ii. 109 General and Physical Chemistry. Spectrochemistry of Unsaturated Compounds. KARL AUWERS (Bev. 191 1 44 3679-3692. Compare Eisenlohr this vol. ii 2).-In part polemical. Eisenlohr has failed to acknowledge sufficiently the work of Briihl on the influence of conjugation 011 optical properties. Exception is taken to the statement that the introduction of several disturbing groups turn8 an optical exaltation into a depression which is claimed to be especially the case in crossed conjugations ; the experimental evidence is not considered suficient to justify this view. The optical effect when a simple conjugation is converted into a crossed conjugation depends on two factors the depression produced by the influence of the entering group on the conjugation and the exaltation due to the new double bond.The final result depends on the magnitude of each of the opposing factors. I n the case of crossed con jugations with partial valencies the new valencies compensate t o some extent for the disturbing influence of the substituent. The original must be consulted for the further criticisms of Eisenlohr’s argument. E. F. A. Refractive Indices of Mixed Liquid Crystals. PAUL GAUBERT (Con@. rend. 1911 153 11\58-1160. Compare Abstr. 1911 ii 949),-The refractive indices of optically positive mixed liquid crystals follow fairly closely the ordinary ,law of mixtures fur isomorphous solids. p-Azoxyphenetole and p-azoxyanisole influence in the same way the values of the indices for cholesterol octoate. Anomalous results with optically negative crystals are explained by supposing that these two substances exist in a hitherto unknown negative phase and that crystals of the latter form mixtures with negative crystals of cholesterol salts.w. 0. w. Heterogeneous S t r u c t u r e of ‘‘ Fluid Crystals ” of Par- azoxyphenetole. HELENE DEISCHA (Zeitsch. Krt~st. Man. 191 1 50 24-32).-The pecularities of structure and the birefringence shown by fused parazoxyphenetole under the microsLope are regarded as due not to any crystalline structure but to the capillary attractions and tensions in a colloidal solution forming a frothy mass or emulsion. L. J. S. [Spectroscopy of Oxygen.] WALTER STEUBING (Ann. Phy~ik 191 1 [iv] 36 1077-1080).-Polernical against Kayser (Abstr. 1911 ii 785). H. &I. D. C. PORLEZZA (Atti K.Accad. Lincei 1911 [v] 20 ii 584-58’7).-l’he paper gives the results of the author’s measurements of the line spectrum of nitrogen from the red to the green the nitrogen being enclosed in Geissler tubes at a pressure of about 100 mm. Line Spectrum of Nitrogen in a Geissler Tube. R. V. S. VOL. c11. ii. 8ii. 110 ABSTRACTS OF CHEMICAL PAPERS. Line Spectrum of Nitrogen in a Geissler Tube. C. YORLEZZA (Atti R. Accad. Lincei 1911 [v] 20 ii 642-645. Compare preceding abstract).-The author concludes his tables of wave-lengths for this spectrum giving here those of the blue and ultra-violet regions. R. V. S. Spectrum of Boron. SIR WILLIAM CROOKES (Proc. Roy. Soc. 1911 A 86 36-41).-The spectrum of the spark passing between electrodes of crystalline boron has been examined.Certain lines on the photographs have been traced to calcium and aluminium and wheu these are eliminated it is found that the ultra-violet spectrum of boron consists of three lines of wave-lengths X = 3451.50 2497.83 and 2496.89. The fourteen other lines given by Eder and Valenta and the five other lines recorded by Exner and Haschek could not be detected on any of the photographs in spite of excessively long exposures. H. M. D. Spectrum of the Magnesium High-frequency Arc. ELLEN O’CONNOR (Phil. itfag. 1912 [vi] 23 94-101).-The spectrum obtained from an arc between magnesium electrodes has been examined with a high-frequency oscillatory discharge consisting of undamped oscillations. Photographs have been obtained which show the influence of the inductance capacity and frequency on the nature of the spectrum.A comparison of these indicates that the spectrum becomes richer in lines as the frequency increases and the continuous background which is seen in the case of the low-frequency discharge gradually disappears as the frequency is increased. H. M. D. Measurements of the Wave-lengths of Normal Lines in the Iron Spectrum. P. EVERSHEIM (Ann. Physik 1911 [iv] 36 1071-1076. Compare ibid. 1909 [iv] 30 815).-Further measure- ments of wave-lengths of iron lines have been made between X = 4482 and X = 3371. The data are in excellent agreement with the correspond- ing measurements made by Fabry and Buisson the maximum differ- ence amounting to 0 4 0 3 Angstrom units. H. Ilil. D. Influence of the Solvent on the Position of Absorption Bands in Solutions. T.H. HAVELOCK (Proc. Roy. Soc. 1911 A 86 15-20).-,41i attempt has been made to measure the effects which can be ascribed to the operation of Kundt’s rule in referenw to the displacement of the absorption bands of dissolved substances towards the longer wave-lengths in virtue of the refractive or dispersive power of the solvenf. H. M. D. Absorption Spectra of Phosphorescent Substances. BERN- HARD WALTER (Physikal. Zeitsch. 191 2 13 6-1 I).-Photographs have been obtained of the absorption spectra of the phosphorescent alkaline earth metal sulphides containing bismuth as ‘‘ active impurity.” The spectra are characterised by strong absorption bands and i t has been found that only those lines which are absorbed are capable of producing a well developed phosphorescence effect.GENERAL AND PHYSICAL CHEMISTRY.ii. 112 By reference to Lenard's obsorvstione it is showu that tlie wave- lengths of those rays which excite the phosphorcscence of the alkaline earth metal sulphicles are nearly independent of the nature of the active metallic impurity but vary considerably according to the particular alkaline earth metal present. This cannot be reconciled with Lenard's electronic explanation of phosphorescence but appears to favour the older chemical theory. H. M. D. The Fluorescence of Anthracene. LOUISA STONE STEVENSON (J. Physical Chem. 1911 15 845-865).-From :t chemical point of view the fluorescence of anthracene may be due to the following circumstances (1) A continuous reaction generating i1 small amounf of light the intensity or wave-length of which is sufficiently altered by incident light to render the emitted light invisible.(2) 'l'automerio change incident light beiug absorbed aud fluorescent light emitted during the formation of one or both isodynamic forms. (3) Dissocia- tion under the influence of light fluorescent light being generated either in the dissociation or i n the re-association of the molecule. A number of highly fluorescent substances in solution failed t o produce any effect on the photographic plate under the most favourable conditioris. Heme the conclusion is drawn that explanation (1) is improbable although the plate was also unaffected by the gradual dehydration of quinine sulphate in which it is supposed light of very low intensity is emitted.The tautomeric hypothesis (2) in the form suggested by Hewitt is discredited by the fluorescence of dihydroanthracene which has no para- bond. A ,consideration of all the known anthracene derivatives shows however t h a t the meso-ring is undoubtedly concerned with the fluorescence. Anthraquinone and its derivatives are not fluorescent but oxanthrol derivatives may exhibit fluorescence. A tautomerism due to the oscillation of an hydrogen atom or equivalent group between the y-position and some other part of the molecule is considered improbable in the case of dichloroanthracene tetrachloride where all the immediately available positions for chlorine atoms are filled. Hypothesis (3) cau be applied to all the known fluorescent derivatives of anthracene.The author finds t h a t the chlorination of molten anthracene to y-dichloroanthracene is accompanied by the emission of a bluish-white light. The marked fluorescence of y-dichloro- anthracene is well explained by the dissociation of the compound with absorption of light and its re-formation with emission of light of a different wave-length. Tesla rays and P-rays from radium excite antbracene dichloroanthracene and dihydroanthracene to luminescence whilst even anthraquinone may be rendered fluorescent in a powerful cathode beam. The formation of dianthracene under the influence of light probably has no connexion with the fluorescence of anthracene. Anthracene derivatives which have no para-bond and which cannot polymerise to dianthracenes are often fluorescent.Light has no appreciable influeuce on the oxidation of arithracene a t 140-150" by gaseous oxygen. The cfiloriiiatiori of anthracene in the dark yields dichloroanthra- 8-2ii. 112 ABSTRACTS OF CHEMICAL PAPERS. cene. produced. product is hexuchloroanthritcene tetrachloride. obtained in the dark with red phosphorus as catalytic agent. I n presence of iron filings dichloroanthracene tetrachloride is I n direct sunlight with or without halogen carrier the The same product is R. J. C . CkgStalliUe Liquids. H. VON WARTENBEKG (Physikd. Zeitsch. 19 1 1 12 1230. Compare Abstr. 1911 ii 952).-The fact that crystalline liquids behave like uniaxial crystals when examined in an electrical fieid at right angles t o the linesof force has also been demonstrated by Mauguin (Compt.Tend. 1911 152 1680). H. M. D. Chemical Action of Ultra-violet Rays Synthesis and Decomposition of Water. IWAN I. ANDHEEFF (J Ru88. Phys. Chem. Xoc. 1911 43 1342-1364).-After discussing recent work on the chemical effects of ultra-violet light the author debcribes his own experiments which have been carried out with a mercury lamp at various voltages and currents. With a voltage of 220 hydrogen and oxygen combine to form water with moderate rapidity. The velocity of the reaction which terminates in from five to twenty hours is independent of the concentrations of the reacting substances and is in good agreement with the expression dxldt = C ; it increases almost proportionately with the intensity of the light. The mean value of the temperature-coeflicient of the reaction is 1.1.Also water decomposes under the influence of ultra-violet light in the field of action of which an equilibrium is estitblislied between hydrogen oxygen and water this equilibrium differing from the thermal equilibrium at the corresponding temperature. When the initial product is a mixture of hydrogen (2 vols.) and oxygen ( 1 vol.) the final position of equilibrium is the same as when water is employed. With increase of the intensity of the ultra-violet light the equi- librium is displaced in the direction of the dissociation and the percentage of water decomposed increases proportionately with the increase of intensity. T. H. P. Action of Ultra-violet Light on Sodium Thiosulphate. LOUIS MARMIER (Compt. rend. 19 12 154 32-33).-Af ter exposing a solution of sodium thiosulphate to the rays from a quartz-mercury lamp for five minutes sulphur is deposited and the liquid found to contain sodium hyposulphite.The latter however is completely destroyed after a further exposure of seventy minutes. The formation of a hyposulphite can only be detected in solutions containing less than w. 0. w. HANS EULEH (Arkiv. K e m . Nin. Geol. 1911 4 No. 8 l-lO).