of an electromagnetic agitator of the Kuenen type. On approaching the critical temperature the meniscus disappears and the so-called " critical opalescence " is seen.The point of maximum opalescence is held by some to be the critical temperature (compare Nernst " Theoretical Chemistry "). The opalescence usually exteuds over a range of 0*03O but by means of the Kuenen agitator the range could be extended to 0.3Oto 0.4". The opalescence is therefore due to the formation of a n emulsion of gas and liquid which being of nearly equal density do not separ- ate. The true critical point is taken as the point a t which the opalescence disappears. This could be estimated t o within +_ 0.05" and is about 0.15' higher than the point of maximum opalescence. 16 is possible that an invisible opalescence persists at an even higher temperature since i n experiments made without the stirrer a slight diffraction could still be observed a t the spot where the meniscus had been.Critical data to the nearest 0.05 atm. and 0.05' are tabulated for hydrogen chloride ammonia carbon dioxide sulphur dioxide nitrous oxide hydrogen sulphide acetylene ethylene ethane and cyanogen. R. J. C. Critical Constants of Ethane Carbon Dioxide and Sulphur Dioxide. ETTORE CARDOSO and It. BELL (J. China. phys. 1912 10 497-503).-Xthane prepared from ethyl iodide by the Grignard reaction contained a i r and ethyl iodide and ether vapours as well as other impurities which were removed by suitable absorbents followed by fractional distillation. Ethane prepared by Frankland arid Kolbe's method by cautious addition of propionitrile to metallic sodium in :I vacuum was readily purified by absorbents only. The purified ethane having m.p. - 172.5" and b. p. -S4.l0 possessed a n agreeableGENERAL AND PHYSICAL CHEMISTRY. ii. 111 ethereal odour. I t s critical constants mere tc = + 32*10°F O * l O o and pc = 48.85 f. 0.10 atm. Carbon dioxide prepared by heaticg sodium hydrogen carbonate was washed and fractionated ten times. Its critical constants were t - + 31.00" +_ 0*103 and p c = 72-83 t- 0.10 atm. The critical opalescence mas observed as low as 30*60° but the point of maximum opalescence could not be decided. Commercial sulphur dioxide was washed and purified by ten distillations. I t s critical constants were t = + 157.15" t 0.10" and p = 77.65 0.10 atm. The point of maximum opalescence could n o t be observed owing to an opzlescent deposit on the glass tube. This appeared to result from the action of light and heat on the sulphur dioxide but i t was ascertained t h a t even 300 hours' heating did not alter the critical constants of the gas.R. J. C. Critical C o n s t a n t s of Ethylene N i t r o u s Oxide and Hydrogen Sulphide. XTTORE CARDOSO and E. ARNI (J. C h k . phys. 1912 10 504-508).-Ethylene prepared by the action of sulphuric acid on ethyl alcohol and purified by washing and fractional distillation had m. p. - 169.00' and b. p. - 104.3'. The critical constants were found to be t = +9*50'+_0*10' and pc=50*65 +_ 0.10 a t m . The critical opalescence extended from + 9.05" t o 9-50'. Nitrous oxide was prepared by the action of saturated aqueous sodium nitrite on hydroxylamine hydrochloride in a vacuum. When washed and fractionated ten times i t gave a perfectly colourless solid.The critical constants were tc = + 36-50'? O.lOo and pc = 71.65 & 0.10 atm. The critical opalescence extended from 36.0"to 36*5O and had its maximum a t about 36.3'. Hydrogen sulphide prepared by the action of hydrochloric acid on purified precipitated iron sulphide in a vacuum was mashed in water containing iron sulphide in suspension and then dried and fractionated fourteen times in a vacuum. The purified gas of f. p. - 83' and b. p. - 60.2' was without action on clean mercury and was perfectly stable at the critical point. The critical constants were tc = + 100.40" +_ O * l O o and pc = 89-05 5 0.10 atm. R. J. C. Critical Constants of Ammonia. ETTORE CARDOSO and (&file.) A. GILTAY (J. Chim. phps. 1912 10~514-516).-Arnmonia was prepared by heating purified ammonium chloride with excess of calcined marble.It was dried by barium oxide and sodium wire and distilled four times through the same reagents. Four further fractional distillations served to remove traces of air and hydrogen. tc = + 132.90" 2 O.lOo and pc = 112.30 maximum at about 132.75'. The critical constants were 0.10 atm. The critical opalescence was observed from 132.6" to 1 3 2 . 9 O with a R. J. C. C r i t i c a l Constants of Hydrogen Chloride. ETTORE CARDOSO and A. F. 0. GERMANN (J. Chiin. phys. 1912 10 517-519).- Purified sodium chloride was decomposed with pure concentrated sulphuric acid in a vacuum the 'apparatus being constructed with fused joints throughout. The hydrogen chloride was dried with sulphuricii.112 ABSTRACTS OF CHEMICAL PAPERS. acid and phosphoric oxide and fractionally distilled twelve times in all first through phosphoric oxide and afterwards without it. The solidi- fied hydrogen chloride was absolutely free from the pink t i n t usually attributed to a compound of phosphorus or t o nitric oxide. It had m. p. - 111-4° and b. p. - 83.1'. The critical constants were t,= +51-40° +_O.lOo andp,.=81*55+0.15 ntm. R J. C. Free Energy of Chemical Substances. Introduction. GILBERT N. LEWIS (J. Amer. Chem. Soc. 1913 35 1-30).-This paper is introductory to a series in which the free energy of chemical substances will be systematically studied. A detailed account i s given of the general methods used in free energy calculations including the con- sideration of the notation and fundamental units the laws of energy the change of energy with pressure the relation between activity and fugacity and the free energy free energy and the equilibrium constant free energy and E.M.F.and the influence of temperature on the free energy change. E. G. Tempering [of Metals] without Deformation. MAURICE HANKIOT (Compt. rend. 1912 155 1502-1504. Compare A. 1912 ii 1137).-Cubes of a number of metals and alloys were submitted to high pressure when immersed in vaseliu and i t was found that whilst the cubes were not deformed and the internal structure as examined microgrsphically for brass was iinaltered the substances had under- gone considerable tempering as was shown by measuring their hardness elongation and breaking strain before and after the compression.W. G. The R e c t i l i n e a r Diameter for Argon. EMILE MATHIAS H. KAMERLINGH ONXES and C. A. CROMMELIN (Proc. K. Akad. Netensch. 19 12 15 667-673).-8 preliminary account is given of the apparatus and mode of procedure adopted in the determination of the densities of liquid and saturated vapour at a series of different temperatures. The mode of calculating the requisite data from the experimental observations is described but no results are corn mun ica ted. H. 11. D. Specific Gravities of Saturated Aqueous Solutions of Various Salts at Different Temperatures. N. A. TSCHERNAJ (J. Buss. Phys. Chem.. Soc. 19 12 44 1565-1576).-The author has determined the specific gravities (weights of 1 c.c.) of saturated solutions of a number of salts at intervals of 10' from 0' to 70" (or 90'). These salts may be divided into two groups (1) those forming saturated solutions which increase i n specific gravity with rise of temperature and with in- crease in the salt-content potassium chloride nitrate and sulphate sodium and barium nitrates; ( 2 ) those with which the reverse takes place sodium chloride. These salts also show different behaviour with reference to the variation with temperature of' the quotient dP/d V where d P represents the increase per 10' of the amount of salt (grams) dissolved by 100 grams of water and d V the corresponding increase in volume (c.c.) ofGENERAL AND PHYSICAL CHEMISTRY. ii.113 the solution. With rise of temperature from 0’ to 70° this quotient shows a continuous increase from 0.12 t o 0.5 with sodium chloride remains almost constant with potassium chloride (1.69) sodium nitrate (1.9) potassium nitrate (2-1) and potassium sulphate (2*2) and con- tinually diminishes from about 5 to 1.6 with barium nitrate.T. H. P. Compressibility of Gases. GEORGES BAUME and E. WOURTZEL (J. Chirn. phys. 1912 lo,’ 520-522).