-The degradation of lactic acid to carbon dioxide and alcohol takes place in pure aqueous solution under the influence of ultra-violet radiations. The rate of decomposition of the lactic acid does not depand on the concentration of the acid within 0.6% of thiosulphate. Biochemical Reactions in Light.GENERAL AND PHYSICAL CEEMISTRY. ii. 113 the limits studied 095N to N.The condensation of acetaldehyde t o aldol also takes place without a catalyst on exposure to ultra-violet light. The bearing of these results on the degradation and synthesis of plant acids is discussed. E. F. A. The a-Particles Emitted by the A c t i v e Deposits of Thorium and Actinium. E. MARSDEN and T. BARRATT (proc. Physical L!OC. 1911 24 i 50-61. Compare this vol. ii 6).-Probabilitp con- siderations have been applied to the a-particles emitted by the C-members of the thorium and actinium active deposits. For a single disintegration the probability of occurrence of a time interval of length between t and t+8t is p ~ - ~ t S t where the a-particles are emitted at the mean rate of p per second. If more than one a-particle is emitted simultaneously or if a product of very short period exists there will be an excess of short intervals above t h a t calculated from the formula.The method of two observers simultnneonsly observing two zinc sulphide screens between which the radioactive substance was placed or in the case of an emanation diffused was adopted each observer recording the observations seen. For the actinium active deposit the agreement between the observed and calculated intervals was good the short intervals being only about 1% of the whole in excess. With the thorium active deposit good agreement was also obtained and as before only a slight excess in this case about 2% of the whole of short intervals occurred. It is possible that this may be explained by imperfections in the experimental arrangement.A determination of the number of scintillations from thorium-C of ranges 4% and 8.6 cm. respectively showed that the number of t h e longer range is 0.646 of the total so that thorium-C2 gives 1.83 times the number of a-particles given by thorium-C,. This ratio held unaltered as the active deposit decayed. The period of the change in which the longer range a-particle is expelled calculated from the range is only second. The distribution of the time intervals shows that if thorium-C bucceeds thorium-C1 its period should be greater than a few seconds and it should be separable by recoil methods which however fail. The conclusion is drawn that thorium-C and -C2 must be simul- taneous products of thorium-B. However there is t h e possibility that the ratio in the number of a-particles given by the two C- members should be as 2 1 but the assumptions necessary in this case to avoid the introduction of branch series seem out of the question. The conclusion of Geiger and Kovarik (Abstr.1911 ii 954) that thorium-D gives two P-particles per atom disintegrating depends on the assumption that thorium-C1 and -C2 give two a-particles per atom disintegrating and is therefore no longer justified. F. S. The /3-Rays of the Radium Family. J. DANYSZ (Compt. rend. 1911 153 1066-1068. Compare Abstr. 1911 ii 840; Baeyer Hahn and Meitner ibid. 567).-The speeds of the separate bundles of P-rays from a tube containing the emanation have been determined with the greatest possible accuracy and many more have been recog-ii.114 ARS'I'RACTS OF CHEMICAT PAPERS. nised. The table gives in the first line the number of the bundle in the second its velocity in terms of that of light calculated according to the Lorentz formula and in the third the relative strength of the bundle ( 8 = strong nz = medium f= feeble v.f = very feeble). 1. 2. 3. 4. 5 . 6. 7. 8. 0.615 0'634 0.660 0'682 0.705 0.718 0-735 0.748 *< . f. 1;. f. 9 f. V f . S . m. 9. 10. 11. 12. 13. 14. 15. 16. 0.i60 0786 0.790 0'862 0'882 0.897 0'920 0.940 f. Y. f. Y. f. f. f. S. 17. 18. 19. 20. 21. 22. 23. 0.943 0.946 0'949 0.957 0,962 0.988 0'996 - - f. .f. In. S. f. The seven strong bundles apparently correspond with those before measured the velocities being more accurately determined. From Nos. 1 to 9 the bundles appear tlo consist of rays travelling a t veloci- ties differing by about 1% from the mean value on either side.The more rapid bundles are perfectly sharply defined. Bundle No. 22 consists of from three to five separate bundles travelling with velocities between 0.98 and 0.99. These extremely fast @-rays like the 7-rays only slightly affect tho photographic plate. The last bundle No. 23 is for this reason very little known. The value given for the velocity 0.996 which corresponds with a value of R p (H= strength of magnetic field p = radius of curvature) of 18,100 com- prises certainly some rays with a value of H 26,000 and of the velocity 0.998. The emanation from 0.3 gram of radium chloride was employed. F. S. Radiant Emission from the Spark. ROBERT W. WOOD (Physikal.Zeitsch 1912 13 32-34. Compare Abstr. 1910 ii 915). -Polemical against Steubing (Abstr. 19 11 ii 838). Further photo- graphs are reproduced which confirm the view that the spectrum under discussion is due to a spectral radiant emission and not to a scattering of the light from the spark discharge by condensed particles of the vapour of the metal as maintained by Steubing. Charges on Thermions Produced in Air and Hydrogen at Atmospheric Pressure. J. (2. POMEROY (Phil. A v c q . 1922 [vi] 23 l'i3-182).-The magnitude of the charge carried by the ions emitted by a strip of heated platinum has been examined by observa- tions on the distribution of the ions when subjected to the action of a uniform electric field. At x temperature which is only just sufliciently high to give positive ions some of the ions from a fresh strip carry charges which are double the atomic charge.When however the strip is heated for some time a gradual change takes place and as the poRitive leak diminishes the proportion of ions carrying a double charge gets smaller and smaller until a steady state appears to be i.eachetl when the average charge is nearly eqaal to the atomic charge. No such change in the average charge occurs in the case of the H. M. D.GENERAL AND PHYSICAL CHEMISTRY. ii. 115 negative ions and at temperatures which are sufficiently high to give negative ions it is found that the average charge of the positive and negative ions is about the same. This result suggested that the negative ions might also be found to carry single charges only if obtained from a strip at a sufficiently low temperature and experiments in hydrogen showed that this is actually the case.H. M. D. Ionisation Produced by Carbon at High Temperatures. JOHN N. PRING and A. PARKER (Phil. Mag. 1912 [vi] 23 192-200). -The emission of ions by carbon has been investigated at tempera- tures between 1300" and 2050" and at different pressures. The negative leak at a given temperature diminishes with the progressive purification of the carbon and with reduction in the pressure. ampere per square centimetre of carbon surface whereas Richardson found 1.3 x 1 0 - 4 at this temperature. Above 1800" the increase in the ionisation with the temperature is very small and the lowest value of the current observed at 2000O was 4 x lo-' ampere.The data obtained in these experiments are incompatible with Richardson's formula and indicate that the large ionisation currents hitherto obtained with heated carbon cannot be ascribed to the emission of electrons from the carbon itself. They are probably due to some interaction between the carbon or the impurities contained in it and the surrounding gases which involves the emission of electrons. The smallest ionisation current cbtained at 1300" was 1.6 x H. M. D. The Emission of Positive Charges by Heated Metals. G. REBOUL and E. GR~GOIRE DE BOLLEMONT (Le Radium 1911 8 406-41 l).-Sur€aces of copper and platinum separated to defiuite distances from one another are heated to various temperatures under various conditions. If the copper surface is in the shape of a cross the deposit on the platinum assumes the same form.The thickness of the deposit increases to a maximum and then decrease8 as the time of heating is increased so that after a certain time no deposit is left. The maximum deposit results in a time the shorter the higher the temperature. The deposit is most abundant in oxygen but occurs and has a grey colour turning black on heating in air in nitrogen and carbon dioxide. The nature of the surface receiving the deposit is not of great import'ance and the platinum may be replaced by other metals and even by porcelain. I n addition to copper a projection of metallic particles occurs with silver as may be put into evidence by opposing it with a gold plate which becomes coated with a white deposit of silver of the form of the silver surface.The maximum in the amount of deposit with copper is explained by the '' fatigue" of the copper. All metallic surfaces are regarded as projecting metallic particles when heated due to the liberation of occluded gas and the mechanical carrying away of the metal by the bursting of the bubbles a process which accounts also for the positive ionisation from heated metals which in every respect is analogous to the phenomena investigated. A t first the projection from the copper surface to the opposed surface is regarded as greater than that of the projection of particles from this A black deposit forms on the platinum.ii. 116 ABSTRACTS OF CHEMICAL PAPERS. surface but afterwards when the gas has been liberated ‘‘ fatigue ” of the copper takes place and the deposit on the opposed surface is again removed by the projection of particles from that surface.F. S. Positive Thermions from the Salts of the Alkaline Earths. CLINTON J. DAVISSON (Phil. Mag. 1912 [vi] 23 121-139).-The nature of the ions emitted a t high temperatures and low pressures by the chlorides fluorides sulphates and phosphates of barium strontium calcium and magnesium has been investigated by the method employed by Richardson. From observations on the influence of a known magnetic field it is found that the truly characteristic thermions emitted by the salts of these metals consist of single atoms of the elements carrying-utiit positive charges. Experiments with zinc sulphate lead to the same conclusion. The R61e Played by Gases in the Emission of Positive Thermions from Salts.CLINTON J. DAVISSON (Phil. Mag. 1912 [vi] 23 139- 147. Compare preceding abstract).-The emission of positive ions by heated strontium chloride strontium sulphate and aluminium phosphate has been examined with reference to the part played by gases in the phenomenon. The results indicate that the increased emission of thermions which is observed when the heated substance is in contact with air carbon dioxide or hydrogen is not due to the emission of molecules of the gas which have been absorbed by the salt and emitted in a charged condition. I n reference to the suggestion that the traces of carbon monoxide or dioxide which are sometimes emitted by heated salts are responsible for the observed ionisation it has been found that the emitted gas is not electrically charged.H. M. T). The Production of Chemically Active Rays in Chemical Reactions. J. MATUSCHEK and NENNING (Chern. Zeit. 1912 38 2l).-A beaker containing zinc and sulphuric acid is placed on a photographic plate in a light-tight case a tinfoil star being int’er- posed. After several hours’ exposure a distinct image is obtained on development. The best results are obtained when the zinc is in the form of a sheet with ribs on the lower surface. Similar results are obtained with hydrochloric and nitric acids and with copper tin and lead the intensity of the photographic action varying considerably. Copper oxide and hydroxide and potassium hydroxide also give chemically active rays when dissolved in acids. Other active reactions are the slaking of lime the setting of Portland cement and plaster the decomposition of calcium carbide by water and the formation of ammonium amalgam.The most active reaction observed is t h a t between sodium silicate and dilute hydrochloric acid which gives a distinct image in less than an hour. The Detection of Actinium Emanation in Minerals containing Actinium. GEORG VON HEVESY (Physihd. Zeitsch. 191 1 12 1213-1214).-An investigation has been made of the point H. M. D. C. H. D.GENERAL AND PHYSICAL CHEMISTRY. ii. 117 noticed by Boltwood (Abstr. 1908 ii 454) that actinium emanation cannot be detected in a uranium mineral by passing air through its solution into a n electroscope. It was found that by the use of vessels of sufficiently small capacity and by filling these as com- pletely as possible with the solution or drying liquid employed the actinium emanation could be readily detected from a solution of pitchblende.One gram of pitchblende in 2.5 C.C. of solution contained in a washing-bottle of 3.5 C.C. capacity gave when an air stream of 0.4 C.C. per second was passed through a leak in the electroscope due t o actinium emanation of two divisions a minute. F. S. The Solubility of A c t i n i u m Emanation in Liquids and Charcoal. GEORG VON HEVESY (Physikal. Zeitsch. 19 1 1 12 1214-1224).-By means of the differences in the ionisations pro- duced in an electroscope wheu the actinium emanation is passed through various liquids at various rates on its way to the electroscope the solubility of the emanation has been indirectly calculated in various liquids.These show the same relative absorbing powers as in the case of the other emanations. In increasing order of absorbing power the liquids examined were saturated aqueous potassium chloride solution water sulphuric acid ethyl alcohol amyl alcohol acetone benzaldehyde benzene toluene petroleum and carbon disulphide. The solubility coeficient of actinium emanation in water is 2 as compared with 1 for the thorium and 0.33 for the radium emanation. Cocoanut charcoal also absorbs the actinium emanation strongly at the ordinary temperature F. S. Radium Emanation Contained in the Air of Various Soils. JOHN SATTERLY (Proc. Camb. PhiZ. SOC. 1911 16 336-355. Compare Joly and Smyth (Abstr. 1911 ii 1048).-The radium emanation contained in the underground air of various soils and at various depths in the neighbourhood of Cambridge has been found on the average to be 2500 times as great as that in the atmospheric air of the same locality.Low barometer usually decreases the amount in the air probably because of the rainfall accompanying it which chokes the pores of the soil. In the tests of the influence of depth the greatest amounts of radium emanation were found at the greatest depths. Continuous withdrawal of the air from the ground did not exhaust the supply of emanation. F. S. A n Attempt made to Detect the Electric Conductivity of Radium-D. L ~ O N KOLOWRAT (Le Radium 191 1 8 401-404).- If the radium emanation is stored in a vessel all of the same potential i t is to be expected that the radium-D formed will be deposited as a uniform film.From the manner in which the resistance of films of platinum or silver suddenly augments when the thickness is decreased between and 10-6 cm. and on the assumption that the density and electric conductivity of radium-D are the same as that of lead it is calculated that the emanation accumulating in six days from 0.25 gram of radium should produce in a capillary tube 5 mm.ii. 118 ABSTRACTS OF CHEMICAL PAPERS. long and 0.2 mm. diameter a film of radium-I) about 4 x 10-6 cm. thick with a resistance of 40 ohms. Capillary tubes of this size provided with two platinum wire electrodes and silvered internally except for the space between the ends of the electrodes had an initial resistance greater than a megohm.Filled with the emanation from 0.27 gram of radium the silvering was rapidly attacked and rendered transparent which was ascribed t o the presence of oxygen. I n one experiment a ring of brown deposit midway between the electrodes and brilliantly metallic when seen by reflected light was seen three or four days after the emanation was introduced but this deposit after three days more gradually and cotnptetely disappeared and was hardly visible even under the microscope. The deposit is supposed to have been radium-U oxidised i n course of time by the presence of a trace of oxygen. I n the second experiment no deposit formed but in a third n non-uniform discontinuous deposit appeared. IKI all cases the re5istance of the tube remained greater than a megohm.The formation of the active deposit appears not to take place uniformly but over certain areas determined by the position of the electrodes aud the shape of the tube. F. S. The Decay-constant of Polonium. ERICH REGENER (Rer. Daut. pJtysikal. Ges. 1911 13 1027-1033; Le Radium 1911 8 458-46 1 ).-The half-period of polonium determined by counting the scintillations produced on a zinc sulphide or diamond screen was 129.5 days the total number of scintillations counted being 16,800. This period is less than that usually given which varies in different measurements from 134.5 to 143 days. An examination was made of the effects of unsaturation on the period of decay in ionisation measure- ments. The ionisation of a strong polonium preparation was measured by a galvanometer at different voltages over a period of nearly a year.A t insufficient voltage to produce saturation the period found was higher than when higher voltages were employed The true half-period deduced from these measurements was 136 days (+_ 0.5 day). F. S. Radioactivity of Marsh Gas. JOHN SATTERLP (Proc. Carnb. Phil. SOC. 1911 16 356-359).-The inflammable gas drawn from stagnant backwaters and ditches around Cambridge during a long spell of fine weather was rich in radium emanation to the same extent approximately as the air drawn from depths of three to five feet in the soil of the same locality. F. S. The Radium Content of Various Fresh and Sea Watere and Some Other Substances. JOHN SATTEBLY (Proo. Can& Phil. SOC. 1911 16 360-364).-The amount of radium present in various fresh waters in the neighbourhood of Cambridge was usually less than gram of radium per litre which is some hundreds of times less than accounts for the quantity of emanation found in the water in the case of the tap water and spring waters.The water of the Cam contains much less dissolved emanation than these. Three specimens of coastal sea water contained about gram of radiumGENERAT AND PHYSICAL CHEMIS’I‘RY. ii. 119 per litre which is in agreement with Eve’s and much smaller than Joly’s results. F. S. The Apparatus Used f o r the Determination of the Radio- activity of Springs. FERDINAND H.ENRICH and FRITZ GLASER (Zeitsch. angew. Chem. 1912 25 16-19).-The ‘; fontaktoscope” of Engler and Sieveking has been compared with its improved form the ‘‘ fontaktometer ” of Mache and St.Meyer. Similar results were obtained with the two instruments and it is held that the former is advantageous on account of its cheapness its ease of working and its transportability. F. S. Electrical Resistance of‘ Special Steels. OCTAVE BOUDOUARD (Compt. rend. 1912 154 1475-1478).-The influence of different percentages of carbon on the electrical conductivity of nickel manganese chromium and tungsten steels of varying composition i s shown in tabular form. I n nickel steels the coefficient of resistance increases with the percentage of carbon. The curve showing the variation of the coefficient with the proportion of manganese shows a maximum corresponding with the compound NiFe,. I n manganese steels the resistance shows a maximum when 12-13% of manganese is present the amount of carbon having little influence on the value.The resistance varies in a very irregular way with chromium steels whether they are rich or poor in carbon. Tungsten steels show an increase up to a certain concentration then a minimum followed by a further increase. Benedick’s formula for the conductivities of steel expressing the conductivity as a linear function of the sum of the concentrations of the different elements present expressed in terms of their carbon equivalents is found t o hold good for the special steels studied. w. 0. w. Conductivity of Pure Ethyl Elther. J. CARVALLO (Com.pt. rend. 191 I. 153 1144-1145).-‘l’he author considers that tjhe ether employed in Schroder’s experiments (Abstr.1909 ii 462) was not pure and that his values for the conductivity were too high. With ether purified over sodium and in the absence of air a current of 0.93 x 10-10 amperes traversed the liquid on establishing a difference in potential of 425 volts but this diminished progressively owing to the experiment being conducted in glass vessels. It will be neces- sary to employ metallic apparatus with para5n insulation to obtain accurate values. w. 0. w. Influence of Temperature on the Density and Ellectrical Conductivity of Aqueous Salt Solutions. HEINRICH CLAUSEN (Ann. Physik 1912 [iv] 3’7 51-67. Compare Heydweiller Abstr. 1910 ii. 106 398).-Tho influence of temperature on the increase of density due to the ionised and un-ionised portions of the electrolyte has been examined in the case of aqueous solutions of twelve salts and of nitric acid by measurements of density and electrical conduct- ivity at 6O 18” and 30°.From these data the constants A and B in the formula As=- &i + &( 1 - i) have been calculated A being theii. 120 ABSTRACTS OF CHEMICAL PAPERS. difference between the density of the solution and that of water and i the degree of ionisation. The constant A decreases with rising temperature whilst B also decreases for most of the salts examined but to a much smaller extent. I n the case of potassium chloride bromide and iodide and of nitric acid B increases as the temperature rises. From this i t follows that tbe influence of the ionisation on the change in density diminishes as the temperature is raised and the alteration is such t h a t the values of A - B for the different electrolytes become more nearly equal.H. M. D. Dependence of the Conductivity of Binary Normal Electro- lytes on the Concentration. PAUL HERTZ (Ann. Physik 1912 [iv] 37 1-28).