-’rhe relation of pressure to volume in sulphur dioxide methyl ether and chloroform vapour has been represented by the equation pv= 1 +a(l - L/L,) where L is the density at pressure p and L the weight of a normal litre of the gas considered (compare Baume A. 1908 ii 372). The deviation co- efficient a is numerically the same as Berthelot’s deviation coefficient Al iu the case of permanent gases but with easily liquefkble gases A = u/(1 +a).This equation may be reduced to A = a - a2 when cc3 is negligible. R. J. C. Adsorption and Saturated Surfaces. ROBERT MARC (Zeitsch. pliysikul Cf~em. 1913 81 641-694 Compare Schmidt A. 1910 ii 1041 ; 1911 ii 969).-L’he adsorption of a large number of substances on the cryst-tl surfaces of barium sulphate bdrium carbonate rhombo- hedral calcium carbonate strontium carbonate ledd carbonate and lead sulphate was investigated by measuring the change in the refractive indices of the solution of many substances by means of a n interferomet w. I t was shown that colloidal substances are easily adsorbed on crystal surfaces whereas crystalline substances are only adsorbed t o a very slight extent. Crystalline substances adsorb more if they are capable of forming isomorphous or mixed crystals with the adsorbing substances ; thus potassium nitrate is adsorbed by barium carbonate whil-t sodium nitrate is not ; on the other hand sodium nitrate is adsorbed by rhombohedra1 calcium carbonate whilst potassium nitrate is not.The adsorption Lothermals can be divided into three types (1) Those in which a stronger bend is present than is demanded by the exponential formula and which reach the satura- tion value continuously. (2) Those which agree at all points with the exponential formula. I n this case even though the measurements were made very close together it was impossible t o determine whether there was a break in the curve or not. (3) Those curves of substances which at low concentrations have a great tendency to be adsorbed and give a very steep curve and reach the saturation point at very low concentrations.The adsorbed substance is the factor which conditions the type of curve in any case. It is shown that when the saturation values for a series of substances with one given adsorb- ing medium have a given ratio then the ratio of the saturation values for those substances with another adsorbing medium will be the same. The formulae of Schmidt and Arrhenius are discussed and it is shown that the Arrhenius formula holds well for Substances giving curves of the first type but for curves of the second type it holds only over a portion of the curve and ceases t o hold some distance before the saturation value is reached whilst for curves of the third type itii.114 ABSTRACTS OF CHEMICAL PAPERS. does not hold at all. It is however shown that; the Arrhenius formula probably represents ideal conditions which axe only existent when molecular compounds are not formed in the adsorption layer. J. F. S. Adsorption. VII. Chemical Hysteresis of Starches. ADAM V. RAKOVSKI (J. Russ. Yhys. Chem. Soc. 1912 44 1722-172S),- The investigations here described are a continuation of those published in papers I. and 11. (A. 1911 ii 470) and deal with curves inter- mediate t o those of hydration and dehydration of Bermuda arrowroot and rice starches. These intermediate curves are found to be virtually reversible but not throughout their whole length. Detailed study of such curves would however be possible only with colloids showing a considerably greater separation of the curves of hydration and dehydration than is the case with starches. The presence of about 1% of mercuric chloride in potato starch exerta no appreciable influence on the course of either hydration or dehydration.T. H. P. Relation between the Conductivity of Acids and their Absorption by Hide. ANDRB BROCHET (Compt. rend. 1912 155 1614-1617).-Measurements have been made of the amount of acid absorbed by shaking 10 grams of hide powder for four hours with 200 C.C. of a n N/lO-solntion of the acid concerned containing 100 grams of sodium chloride per litre. lVleasurements have been made with three classes of acids varying in conductivity namely (1) good conductors (2) moderate conductors (3) poor conductors. In class (1) inorganic acids and organic acids of the type of trichloroacetic acid were examined.The absorption of acids by the hide is a general phenomenon and is the result of a chemical combination the amount of acid absorbed being proportional to the chemicai equivalent of the acid. The fraction of the gram-equivalent of acid absorbed by one kilo. of dry hide whilst constant for each class does show a slight diminution with diminution in conductivity ; but while the absorption diminishes only in the ratio 3 2 the conductivity diminishes as 100 1. W. G. Some Properties of the Tbiocyanate Ion. HERBERT FREUND- LICH and A. N. SEAL (Zeitsch. C’ham. lnd. Kolloide 1912 11 257-263).-The lyotropic properties of solutions of potassium chloride and potassium thiocyanate have been compared in a quantita- tive manner with the object of determining the position of the thio- cyanate ion in the lyotropic series.The properties examined were the following the influence of the electrolyte on the solubility of benzoic acid the effect on the surface tension the absorbability by animal charcoal the retarding influence on the rate of ester saponification and the elfect on the rate of increase of the viscosity of gelatin solutions. The observations show that the lyotropic influence of the thiocyanate ion is distinctly smaller than that of the chloride ion and that its position in the series of anions is close to the iodide end in theGENERAL AND PHYSICAL CHEMISTRY. ii. 115 series fluoride sul phate,. phosphate chloride nitrate bromide iodide. In this series the ljotropic influence diminishes from left t o right.H. ItT. D. Surface Tension of P r o t e i n Solutions. 11. FILIPYO BoTTAZZr and E. D’AGOSTIKO ( A t t i 12. Accad. Lincei 1912 [v] 21 ii 561-572. Compare Bottazzi A. 1912 ii 1043).-The paper records in tables and curves the results of further experiments with Traube’s stalag- mometer on solutions of serum-albumin t o which various quantities of hydrochloric acid sodium hydroxide or sodium chloride were added. From the experiments i t appears that the value n (number OF drops) depends entirely on the undissocieted molecules of the protein or of its salts for a n increase in the concentration of these molecules increases n (that is the surface tension is lowered). I n order t o obtain the maximum diesociation of the albuminate it is necessary to add more acid or alkali in weak solutions than in stronger ones.This indicates the presence of salts of a weak base (or acid) and for the same reason the maximum dissociation will be reached only when more than the equivalent amount of sodium hydroxide or of hydrochloric acid is added. From the amounts thus added i t appears that the molecular weight of the albumin ehould be somewhat above 1000. From the curves it also appears t h a t the albumin chloride is much less disso- ciated than the sodium albuminate a t the same dilution. The addition of sodium chloride diminishes the dissociation of both the albumin salts but its influence is much greater on the chloride than on the albuminate. The addition of sodium chloride also diminishes the dissociation of the free albumin probably owing to the formation of undissociated secondary products.WILLIAM E. S. TURNER (J. Chim. phys. 1912 10 467-469. Compare Guye A . 1911 ii 1067).