-A theoretical paper in which the author deduces a relationship between the conductivity and the concentration of the form A - X -function of ,qc where X is the molecular conductivity of the electrolyte at concentration c and A the Conductivity at infinite dilution. By euitable variation of the dimensions of the system of co-ordinates from case to case it is shown that the relationship between conductivity and concentration can be represented for all binary electrolytes by a single curve. H. M. D. Conductivity of Certain Salts in Methyl and Ethyl Alcohols at High Dilutions. HENRY R.KREIDER and HARRY C. JONES (Amer. Chem. J. 1911 46 574-585).-1n continuation of earlier work (Abstr. 1911 ii 362) on the conductivity of very dilute solution8 of salLs in methyl and ethyl alcohols measurements have now been made of the conductivities at 0' and 25' of similar solutions of sodium bromide lithium bromide potassium thiocyanate and cobalt bromide. The values of pa were found in most of these cases. The ratio of the vahe of pa for a certain saIt in one solvent to that for the same salt in another solvent is nearly constant for salts which are solvated to approximately the same extent. When one salt is solvated very much more than another the value of pa is generally less for the former.The ratios of poo for certain salts in t w o different solvents deviate from the ratios of the fluidities of the solvents since the value of pa in the solvent which has the greater molecular mass is always less than would be expected from the ratio of the fluidities. The greatest difference between these ratios is shown by the most highly solvated salts. E. G. Afflnity Measurements in Alcoholic and Aqueous Alcoholic Solutious. ERIK HAQULUND (Arkiv. Kern. Mil.&. Geol. 1911 4 No. 11 1-31).-By means of conductivity measurements the author has investigated the ionisation of various acids and bases and the hydrolysis of certain salts in 80 and 99.8% alcoholic solutions. At 18' the mobility of the ammonium ion was found to be 17.9 in 80% alcohol that of the hydroxyl ion 26.1 in 80% alcohol and 16.5 in 99.8% alcohol.Conductivity data were obtained for salicylic cinaamic and acetic acids and for p-nitrophenol ; also for piperidine ammonia,GENERAL AND PHYSICAL CHEMISTRY. ii. 121 piperazine mil diethylamine in both SO% and 99.8% alcohol. For small variations of concentration i t is found that the iouihation of both acids and bases is in harmony with the requirements of the law of mass action. The degree of hydrolysis was examined in the case of aniline salicylate a-picoline salicylate piperidine salicylate a-picoline acetate ammonium acetate and ammonium p-nitrophenoxide. The extent to which these salts are hydidysed iu aqueous alcoholic solutions increases with increase in the alcohol content of the solvent. Conductivity determinations were also made in order to ascertain whether salt formation occurs when certain ketones and aromatic nitro-compounds are brought into contact with sodium hydroxide in alcoholic solution.I n the case of P-nitrophenylethylene salt formation takes place at a sufficiently slow rate to allow of the process being followed by means of conductivity nwasurements. The rate of progress in 80% alcohol indicates that the reaction is bimoleculnr but in 99.8% alcohol the corresponding values of the velocity coefficient diminish appreciably as the reaction progresses. Measurement of Electrical Conductivity. ALFONS KLEMEKC (Clbem. Zeit. 19 11 35 1420).-A rapid and sufficiently nccurate method is described of making up solutions containing 1 grmu- equivalent or one mol.in 32 64 . . . . . litres (the usual dilutions in which electrical conductivity data are tabulated). Electric Properties of Substances in Relation to their Allotropic State. NICOLAUS A. HESEHUS (J. Buss. P h p . Chm. Soc. 191 1 43 Yhys. Part 365-371).-Study of the three allotropic forms of carbon-diamond graphite and charcoal-is in accord with the view previously advanced (J. Russ. Phyrr. Chern. Soc. 1901 33 Phys. Part 19 ; Abstr. 1911 ii 13) and indicate3 that the differences between the contact electrifications of allotropic forms of a substance depend not only on the surface density but also on the capacity of the atoms for ionic dissociation. Surface Effects between Mercury and Certain Solutions and a n Electrochemical Method of Estimating Dissolved Oxygen. SAMUEL W.J. SMITH and WILLIAH F. HIGGINS (Trans. Faraday Soc. 1911 7 64-77).-The influence of sodium sulphide on the difference of potential between mercury and certain salt solutions has been investigated (compare Smith and Moss Abstr. 1908 ii 343). The cells are connected in turn with a constant electrode and the X . M . P . of the combination determined with an electrometer. Uaing i V / l O - solutions of potassium chloride bromide snlphate aud hydroxide and pure water a sudden change of slope of the E.M.F.-concentration curve occurs at about 0.0004 equivalent of sodium sulphide. This break slowly disappears even when air is excluded from the solutions and is replaced by one between 0.0005 and 0.0006 equivalent. This is the point at which oxygen is completely removed from the solution by the oxidation of sulphide to thiosulphate.This is contirmed by further experiments with old solutions. ‘l’he method may be used tor the eetimation of dissolved oxygen. H. M. I>. G. S. T. H. P.ii. 122 ARBTRACTS OF CHEMICAL PAPERS. The initial breakobserved a t 0.0004 equivalent is atkibuted to t h e formation of mercury sulphide which is a maximum at this concen- tration further additions of sodium sulphide diminishing the thickness of the film. This is confirmed by observations of the colours of the tarnished films. The surface adsorption of sodium sulphide must be greater than that of oxygen. C. H . D. Anode and Cathode Falls of Potential and the Minimum Potential in Chlorine. R. RAISCH (Ann. Physik 1911 [iv] 36 907-Y28).-Measurements of the fall of potential at the electrodes have been made in tubes filled with chlorine a t different pressures and provided with electrodes of aluminium platinum mercury and magnetite.In the case of magnetite electrodes which are only slowly acted on by chlorine the observed potential diEerences are not much larger than those obtained with other gases. With these electrodes tho fall of potential at the anode was found to be independent of the strength of the current but increased slightly with increasing pressure. When such Cubes were left filled with chlorine for several weeks in an undisturbed condition the anode potential a t a pressure of 2.41 mm. was found to have risen from 41 t o 300 volts. On passing a current through the tube the potential difference fell rapidly however to its original value.These observa- tions indicate that the large potential falls which are exhibited by tubes containing sluminiutlr platinum or mercury electrodes are due to the formation of a surface layer by the action of the chlorine on the metals. When such electrodes are pulverised by the paspage of N strong current the anode and cathode falls of potential are found to undergo considerable diininution as a consequence of the reuoval of the surface film. H. M. D. Deposits Obtained f r o m Flames by Electricity. BRUNO “HIEME (Zeitsch. physikcd. CILem. 191 2 78 490-499).-When ttwo electrodes between which a considerable difference of potential subsists are brought near togeLher in a coal gas or candle flame carbon 1s immediately deposited on t.he cathode. In certain circumriltances tohe deposited carbon forms long.narrow growths extending some distance into the flame. The smallest potential giving rise to these structures is 12 volts; they increase in size up t o about 25 volts and beyoud that yoinc do not undergo much further chmge. Similar deposits are obtained when the electrodes are connected with au induction machine giving a short spark. When the poles are reversed the carbon separates in large flakes from the electrodes. It is suggested that very finely divided soot might be prepared in this way and a continuous arrangement for this purpose is described. By means of the following arrangement metals can be deposited in small amount from their salts. Two iron wires were connected with the poles of an induction coil and held at right angles to each other in the flame.A platinurn wire on which were some copper sulphnte crystals was held so as to touch t h e positive pole and after the spark had passed nietallic copper was found on the platinum wire. In a similar way uiinute arriouuts of mercury and apparently also of sodium,GENERAL AND PHYSICAL CHEMISTRY. ii. 123 have been obtained from their salts in the flame. g!he investigatioii is being continued. G. s. Potentials of the Most Important Standard Electrodes. FRIEDRICH AUERBACH (Zeitsch. Eluktrochem. 191 2 18 13-1 8). -As zero for electrode potentials should be taken that of the normal hydrogen electrode at the same temperature as the electrode in question and with a hydrogen partial pressure of 1 atmosphere.For the calculation of the electrode potentials eh referred t o the hydrogen electrode as zero the ternperature-coefficients of the calomel electrode found by Richards (compare Abstr. 1898 ii 7) should not be used more particularly as they are of uncertaiu value from the effect of thermo-E.iW.F.'s which cannot be accurately allowed for. The dependence of Eh on the temperature can be determined directly or can be calculated from the heat of reaction by means of the Helmholtz formula. For the decinormal potassium chloride-calornel electrode eh is practically independent of the temperature between 0' and 30". The following are the most probable values of ch for the more important standard electrodes (concentrations in gram formula weights per litre) eh in volts at Electrode.IIg I Hg,Cl I KC1 0.1 ......... Hg 1 HgO I KOH 1.0 ......... Hg 1 Hg,Cl 1 KC1 1.0 ......... Hg I H#,SO 1 H$O 0'1-0*5 Hg I HgO I NaOH 1-0 ......... Hg I HgO I NaOH 0.1 ......... 7 ~~ ~ 7 0". 18". 25". t- 0.337 -t- 0 *337 + 0.337 + 0.289 + 0'286 + 0.285 + 0.130 - + 0.107 + 0.133 - + 0.111 + 0.184 - + 0.166 - + 0.68 - 0. s. Aluminium Anode-Films. G. E. BAIRSTO and 11. MERCER (Trans. Furadccy Xoc. 1911 7 1-29).-An examination of the results of previous workers shows that the aluminium anode-film is made up of two layers an outer thick conducting film which varies in thickness with the electrolyte and an inner thin dielectric film the thickness of which is independent of the electrolyte. The electrolytes now etudied include potassium ferro- and ferri- cyanide ammonium arsenite and arsenate and ammonium moly bdate of which only the last gives good results giving a high resistance very little affected by temperature.The capacity i n sodium phosphate solutions decreases with rise of temperature whilst in ammonium borate it increases considerably a large part of the increase persisting after cooiing. The addition of chlorides bromides and other salts which have little or no valve action to a phosphate electrolyte produce a very large increase in the leakage current. The capacity decreases slowly with the time of formation. The higher the voltage a t which the film irJ formed the longer is the time required for. the capacity to fall to its minimum value. The product of capacity and voltage of formation decreases wlth increasing voltage.ii.124 ABSTRACTS OF CHEMICAL PAPERS. Experiments with alternating currents using an oscillograph show that the asymmetrical conductivity only becomes marked a t high voltages. C. H. D. Configuration of Equipotential Line8 in an Electrolyte. ANDRE BBOCHET (Compt. rend. 1911 153 1150-1152).