-The author agrees with Guye that liquids which give a high value of Ramsay and Shields’ constant are not necessarily dissociated since error may arise from abnormality in the surface film of the liquid but cannot agree that in his experiments criticised by Guye the surface films of diphenylamine and phenylurethane were oxidised. The value obtained with diphenylamine was in accord with t h a t obtained by Dutoit and Friedrich (A. 1900 ii 194). Mode of I o n i s a t i o n of Sulphuric Acid i n Dilute Aqueous Solutions. JOSEPH A. MULLER (Comnpt. rend. 1912 155,1499-1502).-From a series of determinations of the coefficients of ionisation and heat of dilution of aqueous solutions of sulphuric acid the author deduces that in di1ut.e aqueous solutions sulphuric acid is ionised i n t o the two ions S0,H’ and H’ and that this ionisation is accompanied by development of heat within the limits of the experimental temperatures Applicat.ion of the Theory of Chemical Potential t o the Thermodynamical Theory of Solutions. 111. Action of G r a v i t y on a Solution. The Solute Potential. E x t e n s i o n of the Theory. SYDNEY A. SHORTER (Phil. Mag. 1913 [vi] 25 31-42. Compare A. 1912 ii 24 437)-From a general theorem R. V. S. Molecular Complexity in the Liquid State. R. J. C. (1 4-38’) w. G.ii. 116 ABSTRACTS OF CHEMICAL PAPERS.enunciated by Gibbg the author deduces a formula for the influence of gravity on a binary ruixture. This formula is found to be in accordaoce with the equation obtained directly by Dtihern. Formtile are also given for the practical calculation of the gravity effect from osmotic pressure vapour pressure and freezing-point data. I n the second part of the paper the theory given in Parts I and TI. (Zoc. c i ~ ) is extended to solutions containing any number of non-volatile solutes. Exact formulae are deduced connecting the osmotic pres,sure vapour pressure and freezing poiot with the lowering effect exerted by the solutes on the chemical potential of the solvent. Thermo- dynamically the solutes may be regarded as a single substznce,and hence the formulae relating t o a binary solution may be generalked i n a very simple manner.H. M. D. Studies of the Processes Operative in Solutions. XX. The Conversion of Ammonium Cyanate into Carbamide Especially as Influenced by Alcohols. ERIC WALKER (Proc. Roy. Soc. 1913 A 8'7 539-554. Compare J. Walker and Kay T. 1897 489).-Experiments have been made t o determine the influence of ethyl propyl and isobutyl alcohol on the rate of trans- formation of ammonium cyanate into carbamide when the ratio of cyanate t o water in the solution 1s kept constant. The measurements were made at 40" and the rate of change at any moment way obtained by determining the tangent t o the smooth curve drawn through the points representing the cyanate concentrations after a series of time intervals. I n aqueous solution the rate of change varies approximately as the square of the concentration of the cyanate as found by Waiker and Hambly ( r.1895,?'1,746) but the carbamide and ammonium carbonate which are formed have some influence on the velocity. Sirice the carbonate is formed very rapidly a t the beginning of the experiment the velocity coefficient is abnormally high at first and during the later stages of the change the coefficient is again too large because of the accelerating effect of the ammonium carbonate. The addition of carbamide t o t.he original solution has also the effect of increasing the mean velocity coefficient a result which is probably due to its influence on the side-reaction by which ammonium carbonate is produced. The data obtained in a series of experiments in which the molar ratio of ammonium cyanate to water was 0.15 100 and iu which tho number of mols.of ethyl alcohol was increased from 0 to 180 per 100 of water show t h a t the rate of change is increased in proportion to the amount of alcohol present. When the alcohol and water are presert in equimolar proportions the rate of change is however only about five times as large as the rate in the absence of alcohol (compare Walker and Kay From comparative experiments in which ethyl propyl and isobutyl alcohols were added to the aqueous solution of the cyanate it appears that the activity increases considerably with the molecular weight of the alcohol. The percentage increase in velocity per mol. of added alcohol was found t o be 3-30% 4-92% and 7.3% respectively for a solu- loc.cit.).GENERAL AND PHYSICAL CHEMISTRY. ii. 117 tion containing 0.15 mol. of cyanate per 100 mole. of water. The order in which the alcohols are arranged is the same as that obtained on contrasting their activities as precipitants of salts from aqueous solutions and their physiological activity as hormoneq. It is supposed t h a t the action of the alcohols is largely mechanical the larger molecules having the greater effect both on account of their size and their greater mobylity by reason of their smaller affinity for water. H. M. I?. Studies of the Processes Operative in Solutions. XXI. Hydrolysis of Sucrose by Dilute Acids. FREDERICK P. WORLEY (Proc. Roy. SOC. 1912 A 87 555-563).-The conclusion drawn by Armstrong and Caldwell (A.1904 i 1070) t h a t the hydrolytic action of dilute acids on sucrose is closely analogouq to that of enzymes has been subjected t o examination in a further series of experiments with dilute sulphuric acid at 25". I n three comparative observations the solutions contzined 0-01 mol. acid 200 mols. of water and 1 2 and 4 mols. of sucrose respectively. From the measured time changes in the rotation of the solutions i t is evident t h a t the rate of hydrolysis diminishes as the reaction proceeds in accordance with the requirements of the mass law. The view that the hydrolysis in dilute acid solution is characterised by a n initial period in which the sucrose is hydrolysed at a constant rate finds no support from the data which have now been obtained. Thero is therefore a considerable difference between the sucroclastic action of dilute acids and small quantities of enzymes.It is pointed out that this conclusion has already been drawn by Rosanoff Clark and Sibley (A. 1912 ii 34) but t h a t the arguments put forward by these authors are unsound. Studies of the Processes Operative in Solutions. XXII. Hydrolysis of Sucrose by Sulphuric Acid ; Improvements in Polarimetric Apparatus. FREDERICK P. WORLEY (Pvoc. Roy. Xoc. 19 12 A 8'7 563-581).-Similar experimeuts to those described i n previous papers have been made with sulphuric acid as a catalyst in order to ascertain the influence of dilution on the hydrolytic activity i n the case of a dibasic acid. The various substances were used in the proportion of 1 mol. of sulphuric acid 30 to 200 mols.of water and either 0.125 or 0-25 mol. of sucrose. From the velocity coefficients at the different dilutions the value of the apparent molecular hydration has been calculated. This increases from 12.2 t o 15.8 when the molar ratio of water t o acid increases from 30 to 80 remaining practically constant on further dilution. By comparing the results with those obtained for hydrochloric nitric and p-dichloro- and p-di- iodobenzenesulphonic acids it is found t h a t the maximum apparent hydration is reached at a n earlier stage in the case of sulphuric acid than when the monobasic acids are employed as catalysts. Prom the final and initial rotations of the differently concentrated solutions it ap ears t h a t the degree of optical inversion varies with of the rotations was found to be 0.342 whereas 0.298 was obtained for the most dilute solution.This variation is attributed t o the influence H. hl. D. the dilution. P n the case OF the most concentrated solution the ratio VOL. CIV. ii. 9ii. 118 ABSTRACTS OF CHEMICAL PAPERS. of the acid on the rotatory power of the sugars present the lzevulose being responsible in all probability for the major part of the effect. I n view of this phenomenon and of the influence of the mutarotation of the dextrose and laevulose it might be doubted whether the polarimetric method affords a satisfactory measure of the rate of hydrolysis of sucrose. A theoretical examination of the subject from this point of view shows however that the velocity coefficient is not affected either by an alteration in the rotatory power of the invert sugar as a consequence of the presence of the acid or by the muta- rotation of the monoses.Considerable improvements have been made in the apparatus employed for the polarimetric study of chemical changes and these together with the precautions which must be observed in accurate work are described in detail. H. M. D. Studies of the Processes Operative in Solutions. XXIII. Hydrolysis of Metbyl Acetate by Acids. FREDERICK P. WORLEY (Proc. Boy. Xoc. 1912 A 87 582-603. Compare Armstrong and Watson A. 1907 ii 149).