-Lines of eyuipotential surfaces in a copper sulphate solution have been deter- mined and are reproduced in the form of curves. The method employed in mapping them was to plunge a copper wire into the liquid to different depths; the sound being connected to a potentiometer was not traversed by a current so long as the fall of potential between it and the t.wo electrodes was the same as the resistances of the two branches of the potentiometer.A Deprez-d’ Arsonval galvanometer was found most suitable for the purpose. w. 0. w. Configuration of Equipotential Lines in an Electrolyte. J. DELVALEZ (Compt. rend. 1912 154 1474-1475. Compare ibid. 121 492).-A claim for priority against Brochet (preceding abstract). w. 0. w. Study of the Hydrogen Electrode of the Calomel Electrode and of Contact Potential. N. E. LOOMIS and SALOMON F. ACREE (Amer. Chem. J. 1911 46 585-620).-The methods usually employed for determining hydrogen ion concentration in the study of organic reactions such as the hydrolysis of esters present serious difficulties in their general application. A study of the hydrogen electrode has therefore been undertaken in order to ascertain whether it can be satisfactorily used for the purpose.The relative efficiency of the various forms of apparatus hitherto used has been investigated and new forms have been devised to eliminate various sources of error. As the accuracy of the work depends largely on the constancy and ease of reproduction of the calomel electrodes used with the hydrogen electrode a large number of calomel electrodes have been prepared and measured against each other in order to obtain an average value of the potential. Many platinum electrodes have also been prepared and compared with one another. The hydrogen electrode has been compared directly with the calomel electrode and experiments have been made to determine the efticiency of various solutions in eliminating contact potential. All the experiment8 were carried out at 25’.The following results were obtained. Calomel electrodes N/’lO-KCl-HgCl-Hg can be prepared of which the variation does not exceed 0.10 millivolt during the first three weeks but afterwards slowly increases. Platinum electrodes can bo prepared which when used as hydrogen electrodes in H/ 1 0-hydrochloric acid show a deviation from the mean value of less than 0.10 millivolt. The E.M.P. of the system H,-Pt-N/ 10-HCl-N/lO-KCl-HgCl-Hg is 0.4266. Potassium chloride was found to be by far the most efficient of the salts tried for eliminating contact potential. A saturated solution of this salt eliminates almost completely the contact potential of systems cousisting of potassium chloride and hydrochloric acid.GENERAL AND PHYSICAL CHEMlYTRY. ii.125 The value of the potential of the electrode N/lO-KCI-HgCl-Hg is 0.3390 if the dissociation of N/lO-hydrochloric acid is 92*2% as found by Noyes; i t is 0.3373 if the dissociation of the acid is 86% and 0.3355 if the contact Dotential is assumed to be 0.0284 a value calculated from Lewis ahd Rupert's data (Abstr. 1911 ii 364). E. G. Application of the Hydrogen Electrode to the Measure- ment of the Hydrolysis of Aniline Hydrochloride and the Ioniaation of Acetic Acid in the Presence of Neutral Salts. N. E. LOOMIS and SALOMON P. ACREE (Arne?.. Chem. J. 1911 46 621-637. Compare preceding abstract).-The hydrolysis of aniline hydrochloride has already been studied by means of the hydrogen electrode by Denham (Trans. 1908 93 48). The results he obtained are now discussed and an account is given of a further investigation.It has been found that if the value of the electrode N/lO- KCl-HgC1-Hg is taken as 0.339 the hydrolysis of a N/16-solution of aniline hydrochloride is 2.19% and that of a N/S%solution is 3.05%. If the value 0.3355 is employed the hydrolysis becomes 1.81% for the N/16- and 2.56% for the N/3%solution; these values agree closely with bhose of Bredig (Abstr. 1894 ii 229) and Tiszard (Trans. 1910 07 2492). It is evident therefore that the hydrogen electrode is of service for the accurate determination of the relations between conductivity and hydrolysis. It is shown that the addition of N/lO-potassium chloride to N/4- acetic acid increases the hydrogen ion conductivity by about 4.5% of its original value. A bibliography is appended.Difflculties in the Use of the Hydrogen Electrode in the Measurement of the Concentration of Hydrogen Ionls in the Presence of Organic Compounds. L. JUNIUS DESHA and SALOMON F. ACKEE (Anzer. Chern. J. 1911 46 638-648. Compare preceding abstracts).-When an oxime is formed by the reversible reaction between a csrbonyl compound and a hydroxylamine salt the oxime may exist in the solution as the free oxime oxime salt and oxime cation. The concentration of the oxime cation can be measured by determining the amount of free hydrogen ions which disappear in its formation. Only two methods seemed likely to prove satisfactory for estimating the hydrogen ion concentration during the reaction namely sucrose iuversion and the potential of the hydrogen electrode.The former method has given good results in experiments which will be described in a siibseyuent paper. An account is now given of the difficulties attending the use of the hydrogen electrode. It has been found that some organic compounds are decomposed in presence of the hydrogen electrode and especially when oxygen is present and the h!.Jf.F. observed may not correspond with the con- centration of the hydrogen ions present. I n several cases in which hydroxylamine hydrochloride and acetone were both present the maximum potential obtained indicated a hydrogen ion concentration about twice as great as was possible if the hydroxylamine and acetoxime hydrochlorides had been completely hydrolysed. E. G. VOL. CII. ii. 9ii 126 A HYTRACTS OF CHEMICAL PAPERS.In most experiments with the hydrogen electrode the true E.Jf.Ii’. can be ascertained within thirty minutes but if the platinum electrode and the reacting liquid are previously saturated with hydrogen the E.M.F. can be measured within a millivolt in five to ten minutes. Ammonium nitrate does not entirely eliminate contact potential. Capacities of the Electrolytic Valve Effect .in Fused S a l t s and in Absolute Sulphuric Acid. GUNTHER SCHULZE (Zeitsch. Elektrochem. 1912 18 22-29. Compare Abstr. 191 1 ii 365).-In continuance of the experiments on electrolytic valve action the thick- nesses of the active layers formed on aluminium and tantalum in the presence of fused salts have been calculated on the assumption that the dielectric constant is unity. The capacity measurements were made by a method which is fully described after the cells had been formed for thirty minutes at a potential of 50 volts.The thickness of the layer formed on tantalum which in aqueous solutions is independent of the nature of the electrolyte and the con- centration attainable a t Oo increases regularly a t higher temperatures as the proportion of water is further diminished until the fused salt is reached. The same phenomenon is observed in the transition from dilute to absolute sulphuric acid both with tantalum and aluminium. For tantalum the thickness in dilute electrolytes at 0’ is 4*10pp in absolute sulphuric acid a t 10’ 6*69pp in fnsed silver nitrate 7*12pp and in fused potassium nitrate ll.1p.p. With the different fused salts the thickness of the layer is the greater and the capacity the smaller the higher the melting point of the salt and when the capacities are plotted against the melting points a straight line is obtained which on extrapolation cuts the axis at 510° as if above this temperature no valve effect was possible. E.G. G. S. The After-glow of Electric Discharge and Kindred Pheno- mena. (Hon.) ROBERT J. STRUTT (Proc. Physical Soc. 1911 24 1-8).-It has been found previously that the ozone obtained from a Siemens’ ozoniser supplied with oxygen a t atmospheric pressure does not give any after-glow when mixed with nitric oxide although this effect is obtained when the ozone is obtained from a low pressure dis- charge. Further experiments have shown however that the absence of the after-glow in the first case is due to the very considerable dilution of the ozone by the oxygen for if the ozonised oxygen from the Siemens’ tube a t ordinary pressure is concentrated by fractional distillation the greenish-yellow af ter-glow is readily obtained.Under similar conditions a blue glow appears when the ozone is mixed with hydrogen sulphide. The increased brilliancy of the after-glow which has been found when ordinary or fuming sulphuric acid or sulphur were present along with air in the discharge tube is now shown to be due to the removal by these substances of organic matter which is prejudicial to ozone. When these organic impurities are removed the sulphur compounds have no influence on the after-glow.QENERAL AND PHYSICA I CH ICMIS'I'RI'. ii. 127 Pure dry oxygen exhibits no after-glow and the white glow observed with imperfectly dried oxygen is probably due to the interaction of ozone with soulo substance produced by the action of the discharge on water vapour.The luminosity which is emitted when spring water is shaken up with ozone has been traced to oxidation of the peaty matter present. H. M. D. This substance is not however hydrogen peroxide. Decomposition of Nitrogen Peroxide in the Electrical Glow. J. ZENNECK and B. STHASSER (f'hysjhd. &tsch. 1911 12 1201-1204. Compare this vol. ii 1G).-Spectral observations have been made of the light emitted by the diEerently coloured layers which are obtained when a current of nitrogen peroxide is passed through a long narrow discharge tube. The orange-yellow greenipb- yellow and bright red glows of the first.third and fourth stages exhibit spectra which have been recognised as due to nitrogen peroxide nitric oxide and nitrogen respectively. In the fourth stage some oxygen lines are also present. The violet-blue glow of the second stage is probably due to an intermediate labile oxide its spectrum differing from that of the peroxide by the absence of the continuous portion and from that of nitric oxide by a difference in the relative intensities of the lines. As a result of the investigation of the different regions for after-glow effects it is found that these are readily observed in the case of the first third and fourth stages but the seoond stage does not appear to show any after-glow. H. M. D. Magnetic Temperature-coefficients of the Ferro-magnetic Elements in Corresponding States.J. K. ASllWORTH (Phi[. Mag. 1913 [vi] 23 36-45).-Experiments have been made on the varia- tion of the residual magnetic intensity of magnots made of nickel iron and cobalt for variations of temperature between that of the room and the magnetic critical temperature. If the fractional magnetic. intensity (Z/Io) where I. denotes the intensity a t absolute zero is plotted against the temperature measured as a fraction of the magnetic critical temperature it is found that the experimental data for the three metals can be represonted by a single curve. The critical temperatures of nickel and iron were found to be 388" and 785" respectively and 1090' was adopted as the value for cobalt. From this result i t follows that the characteristic tomperature-coefficients of magnetic intensity of the pure annealed ferro-magnetic elements in corresponding states are inversely proportional to the absolute magnetic critical temperatures The law of variation of the magnetic intensity of t h e ferro- magnetic elements with temperature is therefore of the same kind as that which expresses the variation of the density of a liquid and it is shown that the coefficients involved are of the same order of magnitude.H. M. D. Dependence of the Coefficient of Expansion on the Tem- perature. CHARLES L. LINDEMANN (Physikul. Zeitsch. 19 1 1 12 1197-1 199).-The thermal expansion of rods of aluminium copper 9-2ii. 128 ABSTRACTS OF CEEMICAL PAPERS zinc silver and lead has been measured at low temperatures over the separate ranges afforded by the use of liquid hydrogen liquid air liquid oxygen and ordinary room temperatures When the mean expansion coefficients thus obtained are divided by the mean atomic heats calculated from Nernst and Lindemann's equation (compare Abstr.1911 ii 466) numbers are obtained for each metal which are approximately independent of the temperature. This result is in accordance with the relationship anticipated by Griineisen (Ve.rh. deut. phys. Ges. 19 11 13). H. M. D. The Linear Expansion of Solid Elements as a Function of the Absolute Melting Point. A. STEIN (Zeitsch. anorg. Chenr. 1913 73 270-273).-Lf u is the linear coefficient of expansion of a solid element and v its atomic volume the quantity u O / 2 ) may be considered as the linear expansion of the atomic volume.h e product u gGT where T is the absolute melting point is practically constant over a large range OF melting points being however high for elements of very low melting point (P K Na S) and low for the infusible platinum metals. The regularity becomes more pronounced when the expansion is measured a t low temperatures the coefficient of the more fusible elements being more variable. The relation may be expressed in the form that the atomic heats of solid elements being equal the force resisting the separation of the atoms with increase of tempera- ture is proportional to the absolute melting point. Cubical Expansion of Fused Silica and the Variation of the Boiling Point of Aniline with Preesure.FREDERICK J. HARLOW (Proc. Physical Soc. 191 1 24 30-39).-Neasurements of the cubical coefficient of expansion of fused silica have been made by the weight thermometer method the data being referred to the absolute coefficient of expansion of mercury. For the interval 0-loo3 the cubical expansion coefficient is given as 99.8 x and for 0-184" as 144.7 x lo-*. The former value is considerably smaller than that calculated from previous measurements of the linear expansion but the fact that the coefficient has been shown to change in sign a t about - 80' points to its probable correctness. From measurements of the boiling point of aniline a t pressures from 74 to 78 cms. of mercury it is found that the temperature rises 9-49" per cm. of mercury. C. H. D. H. M. D. Change of Volume Accompanying the Fusion of Crystals. The Thermal Expansion of Grysfals and of their Products of Fusion.HEINRICH BLOCK (Zeitsch. physikal. Chem. 19 12 78 385-425).-The changes of volume which occur in the fusion of crystals was determined by drawing the fused substance into a capillary tube not more than 1.5 mm. wide sealing one end of the tube and allowing the substance to solidify slowly a t a temperature very little below its melting point whereby the occurrence of air spaces is excluded. The position of the upper surface of the solid is then carefully noted as is that of the upper surface after fusion. The tube is then carefully calibrated with mercury and from the data thusGENERAL AND PHYSICAL CHEMISTIIY. ii. 129 furnished the change of volume on fusion is obtained.When impure substances are used the impurities rise t o the surface of the liquid in the capillary tube and can be removed after which the d e k i t e read- ings are taken. The thermal expansion of the crystals was obtained by determining the density at different temperatures by the floating method and the density of the liquid was taken in a special form of dilatometer. These observations gave a n alternative method of obtain- ing the change of volume on fusion and the results are in very good agreement with those obtained directly. The physical constants in question were found for thirty-one substances (organic compounds with one exception) and the results are given in tabular form. ‘They are in most cases in good agreement with those of previous observers.The available data for changes of volume on fusion including those for a number of inorganic compounds are also tabulated. No relationship has been found between tho molecular weights of substances and the changes of volume on fusion. On the other hand Tammann’s formula 6v = I\(d,v’/dT- dpv”/dI’) where 6v is the change of volume on fusion in C.C. per gram 26 the absolute temperature of fusion and dtv‘/dT and d,v”/dT the change in volume of liquid and solid respectively (in C.C. per gram) for a change in temperature of 1’ (compare Krystnllixieren und Schmelzen Leipzig 1903) is valid for the great majority of substances examined. Thermodynamics of the Equilibria in One Component Systems. I. GUSTAV TAMMANN (Ann. Physik 191 1 [iv] 36 1027-1054).-Frorn a consideration of the theory of thermodynamic potential in its application to one component systems it is shown that two forms of equilibrium curves are possible.One of the two groups of curves is characterised by a critical point but such a point is absent in the second group. From this it follows that the limits of existence of the two phases corresponding with an equilibrium curve of t h e first type are incompletely defined whereas the conditions of existence of the phases corresponding with curves of the second type are completely determined by closed systems of curves on the equilibrium diagram. Similar results have already been obtained on the basis of molecular considerations and the theoretical deductions are in agreement with experimental observations.A Correlation of the Elastic Behaviour of Metals with Certain of their Physical Constants. JOHN JOHNSTON (J. Wash- ington Acud. Xci. 1911 1 260-267).-When pressure acts OLI a solid phase but not on the liquid phase produced from it the effect is always to lower the melting point. By means of a thermodynamical equation (Lewis Abstr. 1908 ii 465) t h e pressure required t o pro- duce melting at 25” is calculated for a number of metals for which the densit’y and the latent heat of fusion are known. The values obtained increase from 64 atmospheres for potassium to 21300 for platinum. The order is found bto be the same as that of decreasing ease of flow and compressibility and of increasing Young’s modulus and modulus of rigidity. This indicates that the flow of metsls is due t o conversion into liquid ,under the influence of pressure in accordance with the views of Beilby (Abstr.1904 ii 647). G. S. H. M. D. C. H. D.ii. 130 ARSTRACTS OF CHE.MICXL PAPERS. Neutral Salt Action as Exhibited in the Freezing Points of Mixtures in Aqueous Solution. ALBERT C. D. RIVETT (&fed. A'. Vetenek. Nobelinst 191 1 2 No. 9,1-32).-Measurements of the freezing points have been made for aqueous solutions of fourteen different salts and mixtures of these with (1) ethyl acetate and (2) sucrose. The freezing-point depressions for the two series of mixtures are always greater than the sum of the individual depressions due to the salt and the non-electrolyte. This departure from additivity is attributed to mutual influences operating between the salt and the non-electrolyt e and in the absence of evidence as to the relative erttents of these influences the relative increases in the freezing-point lowering are calculated on the assumption that the effect is entirely due (1) t o the action of the salt (2) to the action of the non-electrolyte.The data thus obtained show that the influence of the salt (or non-electrolyte) is independent of the concentration of the non-electrolyte (or salt) and is in general directly proportional to its own concentration. The pro- portionality doesnot hold in the case of sucrose and the more concen- trated solutions of lithium chloride magnesium chloride and magnesium nitrate. The order of the salts is not the same for the two non-electrolytes and this is considered t o furnish a strong argument against any explanation of the observations which is based entirely on the formation of hydrates.I n tbe case of sucrose there appears to be no connexion between the influence of the different salts on freezing-point depression and on t. he veloci t,y of inversion. H. M. D. 2 8-Dimethylpyrone as a Solvent. G . POMA (Gaxxetta 191 1 41 ii 515-538).-From determinations of its latent heat of fusion (56.0 calories) the cryoscopic constant of dimethylpyrone is found to be 58.7 whilst the mean deduced from cryoscopic measurements with stable organic substances (azobenzene mthracene phenanthrene and phenanthraquinone) is 64.6. The molecular weights of benzoic acid and thymol in dimethylpyrone are slightly below the normal. The molecular weights of some electrolytes in dimethylpyrone solution were found to be much lower than the theoretical values These results accord with those obtained from the measurement of the electrical conductivities of the solutions in some cases but not in others. Measurements of the molecular surface energy and its temperature- coefficient show that dimethylpyrone is slightly associated.It also has considerable dissociating power but this is not manifested in the same sense as that of water so t h a t in the series of alkali halides the degree of dissociation is greatest for sodinm iodide and least for lithium chloride the dissociation of lithium bromide being intermediate between these two. The following other constants have been determined D137 0.9953 DfS3 0 9483; specific heat 0-looo (solid) 0.368; specific heat 152*3-183O (liquid) 0.550 ; surface tension at 137" 30.76 a t 1S3O 26.31 ; molecular surface energy at 137" 767.1 at 183" 677.5 ; coefficient of molecular association 1.21 (9).R. V. S.GENERAL AND PHYSICAL CHEMISTRY. ii. 131 Capillary Force of Evaporation. CARLO DEL LUNGO (Nuovo Cim. 19i1 [vi] 2 ii 425-430). -According t o Laplace's theory the molecules of a liquid in the surface layer are subjected to an attractive force directed towards the interior. This force must naturally be greatest on the molecules quite at the surface. From considerations of the work done in evaporation the author deduces mathematically the formula P=EpC/2 where P is the force a t the extreme outermost layer p is the density of the liquid C is the internal heat of vaporisa- tion and E is the mechanical equivalent of heat.From this it is calculated that at 0" P for water is 11,928 atmospheres for alcohol 3733 atmospheres for ether 1320 atmospheres and for carbon disulphide 2214 atmospheres The results are in fair agreement with those obtained by van der Waals by an alternative method. At looo the calculated value of P for water is 9852 atmospheres. I n the second part of the paper an alternative deduction is given of Sir J. J. Thomson's formula connecting the saturation pressure of a vspour and the curvature of its surface. G. S . Relations betwsen Critical Temperatures Boiling Points and Expansion Coefficients of Liquids. PHILIPPE A. GUYE ( Z'rans. Fm+adcq SOC. 19 1 1 7 11 9-1 21 Compare Prideaux Abstr.1911 ii 368).