-With the object of determining the apparent molecular hydration of the acid by the method employed in connexion with the hydrolysis of sucrose (P. 1910 26 298) measurements have been made of the rate of hydrolysis of methyl acetate under the influence of hydrochloric acid.The experiments were carried out a t 25" the molecular proportions of the substances employed being 1 cf methyl acetate 120 of water and from 0.6 to 4 of hydrochloric acid. Under these conditions the proportion of methyl acetate ultimately hydrolysed is never greater than 95% and in consequence it is necessary to take the reverse chango into account in calculating the velocity coefficient of the hydrolysis the reaction taking place in accordance with the equation dx/dt = k(a - z) - k1x2 in which a is the initial concentration of the ester and x the amount transformed after time t. For a given ratio of water and methyl acetate the proportion of the latter which is hydrolysed when equilibrium is attained decreases slightly as the concentration of the catalyst increases.In view of this it is suggested that the usual interpretation of the equation k[CH,*C02CH3][H20] = k1[CH3*C0,H][CH,0H] is incorrect and that since hydrolysis is probably effected by the direct interaction of hydrated hydrolyte and hydrated catalyst the factor [H,O] does not necessarily represent the active mass of one of the substances taking part in the reaction but probably expresses the degree of dilution. From the velocity coefficients values representing the apparent molecular hydration of the hydrochloric acid have been obtained. Corresponding values have also been calculated for sodium potassium and ammonium chlorides from the increase which these salts produce in the velocity coefficient.These molecular hydration values are much smaller than the Corresponding numbers obtained when sucrose and raffinose are used as hydrolytes. Whereas with methyl acetate the apparent hydration value increases from 2.5 to 5 when the molecularGENERAL AND PHYSICAL CHEMISTRY. ii. 219 ratio of water to acid increases from 30 to 80 the corresponding series of numbers from the data for the hydrolysis of sucrose increases from 12 4 to 18.7. It is considered probable that the differences involved repre$ent actual differences in t’he condition of the acid in presence of such different hydrolytes as methyl acetate and sucrose. H. M. D. Studies of the Processes Operative in Solutions. XXIV. Nature of the Hydrolytic Process. HENRY E. ARMSTRONG and FREDERICK P.WORLEY (PTOC. Roy. SOC. 1912 A 87 604-623).- The authors discuss the general character of the results obtained in this series of papers and their theoretical interpretation. As indicated in previous communications tha view is put forward that hydrolysis is essentially an associative process which involves the association and direct interaction of two complexes one of which consists of the hydrated hydrolgte and the other of the hydrated catalyst. Such associated systems are being constantly produced broken down and reformed in such a manner thttt while some give rise to the original components others are resolved into the products of change. According to this view hydrolysis is a bimolecular change the second factor being the active mass of the hydrated catalyst and not as is generally supposed the active mass or concentration of the water.The similarity of the results arrived a t by the study of different properties such as electrical conductivity f;;ydrolytic activity and osmotic effects is quite consistent with this view the similarity being due to the fact that the determining factor in all these cases is the interaction involved in the production of the electrolyte from water and the solute. The increase of molecular Conductivity t o a maximum on dilution is attributed partly to a gradual increase in the extent of the interaction between the solute and water and partly to the gradual simplification of the complexes formed by the dissolved substance. On the other hand the decrease of electrical activity as exemplified by the hydro- lytic activity of the acids to a minimum on dilution is considered to be the necessary result of a gradual weakening of the acid by further combination with water.Ths greater the extent to which the activity of the acid is used up in combining with water the less must be the residual activity available for hydrolytic processes. Stress is laid on the fact that this associative theory is directly opposed to the generally accepted dissociative hypothesis and i t is claimed that the facts disclosed in the investigation of the processes operative in solution go far to show that this hypothesis is no longer tenable. H. M. D. Intercrystalline Cohesion in Metals and the Formation of Twinned Crystals in Silver. WALTER ROSENHAIP; and DONALD EWEN (J.Inst. Metals 1912 8 149-185).-Many of the properties of metals especially at high temperatures may be accounted for by assuming the presence of an amorphous film between the crystals. Tho hypothesis hits been tested in the following manner. The amorphous modification being unstable should have a higher vapour- 9-2ii. 120 ABSTRACTS OF CHEMICAL PAPERS. pressure than the crystalline. Two specimens of silver are taken one having coarse and the other fine crystals. The latter which has a larger area of intercrystalline boundaries in unit volume than the former always loses weight more rapidly when heated in a vacuum. Similar results are obtained with zinc and copper. Microscopical examination shows t h a t the intercrystalline boundaries are developed as if by etching by such treatment.Boundaries between twin crystals are not thus affected. Cast silver even without any mechanical treatment shows numerous twinned crystals which are revealed by heating in a vacuum. 0. H. D. Formation of Twin C r y s t a l s by Quenching and its Influence on the Hardness of Metals. CHARLES A. EDWARDS (Internat. Zeitsch. Metallopaphie 191 2 3 179-194).-The light and dark acicular structure observed in quenched alloys of copper and aluminium containing 9-16% Al is not due to the presence of two constituents but to repeated twinning brought about by the mechanical pressure due to quenching. The vitreous modification of the metal is formed at the surfaces of slip thus increasing the hard- ness of the alloys. This explanation is extended to the general case of hardening of alloys by quenching.C. H. D. E t c h i n g at E i g h Temperatures. H. HANNEMANN (Ifiternat. Zeitsch. Mctallogrupivie 1912 3 l76-178).-The method of investi- gating the crystalline structure of alloys at a high temperature by exposure to etching vapours is faulty because the surface etched is not a section through a crystal but the surface of a crystal and segregation is not thereby reveded. Further recrystallisation may take place during the action of the reagent thus producing a false structure. C. H. D. Colloids. PAUL B ~ R Y (J. Chin&. phys. 1912 10 437-453. Corn- pa.re A. 191 1 ii 702).-Colloidal solutions comprise three classes namely (1) Jellies which are solutions of liquids in solids of great cohesive power.