-The formulae of Mendel6eff and of Thorpe and Rucker (Trans. 1884 43 126) are only approximations and when several values of a are determined for the same liquid the differences found are greater than the errors of observation. The absolute boiling point T may be substituted for the critical temperature TC and the relation YJV2 = D,/D = (a'T - I',)/(a'Ik - TI) in which a' is a new constant =3-09 ives good results over a wide range of temperature. The three constants in the equation log Y= c - d log ( A - t ) may be found as shown by Mallet and Friderich (Arch. Sci. phys. r u k 1902 14 50) and good results are obtained for temperature intervals of from 150" to 300". The expansion coefficient may be found by the relation dvldt = ( d log e)/(A - t ) and the results thus obtained are in close agreement with experiment.The constant A is very nearly the observed critical temperature exceeding i t by 5' or loo; c represents the critical density and cld = 3.78. C. H. D. ,s The formula ot Avenarius expresses a more exact relation. Energy Relationships in Vaporisation and Electrolytic Dissociation. SVANTE ARRHENIUS (Medd. K. Vetensk. Nobelirtst 19 1 1 2 No. 8,1-34).-From a thermodynamic consideration of the changes in free and total energy which are involved in the process of vaporisa- tion it is found that the change in free energy A can be represented by the formula A = A + BT+ CT2 in which A represents the molecular heat of vaporisation a t absolute zero and B and C are constants. The change in total energy V can similarly be expressed by U = A - CT2.I n the case of water the constants involved give rise to the special equation A = - 11 394 + 48.742'- 0*00927T2 and it is shown that the values of A calculated from this are in excellentii. 132 ABSTRACTS OF CHEMICAL PAPERS. agreement with the experimental data between - 20’ and + 2 5 0 O . A t higher temperatures deviations occur which are ascribed to wide divergences from the simple gas laws at the high pressures corre- sponding with these temperatures. A similar agreement between theory and experiment is exhibited in the case of n-hexane and acetic acid. The constants A B and C have been calculated for a large number of substances. It is further shown that the free energy changes associated with dissociation processes (electrolytic and non-electrolytic) can be represented by formuJae of the same type and for these the corresponding constants are also tabulated.I n a11 cases the second term ST has a very important influence on the magnitude of the free energy change and on the other hand the formula for U indicates t h a t dUjdT becomes very small in the neighbourhood of absolute zero. A detailed examination of the values of U/T shows that Trouton’s rule does not hold at all accurately in the case of a large number of liquids and that it map lead to erroneous conclusions when applied to determine whether liquids are associated or not. Vapour Pressure of Nitrogen Peroxide. F. E. C. SCHEFFER and J. P. TRE~JB (Proc. R. Akacd. Wetensch. Amsterdam 1911 14 536-549).-Whilst liquids give the same vapour pressure by either the statical or the dynamics1 method solids which dissociate on passing into vapour give different results by the two methods (Ramsay and Young Abstr.1886 ii 410,965). The results obtained for nitrogen peroxide by Guye and Drouginine (Abstr. 1910 ii 1056) by a statical method differ considerably however from those found dynamically by Ramsay and Young and the vapour-pressure curve also shows a point of inflexion a character hitherto unknown. I n order to avoid contact with mercury a glass manometer of the form devised by Jackson (Trans. 1911 99 1066) is used. The statical measurements coincide almost exactly with those of Ramsay and Young and the vapour-pressure curve is of normal form without inflexion. The under-cooled liquid has been compared with the solid and the intersection of the two curves observed.The boiling point is found to be 21*2O and the melting point - 10%’. An approximate estimation of the degree of dissociation is also made. Composition and Vapour Pressure of Solutions. V. Change of the P a r t i a l Pressures of Vapours of Solutions and Mechanical Mixtures with Temperature. 31. S. VREVSKY (J. Russ. Phys. Chem. SOC. 1911 43 1446-1457. Compare Abstr. 1911 ii 256).-When the two liquids do not mix the change of the ratio of their partial pressures with change of temperature is expressed by H. M. D. C. H. D. d y - l - d P 1 dP 1 - --.--2.- dt y dt P dt Y,’ where P and P are the vapour pressures of the two liquids a t tem- perature t and y=p,/p that is the ratio between the partial pressures.The same expression holds for the case when the liquidsGENERAL AND PHYSICAL CHEMISTRY. ii. 133 mix assuming that the addition of the second liquid t o the first lowers the vapour pressure of the latter in the solution according t o -Raoult’s law p = P1x x being the number of gram-mols. of the first liquid per 1 gram-mol. of the mixture. ’\f7hen the formation of the solution of the two liquids is accompanied by a heat effect the change in composition of the vapour with temperature is expressed by where Q is the heat-effect produced on mixing x gram-mols. of the first liquid with (1 -x) gram-mols. of the second and R is the gas constant in heat units; hence in the general case change of the equilibrium between vapour and solution is the result of the combined action of two factors one of them physico-mechanical and the other physico-chemical.With solutions formed with either development or absorptionof heat these two factors act in the same direction in one region of concentration and in opposite directions in another region the limit between the two regions being the solution for which the heat of mixing is a maximum or minimum that is I n the first two cases considered above if between t and t’ the rate of increase of the vapour pressure of the first liquid does not change in the solution the same will hold for the second liquid. But in the general case where a heat-effect is observed on mixing and the temperature-coefficient of the partial pressure of the vapour is not equal to the temperature-coefficient of the vapour pressure of the liquid in the free state the change of composition of the vapour with temperature is expressed by dQ/dx = 0.Various pairs of liquids have been studied the results confirming qualitatively the theoretical deductions. Tbe Distillation of Difficult Boiling Substances. Huao DUBOVITZ (Seifeensied. Zeit. 191 1 38 529-530).-An enquiry into the distillation of mixtures under varying pressures with superheated steam; the author suggests the following formula as a means of estimating the vapour tension of a contained substance of high boiling point (M2P2)/(MlP1) =a where M and I’ are the molecular weight and vapour pressure respectively of the substance of lower and M and P the corresponding-data for that of the higher boiling point.T. H. P. I?. M. G. M. Heat of Formation of Titanium Dioxide. WILLIAM G. MIXTER (Amer. J. Sci. 1912 [iv] 33,45-48. Compare Abstr. 1909 ii 644). -The heat of formation of titanium dioxide has been redetermined by combustion of the finely divided metal in oxygen. The most trust- worthy observations give Ti + 0 = TiO (crystalline) + 218.4 Cal. This is about 1.1% higher than the value obtained previously by the sodium peroxide method. Both values are in complete disagreementii. 134 ABSTRACTS OF CBEMICAL PAPERS. with t h a t given by Weiss and Kaiser (Abstr. 1910 ii 302) which ia only 97.77 cal. H. M. D. Use of Dewar's Vessels in Calorimetry. ALEXIS J. BOGORODSKY (J. Russ. Phys. Chem. Soc. 3 91 1 43 1262-126S).-Dewar flasks which have been used recently by Mathews and Germztnn (Abstr.1911 ii 187) and by Luginin and Dupont (Abstr. 1911 ii 369) as calorimeters were employed by the author for this purpose as long ago as 1901 (Protokol Phil. SOC. Kaxan Univ. 1901 No. 197). The values then obtained for the heats of solution of KCIO and Na2B,07,10H,0 were - 10051 ( - 10040) and .- 26008 ( - 25859) cals. respectively the numbers in brackets being those given by Thomsen. The apparatus used is described and the results of further measurements made with it given. For Na2HP0,,12€€,O the molecular heat of solution gave the vdues 22836-22925 cals. the mean being 22880 (22830). T. H. P. Heats of Solution of Mono- and Di-hydrated Lithium Chlorides. ALEXIS J. BOGORODSKY (J. IZuss. Ph,ys. Chem. Soc.1911 43 1268-1374).-The experiments here described were carried out by means of the apparatus mentioned in the preceding abstract. The mean of the values obtained for the molecular heat of solution of LiCl,H,O at the mean temperature 20.3" is +4121 cals. the value of + (Thomsen) the change of the heat of solution with temperature being 29.02 cals. per degree. For LiCI,ZH,O the mean value of the molecular heat of solution is +9Sl cals. a t the mean temperature 22*2O (b having the value 12.3. From these results together with the value 8444 given by Thomsen for the heat of Solution of LiCl a t 20*7O the following equations are deduced Tic1 + H,O = LiCI,H,O + 4323 cals. ; LiCI,H20 + H20 = LiC1,2H,O + 3162 cals. On account of the lorn transition temperature ( - 15') of the trihydrate the heats of formation and solution of this hydrate were not investigated.The existence of such a hydrate if this were unknown would be indicated by the positive sign of the heat of solution of the dihydrate. The proposition advanced by Thomsen to the effect that the heat of solution of haloid compounds in water is negative when they are completely saturated with water possesses no general significance. T. H. P. A New Method of Determining Vapour Densities. IX. PHILIP BLACKMAN (J. Physical Chern. 1911 15 869-870. Compare Abstr. 1908 ii 157).-The improved apparatus consists of a graduated tube closed by a ground-in stopper carrying a capillary U-tube. The tube is filled with mercury and after the weighed amount of liquid has been introduced is closed and inverted.When the liquid has been vaporised its volume and pressure can be read off. XI. The Dehydration of Copper Su lphate Pentahydrate. PHILIP BLACKMAN (J. Phpsical Chem. 191 1 15 871-S73. Compare Abstr. 1909 ii 643).-Experiments made with the author's apparatus R. J. C. Simple Method for Vapour Density Determinations.GENERAL AND PHYSICAL CHEMISTRY. ii. 1.35 indicate that the dehydration of copper sulphate takes place pro- gressively as a function of temperature and pressure there being no sudden increase in stability as each molecule of water is removed fi,om the hydrate. Tbe dehydration was not carried as far as the monohydrate stage. H. J. C. Dilatometric Researches. New Form of Dilatometer for M i x t u r e s of Liquids. FILIPPO BOTTAZZI and GIUSEPPE EUGLIA ( A t t i R.Accad. Limei 1911 [v] 20 ii 683-627).