(2) Electrical colloids which are suspensions of very fine insoluble powders without solvent power which acquire a n electrical charge in contact with the liquid. Colloids of this class exhibit Brownian motions and are coagulated by reducing their electro- static charges. (3) Mixed colloids consisting OF micella formed from irregular particles of jelly by the action of excess of liquid. These micella exhibit Brownian motions and may acquire an electrostatic charge. They can therefore be coagulated by altering the external osmotic conditions or by electrical methods or by both combined according t o circumstances. The micella in class (3) are to be regarded as minute osmotic cells into which the surrounding liquid diffuses until its pressure is balanced by the surface tension or cohesion of the small particles. The surface tension of the colloidal substance is negative to start with and tends t o approach zero as the jelly becomes saturated.Caoutchouc (Red Tonkin unvulcanised) absorbs approximately the same volumesGENERAL AND PHYSICAL CHEMISTRY. ii. 121 of chloroform benzene carbon disulphide and tetrachloroethane to form saturated jellies before breaking up into micella. I n some cases the surface tension does not attain zero so t h a t the limit of osmosis inwards is reached in the jelly stage and the true solution of class (1) does not pass into a pseudo-solution of class (3) on further addition of liquid. Examples of this are the swelling of gelatin in cold water and of cellulose tetra-acetate in cold tetra- chloroethane.I n such cases micella can usually be obtained from the jellies by heating or even by violent agitation. The jelly separates out again on cooling as with cellulose tetra-acetate unless Brdwnian motion and electrification of the micella supervene as in the cdse of gelatin. R. J. C. Determinations of the Volume of Voids in Silicic Acid Gels. WILHELM BACHMANN (Zeitsch. anorg. Chem. 19 12 79 202-208).- The transierit opacity of silica gels during dehydration has been considered by Tschermak and also by Tarnmann to indicate the appear- ance of a new hydrated phase. On the other hand the fact that the same change is observed when such different liquids as cedar oil or a mixture of olive oil and chloroform are used indicates that the process is one of mechanical imbibition.The view of Zsigmondy (A. 1911 ii 880 ; Bachmann A. 1912 ii 145) that the gel is traversed by minute capillaries accords best with the facts. It is now shown that the weight of different liquids taken up by a gel is proportional t o their density. The gel is prepared by exposing glassy silicic acid to steam and then washing with water and drying over sulphuric acid. T h e e gels are examined the liquid being generally introduced by exposure to its vapour. Measurements with water benzene cbloroform ethyl iodide and acetylene tetrabromide the last being used in the liquid form only give concordant results. c. H. D. Theory of Emulsification. IV. WILDER ID. BANCROFT (J. Physical Chem. 1912 16 739-758. Compare A. 1912 ii 834). -Robertson’s paper on emulsions of oil aud water (A.1910 ii 697) and the article on emulsions in Remington’s ‘‘ Practice of Pharmacy ” (1907) are reproduced in full. Robertson’s work is valuable because he was able to prepare emulsious of water in oil as well as of oil in water. Moreover by means of a n especially efficient shaker he emulsified the whole of his ingredients iu one operation instead of working in the dispersed phase gradually. The author holds t h a t this succeshf ti1 preparation of emulsions of water in oil was due t o the fact that the olive oil used has some solvent power for the emulsifier sodium oleate whereds kerosene and benzene in which the soap is insoluble give no such emulsions. The author does not accept the conclusion drawn by Robertson that one type of emulsion passes into the other at a definite concentration but argues that there may be a range of concentrations within which no emulsion whatever is produced or the emulsions may overlap.The emulsions known t o Pharmacy which contain upwards of 95% of oil are all of the oil in water type with gum acacia egg-albumin orii. 122 ABSTRACTS OF CHEMICAL PAPERS. casein as emulsifying agents. The object of the pharmacist is to divide the oil into minute globules and surround each one with a n adhesive envelope. The exact proportions of oil water and gum are probably not so important as is generally supposed. Composition of the Disperse Phase in Emulsoids. EMIL HILTSCHEK (Zeitsch. Chem. Ind. Kolloide 1912 11 284-286).-0n the assumption that the velocity of displacement of the juxtaposed layers exceeds a certain critical value it has been shown (A.1911 ii 9s) that the viscosity v’ of a n emulsoid is given by the equation 7’ = y~ u2/(uA - l) in which 77 is the viscosity of the dispersive medium and A is the ratio of the volume of the emulsoid to that of the disperse phase. If the viscosity of the pure dispersive medium is taken as unity this equatlion gives A = (q’/(v‘ - l)13. From the viscosity data for a series of glycogen and sodium casein hydrosols the author has calculated the values of A given by this formula. When these values are compared with the values of A’ representing the ratio of the volume of the ernulsoid t o the weight of disperse phase present it is found that the ratio of A’ to A remains very nearly constant if the less concentrated hydrosols are left out of account.The constancy of this ratio indicates that the disperse phase consists at a given temperature of the dissolved substance together with a definite and constant quantity of the dispersive medium. The Existence and Probable Thickness of Adsorption Envelopes on Suspensoid Particles. EMIL HATSCHEK (Zeitsch. Chem. Ind. Kolloide 1912 11 280-284).-The properties of suspensoid systems which depend on the movements of the particles indicate that the volume of the disperse phase is not independent of t h e degree of dispersity. On the assumption t h a t this is due t o the formation of an envelope of the dispersive medium round each suspensoid particle and that the thickness of the covering film is constant for varying degrees of dispersity i t can be shown that the effective volume V’ and the actual volume V of the disperse phase are connected with one another by tho equation Y’/Y= (1 + 3t/r) where 1.is the radius of the nucleus of disperse phase and t is the thickness of the envelope of dispersive medium. From this i t is evident that the effective volume of the disperse phase will increase continuously as the degree of dispersity increases. Experimental measurements have shown that the viscosity of highly disperse systems increases with the degree of dispersity when the proportion of disperse phase is kept constant and this is probably due t o the incr.e:tsing importance of the surrounding envelope of dispersive medium. From OdBu’s measurements of the viscosity of colloidal sulphur solutions of different degrees of dispersity (A.1911 ii 971 ; 1912 ii 240) the arithor has calculated by means of the above formula the thickness of the surrounding envelope t o be 0.87pp. Accepting this value it follows that for colloidal particles of diameter lOpp the volume of the envelope amounts to 62% of the volume of the colloidal sulphur. H. M. D. R. J. C. H. M. D.GENERAL AND PHYSICAL CHEMISTRY. ii. 123 The Equilibrium of a Gas in a State of Binary Dissociation. J. DE BOISSOUDY (Compt. rend. 1913 156 61-64).-A mathematical discussion of the equilibrium of a gas such as nitrogen peroxide acetic acid vapour etc. when partly dissociated into two identical constituents. The equilibrium is expressed by the equation X x2]v( 1 - x) = MT3e-= where x is the degree of dissociation w the volume containing one gram-molecule 5" the temperature .X the energy necessary to dis- sociate the normal molecules into their constituents and M a constant coefficient.W. G. Action of Temperature on the Equilibrium of Nitrous and Nitric Acids Formed from the Oxides of Nitrogen and Water. EMIL BRINER and E. L. DURAND (Compt. rend. 1912- 165 1495-149T).-In a previous paper (A. 1912 ii 1045) the authors studied the effect of pressure and concentration on various systems of oxides of nitrogen and water the temperature remaining constant. They have now estimated. the relative molecular amounts of nitrous and nitric acids formed in the solutions a t varying temperatures and find that increase in pressure of the nitric oxide and diminution of temperature favour the formation of nitrous acid.W. G. The Equilibrium in Acid Solutions of Potassium Salts. 11. ALBERT J. J. VANDEVELDE (Bull. SOC. chirn. Belg. 1912 2G 513-532. Compare A 1912 ii 30).-A continuation of the study of the constitution of the solid phase obtained from acid solutions of potassium salts. I n the former communication the salt chosen was pottssium sulphate and the three acids hydrochloric nitric and sulphuric. To complete this study the equilibrium has been deter- mined (1) for solutions of potassium chloride (2) for solutions of potassium nitrate mixed in turn with one of the three above- mentioned acids in such proportions as to produce a solid phase. I n the systems studied the same results are obtained by having present the same ions in equal quantities independently of the manner of their original combination ; thus from solutions containing 20KCl; 10H2S04 or 10K2S04 20HC1 the solid phase in each case had the composition 8KCI 4KHS0,.I n the system potassium chloride sulphuric acid and water the blid portion was only a single phase consisting of potassium chloride when the sulphuric acid was in the proportion 40KC1 5H,SO or less the solubility of the chloride diminishing as the concentration of the acid diminished. On replacing the sulphuric acid in the system by hydrochloric acid the solid phase was a1 ways potas4um chloride the solubility increasing with diminution in concentration of the acid. In the system potassium chloride nitric acid water the solid phase in all cases contained both potassium chloride and nitrate the amount of the former diminishing and of the latter increasing with rise in concentration of the acid.In the case of potassium nitrate sulphuric acid and water theii. 124 ABSTRACTS OF CHEMICAL PAPERS. solid phase consisted solely of potassium nitrate and its solubility seemed but very slightly affected by the concentration of the acid. On replacing the sulphuric acid with nitric acid the nitrate was found to be less soluble but to an extent independent of the acid concentra- tion. I n the system potassium nitrate hydrochloric acid water the solid portion was a single phase potassium nitrate until the acid reached the concentration 20KN0 20HC1. W. G. Heterogeneous Equilibria between Aqueous and Metallic Solutions.11. Interaction of Mixed Salt Solutions and Liquid Amalgams. GEORGE MCPHAIL SMITH (J. Amer. Chem. Soq. 1913 35 39-49. Compare A. 1910 ii 401).-This work was under- taken for the purpose of studying by an independent method the ionisation relations existing in mixtures of salts. By agitating dilute sodium or potassium amalgam with successive portions of a solution of sodium and potassium chlorides or sulphates a mixture is soon obtained i n which a t equilibrium the concentrations of the salts in the mixed solution are identical with those in the original saIt solution; the amalgams are then analysed. The ion fractions of sodium and potassium in experiments in which the salts were present in equivalent quantities have been calculated on t,he assumptions that (1) the reaction takes place according to the equation KHg,+ Na' S NaHgn+ K' + (m - n)Hg and (2) that in the solutions containing a common ion and having a total salt concen- tration 0*2N the relation Na salt/K salt = Na'/K' is approximately true.The results show that the sodium ion fraction increases with the total salt concentration of a solution. In order to explain this phenomenon it is suggested that sodium and potassium chlorides in mixed aqueous solution form the complexes Na(CI-K*Cl) and K(C1.Na-Cl). The formation of the latter complex would lower the value of the sodium ion fractioqand the conclusion is therefore drawn that the preponderating complex is Na(Cl*K*Cl) and t h a t this ionises into Na' and (CI*K*Cl)' ions. E. G. Influence of Temperature on the Velocity of Chemical Reactions.B. SCHVECOV (J. Russ. Phys. Chem. Soc. 1912 44 Yhya. Part 470-474).-Owing to the inconstancy of the ordinary temperature-coeficient of chemical .reactions which is represented by the expression y = (R,/K,)lO/(t - t l ) where Kl and K are the velocity constants a t the temperatures 8,' and t,' respectively Plotnikov (A. 1905 ii 376) suggested the use of the so-called logarithmic tempera- ture-constant given by a = (logK - logK,)/(t - t,). Comparison of these two equations shows that a would be constant only over intervals of temperature for which y gives constant values. Auerbach (A. 1905 ii 57 1) has indeed shown that Plotnikov's logarithmic temperature- constant is a variable magnitude. From theoretical considerations the author regards i t as more likely that constant values will be obtained for the ratio between the velocities of reaction if these are calculated for absolute temperatures having a constant ratio (p).The logarithmic temperature-coefficient 11 would then be given by the equations 7 = KpT/KT 7 = Kp271'/KpT I.GENERAL AND PHYSICAL CHEMISTRY. ii. 125 . . . . q=KpnP/Kpn-lT. The product of all these equations gives qn= KpnTlKI' or if KpnT= Kz KT= Kl pnT- T2 and 5!'= Tl pn= T2/Tl or n = (log T2 - log T,)/log p. The relation between this logarithmic coefficient cofficient y is expressed by the equation Hence 7 = yKKp"T/KT= (K2/Kl) log ppog T - log TI. and the ordinary logq = w . - T,-*1- . log y. 10 log TI2 - Jog T'l Taking p to be 1.1 that is an increase of 10% in the abcolute temperature the values yI 2 and y2 = 3 for the ordinary coefficient at ordinary temperatures give yl = 7 *23 and qz = 23.02 respectively. On the assumption that the latter magnitudes remain constant the corresponding values for y a t different absolute temperatures will be as follows T.q1 7.23. qa 23.02. 100 y,=8'00 y2 = 27 *02 150 3 '98 8.93 200 2 -84 5.22 300 2.00 3 .oo 1000 1-23 1 *38 These numbers are in agreement with the general variation of the ordinary temperature-coefficient which diminishes with rise and rapidly increases with fall of the temperature. I n the particular case of the decomposition of hydrogen iodide the values of y calculated by means of the above equation agree within the limits of experimental error with the observed values ; thus for the temperature-intervals 300-400° 400-500° and 500-600" Bodenstein (A.1899 ii 637) found 1.89 1-64 and 1.53 respectively the calculated values being 1.94 1.64 and 1.49. The Relation between Oxidation Potential and Oxidation Velocity JULIUS G R ~ H (Zditsch. physikal. Chem. 1913 81,695-712. Compare Bodenstein A 1904 ii 717 ; Bogrim A. 1910 ii 282).- The oxidation of acetaldehyde and formic acid was effected by means of chlorine and the velocity constants determined. Considerable diffi- culty was experienced owing t o the reactions being complicated through the action of the chlorine on water ; however by effecting the changes in t h e presence of a known concentration of nitric acid it became possible to bring the secondary reaction into the calculation and obtain a moderately good velocity constant.Other experiments namely the action of iodine on acetaldehyde the oxidation of aldebyde by cerium ammonium nitrate and the oxidation of chloral and bromal with bromine were tried but were found t o be too complicated to furnish results. The velocity of oxidation by T. H. P. chlorine i n the two cases investigated when results for bromine (Zoc. cit.) show that they oxidation potentials of chlorine and bromine Acetaldehyde. Er. c1 Velocity constants ...... 1'205 0'648 Oxidation potelltial ... 1.334 1-639 compared with Bognaris are j u s t opposite to the thus Formic acid. - Br. el. 3280 431 1-334 1.639ii. 126 ABSTRACTS OF CHEMICAL PAPERS. This anomaly can be explained in two ways (1) The chemical resistance is greater in the case of chlorine than in that of bromine.(2) That the course of the reaction is different in the two cases. The first reason is held by the author to be unlikely the second case becomes likely if it is supposed that the halogen forms a n inter- mediate compound with the reducing substance which breaks iip with greater or lesser velocity to form the end products of the reaction. Attempts were made by spectrophotometric measurements t o confirm the second conclusion but they led to no definite results. J. F. S. Velocity of Reaction of Different Bases with Halogen- substituted Acids. 11. HJ. JOHANSSON (Zeitsch. physikal. Chm. 1913 81 573-589. Compare A. 1912,ii 544; Holmberg A. 1912 ii 443; Senter T. 1912 91 460).-The velocity of reaction of the decomposition of monobromosuccinic acid is investigated in neutral and alkaline solution and aloo in the presence of neutral salts.It is shown that two reactions take place t h e first unimolecular in which the hydroxyl ion reacts with the ion of monobromosuccinic acid with the production of the ion of fumaric acid and bromine ion thus ‘O*CO*CHBr*CH,*CO*O’ + OH’= ’O*CO*CH:CH-CO.0‘ + Br’ + H,O and the second a bimolecular reaction in which the ion of monobrowo- succinic acid forms first the ion of propiolactonecarboxylic acid and bromine ion thus ’O*CO*CHBr*CH,-CO*O‘ = ’O*CO-$!H-YH + Br’ and the lactone ion then reacts with the hydroxyl ion giving the ion of malic acid thus 0-co 0--co ’O*CO.VH.YH + OH’ = ‘O*CO*CH(OH)*CH,*CO*O’. The reaction constants of both reactions are determined the former being determined in neutral solution using the sodium potassium barium and strontium salts of the acid.The reaction constant is found to be k=0*002403 and is independent of the nature of the metal ion. The addition of neutral salts causes the constant to increase slightly but this is held t o be due to the superimposing effect of the second reaction. The second reaction constant wits determined in alkaline solution using the hydroxides of sodium potassium barium strontium both with arid without the addition of the nitrate corre- sponding with the hydroxide. This reaction is shown to depend on the nature and concentration of the cation and follows the empirical rule of Holmberg (Zoc. c i t . ) C,=C[M.]a. I t is found that d=&C,= 0.095 C = 0.093 ; Cha = 0,197 and C = 0.196.J. F. S. Theory of EfBorescence. Influence of the Size of the Crystal. CH. BOULANOER and GEOEGES URBAIN (Compt. rend. 1912 155 1612-1 6 14. Compare this vol. ii 34).-Starting from their law log (m - mt) = logcc + A log (0 - t ) the authors deduce the equa- tion log a’ = log a + (1 - A/3) log p‘/p for the relationship between the efflorescence of two crystals of weights p’ and p respectively.GENERAL AND PEYSICAL CHEMISTRY. ii. 127 They have applied this to two crysta.ls of sodium sulphate and consider that the values found and calculated are in agreement within the limits of experimental error. W. G. Kinetics of Chemical Reactions of Combination Deoxida- tion and Oxidation. 11. E. I. ORLOV (J.Rzcss. Phys. Chem. Soc. 19 12 44 1576-1597).-Exarnination of Spitalsky’s results dealing with thz catalytic decomposition of bydrogen peroxide (A 19 11 ii 36) in the light of the considerations previously advanced by the author (A. 1912 ii 243) leads to the following conclusions. The cat,alytic decomposition of hydrogen peroxide by chromic acid furnishes an example of the transformation of a unimoleculnr reaction of the first order into a unimolecular one of the second order. This transformation is conditioned by the action of an intermediate form of the oxide formed during the reaction and entering the sphere of the catalysis only after the peroxide is docomposed to a certain extent. The influence of this intermediate form is complicated by the fact that one form of catalysis is converted into the other not suddenly but gradually; when the intermediate form comes into action the coefficient f appears in the differential equation dx/dt = k( B -fx)(B +fx).This appearance o f f is dependent on the consumption of a certain proportion (one-fifth to one-eighth part of the original amount remains) of the hydrogen peroxide and the consequent difficulty of oxidation of Cr’” to C”; a more rapid process of oxidising Cr20,(0,H,) to Cr,07H is then initiated Cr(OH) being formed as an intermediate product in the change. I t is hence necessary to assume in t h e kinetics of chemical processes the principle of least loss of time; chemical processes strive to take place in such a way that there occurs the least waste of time. The reducing properties of hydrogen peroxide are explained as due to the quadrivalency of oxygen in the molecules of the peroxide and of water.The oxidation of the hydrogen ions is regarded as a com- bination with the molecule H-OiO-H thus H*OiO*H + 2H’ = H . H,:o:O:H - 2 >Oc::.:.. The first stage of this reaction requires 2 - H time but the second proceeds instantaneously. This hypothesis of the quadrivalency of oxygen necessitates the O:H awumptions that aqueous solutions contain the complex H,:O< O:H possessing reducing properties and that the ions Cr,O,” and the molecules H,O give an intermediate oxide of the type - Cr,O,[_~~>OH,] 4 or This oxide is the oxygen-carrier and takes part in the conversion of CrV1 into CrI’I and of the latter into Crvl again. After combina- tion of the hydrogen ions with hydrogen peroxide the remaining groups (Cr,07)d:: with free affinities also combine with H*O:O*H,ii.128 ABSTRACTS OF CHEMICAT PAPERS. giving H*O=O*H which is highly unstable under the conditions of Spitalski’s experiments (Eoc. cit.) and decomposes into Cr2y 2H’ and 0 but is ~omparat~ively stable under the ordinary conditions of room-temperature and absence of vigorous shaking. I f the catalytic decomposition of hydrogen peroxide by means of potassium dichromate proceeds under ordinary conditions the stable compounds \,/ ant3 I are obtained and the decomposition follows the differential equation dx/dt = k(A - x ) ( B + x). This equation is applicable t o reactions of combination. The form dx/dt = Ic(A - z) represents a particular case.K i n e t i c s of Chemical Reactions of Combination Deoxida- tion and Oxidation. E. I. ORLOV (J. RUS. Phys. Chem. SOC. 1912 44 1598-1623 1633-1658).-The considerations advanced in .previous papers (see preceding abstract) are applied t o other reactions. The decomposition of hydrogen peroxide by an iodide in aqueous solution is shown t o be a unimolecular reaction of the second order decomposition by means of potassium dichromate or molecular platinum proceeding siniilarly. ‘V Cr,O H*0=0- H H.O-- O*H 1i CrO H d r 0 ’ ( 3 - 2 0 7 T. H. P. 111 and IV. The mechanism of these changes is discussed. T. H. P. Universal Significance of the EleDi entary Quantum. OTTO SACRUR ( A m . Physik 1913 [iv] 40 67-86. Compare A 1912 ii 145 1151).-By a process of reasoning similar to that adopted in the previous papers equations have been deduced for the energy and entropy of ideal monatomic solid substances and of monatomic gases The reasoning is based on a more precise definition of the (physical) conception of probability and not on the usual assumption of elemen- tary energy quanta.The only quantities occurring in the equations are certtain general constant8 and also the atomic vibration frequency in the case of the solids and the molecular weight in the case of the gases. The values of the “chemical constants,” which determine the chemical behaviour of the gases and the vapour pressures of their condensation products are calculated for helium neon argon krypton xenon aod mercury I n atmospheres this constant C is given by the equation C = - 2.055 + 1.5 log M where M is the molecular weight of tbe gas.It is shown that the calculated vapour pressures of mercury between 0” and 360’ are in fairly good agreement with the observed values when the “constant” for mercury given by this equation is applied in the calculation of the vapour pressure curve. The calculat,ed and observed vapour pressures of argon a t 84’ nbs. are also concordant H. M. D. The ‘‘ Chemical C o n s t a n t s ” of Di- and Tri-atomic Gases. OTTO SACKUR (Ann. PI~ysik 1913 [iv] 40 87-106. Compare preceding nbstract),-On the assumption that the di- and tri-atomic gases haveGENERAL AND PHYSICAL CHEhlIS’L‘RY. ii. 129 the structure assigned to them by Boltzmann formuh are deduced from which the entropy and the ‘‘ chemical constants ” of these gases may be calculated.The values obtained for hydrogen oxygen nitrogen the halogen and halogen acids carbon monoxide nitric oxide water hydrogen sulphide carbon dioxide and sulphur dioxide are recorded. The “ chemical constants ” are applied in the calculation of the vapour-pressure curves of iodine and ice and of the dissociation constants corresponding with the equilibria 2HCl ZZ H + CI 2HBr Hz+Br 2HI Z H,+I BH,O ZZ 2H,+0 2CO,= 2UO + 0 and 2NO From a comparison of the calculated vapour pressures and dissociation constants with experimental data i t is found that the values obtained for the “ chemical constants ” may be regarded in most cases as approximately .correct. The existence of considerable discrepancies between theory and experiment which is found in certain cases for example the dissociation of carbon dioxide and nitric oxide indicates however that the values of the ‘‘ constants ” cannot be regarded as final and that i t will probably be necessary to modify the theory of poiyatornic molecules before the true values can be derived.H. 11. D. N + 0,. Volume and Valency. 31. SEBALDT (Zeiisch. ph ysikal. Clmtz. 1 9 1 3 81 749-753).-8 theoretical paper in which the relationship between valency arid atomic volume and also other properties of the elements is shown to be periodic. In one diagram the logarithms of the atomic volumes are plotted as ordinates the elements being placed on lines parallel to the ordinate axis which represent the eight groups of the periodic system.This arrangement shows clearly the relationship between the electroaffinity and the volume ; atomic magnetism and compressibility are also shown to be functions of the atomic volume. I n another arrangement the elements are placed round a polar co-ordinate ring in order of atomic weight those o€ even valency to the left of the ordinate axis and those of odd valency to the right ; the logarithm of the atomic volume is plotted on the ordinate. This arrangement brings out most of the relationships betweeu the elements very clearly. J. F. S. The Nature of A u x i l i a r y Valencies. 11. Metal Ammonias. FRITZ EPHRAIX (Zeitsch. physikcd. Chern. 1913,19,513-538. Compare A. 1912 ii 546).-The present paper contains further details on ammonia derivatives of the metal salts.It is shown that errors are likely to come into the prevsure measurements of ammine derivatives owing to incomplete drying and absorption of air by the finely divided compounds. Precautions for avoiding these errors are indicatad. The gaseous pressure at a series of temperatures is determined for the hexammine derivatives of the halogen salts of cadmium zinc man- ganese nickel cobalt iron copper calcium and magnesium. It i s found that the ratio lLl/T1 is a constant for a given pair of salts at all pressures where 5” is t h e temperature a t which the two compounds have the same gaseous pressure. Compounds of this type therefore obey the kamsay-Young rule. The value of the factor varies from 0.9138 for Zn12/ZnBr2 to 1.232 for MnI,/AlnCl,. The temperature a tii.130 ABSTRACTS OF CHEMICAL PAPERS. which the dissociation pressure of the various salts is 500 mm. is given and i t is seen t h a t t h e pressure generally decreases as the atomic volume of the central atom increases. The hexammines of zinc iodide cupric bromide and those of the chloride bromide and iodide of cadmium being exceptional. The value 3E is shown to be constant for all the hexammines except those mentioned above and in this con- nexion i t is shown that taking the mean value of JpT= 14.0 it is possible to calculate the atomic volume of the central atom with fair approximation when T is the absolute temperature a t which the dis- sociation pressure is 500 mui. It is found that when the absolute temperatures for pressures of 700 mm. and 200 mm. are taken the ratio is approximately constant (1.085-1.073). Moduli are calculated for the pressure-temperature relationship of I C1 I Br Br C1 and from these it is seen that generally a t a given temperature the tension is greatest for the chlorides and least for the iodides. From calculations from the Nernst equation logp = - Q/(4,571Y’) + 1.75 log T’+ 3.3 i t is shown that the dissociation of the hexammines of the bivalent metals occurs by first splitting off one molecule of ammonia R(X),GNH R(X),5NH3 + NH,. The heats of formation of the various ammines are calculated both from the Nernst formula and the van’t Hoff formula. A series of notes on the preparation of the ammines used in the investigation is given. J. F. S. The Nature of Auxiliary Valencies. 111. The Region of the Existence of Auxiliary Valency Compounds. FRITZ EPHRAIM (Zeitsch. physikul Chenz. 1913 81 539-542. Compare preceding abstract).-The dissociation of the hydrates and ammines is theoretically considered. It is shown that whilst the ammines dissociate by success- ively aplitting off one molecule of ammonia this is not analogous to the hydrates. The reason is the formation of bolutions in the case of the hydrates. I t is concluded that all the possible hydrates exist if only in a labile condition and in solution. It is also shown why the dissociation curves of the ammines show no transition points whilst those of the hydrate do. What are Bases and Acids? DAKIEL VORLBNDER (J. p r . Chem. 19 13 [ii] 8’7 84-91).-Theoretical. New Shaking Apparatus. WILHELM STEINKOPF and HANS WINTERNITZ (Chem. Zeit. 19 13 3’7 4O).-The apparatus resembles a retort stand where the upright is fitted by means of ball-bearings into a heavy cast-iron base. A short horizontal arm is attached to the upright and connected with the ecceotric of a small motor. The usual retort stand clamps may be used for holding water- and hot-air baths and flasks for the distillation of viscous llquids which other- wise have a tendency to bump. The apparatus may also be used for holding burettes for the titration of boiling liquids. On account of the gentle rotating movement imparted t o the contents of the vessels it is not necessary to use stoppers and hence any risk of introducing impurities is avoided. J. F. S. F. B. H. EL H.INORGANIC CHEMISTRY. ii. 131 Shaking Apparatus which can be Exhausted Fitted with an Inner Temperature Regulator. RICHARD KEMPF (Chem. Zeit. 19 13 37 58-59).-A modification of the apparatus previously described (A 1906 ii 433) the principle remaining the same. The method by which the apparatus may be used in invest,igating catalytic reductions with colloidal or finely divided platinum is described. T. S. P. A Simple Experiment Illustrating the Luminosity of Phos- phorus. DOUGLAS F. Twrss (Chem. Jaws 1913 107 16).-A vertical glass tube 2-2& cm. internal diameter and about 120 em. long. is fitted at the lower end with an indiarubber bung carrying a glass tube which is bent upwards so as to be parallel to and of approxi- mately the same height as the wider tube. A solution of phosphorus in olive oil is introduced into the wider tube so as to reach about 6 inches from the top and steady suction is applied at the mouth of this tube by means of a water pump. Air enters through the narrow tube and a beautiful series of bell-shaped phosphorescent air ttubl)les rises through the column of oil. T. s. P.