-The authors describe a dilatometer intended tor measuring the change of volume occasioned by the mixing of two liquids. The mixture of the two liquids is brought about by the removal of the partition separating them the removal being due to the fusion of the wax which holds i t in position I n its original form the apparatus consists of a stomach- shaped glass bulb which has two openings at the ends of the upper surface. To one of these a graduated capillary is fiised whilst the other aperture consists of a wide tubulure ground on the inside. Into the tiibulure fits a glass tube which is open a t the bottom and is closed at the top by a glass stopper-. This tube serves to hold one of the liquids whilst the main bulb contains the other.I n filling the apparatus a suitable quantity of one liquid is placed in the bulb. The tube just mentioned is closed at the bottom by means of a circular glass plate attached to i t by mesns of a little wax of suitable melting point. To facilitate the introduction of the stopper into this tube the glass plate is provided with a small hole which is kept closed with the finger during the filling and is finally sealed with a spot of wax. The tube thus constructed is filled with the second liquid R glass marble is introduced into it the stoppcr is inserted and the hole in the glass plate sealed as described. The whole is then inserted in the tubulure of the bulb as though it were a stopper. All the joints are well covered externally with wax of high melting point and the apparatus is immersed in a thermostat. When a sufficient tempera- ture is reached the glass plate may be shaken off and the change of volume due to the mixing of the liquids may then be observed.A simplified form of the apparatus differs only in being cylindrical and in having the graduated capillary attached to the top of the tube instead of a stopper so that one of the ground surfaces is dispensed with. This form is recommended. R. V. S. Dilatometric Researches. 11. Preliminary R e s u l t s Re- gardihg Non-colloidal Solutions. FILIPPO BOTTAZZI and GIUSEPPE BUGLTA (Atii 12 Accad. Lincei 2911 [v] 20 ii 627-633. Compare preceding abstract).-The paper gives the chmges of volume observed on mixing solutions of sodium chloride sodium hydroxide sulphuric acid and hydrochloric acid with pure water.The changes in volume were always the same for the same liquids and calculations in the case of sodium chloride solutions showed t h a t the observed changes agreed with those calculated from the specific gravities before and after mixing. R. V. S.ii. 136 ABSTRACTS OF CHEMlCAL PAPERS. Law of Molecular Attraction. DAN TYRER (Phil. Mag. 1912 [vi] 23 101-113. Compare Kleeman Abstr. 1910 ii 22 492 600 932; 1911 ii 34 97).-According to Kleeman the magnitude of the force of molecular attraction is dependent on the distance of separation of the molecules and on the temperature but the author shows that so far as latent heat considerations are concerned it is unnecessary t o assume t h a t the law of attraction includes a tempera- ture function.From the data for the specific heat of various substances at constant volume it is found that the specific heat is considerably greater in the liquid state as compared with the gaseous and from this the conclusion is drawn that the change of energy which accompanies the isothermal expansion of a liquid is not entirely due to a change in the potential energy of molecular attraction but t h a t a liberation of intra-molecular energy must be involved. If it is assumed that the law of attraction can be represented by K2/Sn in which K 2 is constant for a given pair of molecules and S is the distance between them then the author's reasoning leads to the result that n cannot be less than 5 and is probably greater than 7.It is however impossible t o determine completely the lam of attraction from the consideration of the energy changes which accompany the change of volume of a liquid or a gas. Internal Molecular or Intrinsic Pressure. A Survey of the Various Expressions Proposed f o r i t s Determination WILLIAM C. McC. LEWIS (Trcms. Fu~uday ~ o c . 1911 7 94-115).- The expressions proposed by various authors for the internal pressure are reviewed and the assumptions underlying them criticised. I t is shown that all involve one or other of the assumptions ( 1 ) that the pressure K is independent of temperature (2) that K may be repre- sented by a / v t in van der Waals' equation and that a is independent of temperature that is that K/S2 is independent of temperature where 6 is the density.Neither of these assumptions is justifiable although the second is nearly true. The general thermodynamic equation does not allow K to be calculated wit,hout some further assumption. Assuming t h a t the fractional change of K per degree is the same as that of the latent heat per unit volume Li per degree the relation is found K= Li/l - (T/Ll).(8Ll/ST) by the use of which probable values are obtained for ether etbyl alcohol and carbon disulphide. C. H. D. Simplest Example of the Diagram of Hardness [System KBr-KF]. NICOLAI s. KURNAKOFF arid I. B. 17RSHESNEVSKY (J. Buss. Phys. Chem. Soc. 191 1 43 1392-1397).-The existence of eutlectics for the melting curves of binary mixtures of the type MF-MX (where M = K or Na and X = C1 Br or I) was shown by the work of Ruff and Plato (Abstr. 1903 ii 588).In order t o obtain information concerning the composition of the solid phase separating on solidification OF mixtures of potassium bromide and fluoride the authors have investigated the cooling curve by means of the registering pprometer previously dePrcri bed ( Abstr. 1905 ii 10). The melting-point curve consists of two branches meeting in an eutectic point corresponding with the temperature 580" H. M. D.GENERAL AND PEYSlCAL CHEMISTRY. ii. 137 and the composition 40% mol. KF. The addition of as little as 3% of one salt to the other CituEeS a marked eutectic arrest in the cooling curve this indicating that solid solutions are either absent or of very low concentration. I n agreement with this conclusion are the results of the hardness measurements. I t was found that potassium bromide and fluoride like other halogen salts of the alkali metals are plastic substances so that the measurement of hardness was carried out by determining the pressure required to produce flow ; i t has been shown by Kurna- koff atid Schemtschuschny (Abstr.1909 ii 855) that this method gives results in correspondence with those yielded by Brineil's ball test. Measurements of the pressures required to produce flow with potassium chloride bromide iodide and Biloride give values diminish- ing with the atomic weight of the halogen. Similar measurement with mixtures of potassium bromide and fluoride give values lying almost exactly on a straight line joining those for the separate salts this curve exhibiting no peculiarity.at a position corresponding with the eutectic point. These results show that when crystallised from the fused condition the system KBr-KF yields only superposed mixtures of the two salts. T. H. P. Fusion and Pressure of Flow of Mixtures of Isomorphous S a l t s . B. VHSHESNEVSKY (J. Rusa. PIiys. Chem. Soc. 191 I 43 1364-1398).-The investigations here described were made with the view of ascertaining how the hardness and the pressure necessiry to produce flow (compare Kurnakoff and Schemtschuschny Abstr. 1909 ii S55) with salts vary with respect to their capacity of forming solid solutions this being studied thermally. The melting-point diagram of the system KCNS-NaCNS consists of two branches meeting in an eutectic point corresponding with 30% mol.NaCNS. An arrest occurs at 143" owing to the conversion of the a-modification of potassium thiocyanate-stable above 143O-into the /3-form this change being accompanied by considerable develop- ment of heat (compare Gossner Zeitsch. Kryst. Min. 1904 38 136 ). A number of inferior arrests occur between 10 and 40% rnol. NsCNS the most probable cause of these being the formation of a new solid phase XaCNS,3KCNS. Calorimetric investigation of the heat of solution of these fused salts shows the maximum absorption of heat to take place with the composition NaCNS,3ECNS. On solidi tication of their fused mixtures potassium and ammonium thiocyanates form an uninterrupted serios of solid solutions. Owing to tho polyruorphisu of the ammonium salt the cooling curve exhibits four arrest5 (1) at 1 4 P corresponding with the conversion of the liquid t o the a-solid phase; (2) a t 120° indicating the change of the a- into the p-form ; (3) a t looo an eutectic point and (4) at 90° this being accompanied by a marked heat-effect and corresponding with the transformation of the p- into the y-modification (compare Gossner Zoc.cit.). At high temperatures these potassium and ammonium salts give a continuous series of isomorphous mixtures. In order to explain the origin of the two middle arrests the cooling curves of mixtures of I.ii. 138 ABSTRACTS OF CHEMICAL PAPERS. ammonium thiocyanate with the isomeric thiocarbamide were investi- gated. The cooling curve of thiocsrbamide also exhibits a n arrest at looo this being similar in character to that occurring at the same temperature with mixtures of potassium and ammonium thiocyanates.The melting point diagram of ammonium thiocyanate and thio- carbamide is characterised by a eutectic point at 100Oand by two other arrests at 120' and 90° so that the salt undoubtedly exists in the three modifications referred to above. Potaesium and rubidium thiocyanates form an uninterrupted series of solid solutions the system showing a minimum freezing point and being of Roozeboom's third type (Abstr. 1900 ii 132). The system KC1-KI on passing from the fused to the solid condition gives a continuous series of isomorphous mixtures this being confirmed by calorimetric investigation. A t the concentrations 10 and 70% mol. of the iodide the cooling curve shows arrests indicating the breaking-down of the solid solutions the curves characterking this process kiaring the nature of waves.The systems KCI-KBr and KBr-KI giva continuous melting- point curves each exhibiting a minimum which lies at 716" (60% mol. KBr) and 589O (50% mol. K1) respectively. I n either case the cooling curve shows only a n arrest corresponding with the conversion of t h e liquid into the solid phase; hence when the fused mixtures a r e cooled even to lorn temperatures no decomposition of the solid solutions occurs. Calorimetric measurements show however that such decomposition does take place with lapse of time. It is found that the pressure necessary to produce flow in the pure potassium halogen salts varies considerably with the rate at which they have been cooled.The rapid cooling referred t o in the following table was effected by pouring the fused salt into a cold platinum dish and medium cooling by means of a platinum dish previously heated ; for slow cooling the platinum crucible containing the substance is placed in a wider clay crucible situate i n a Fletcher furnace surrounded by asbestos cloth. The pressures are given in kilo:. per sq wm. Rs[~id hlediuni Slvw cooling. cooling. cooling. KCl ............ 28 35 50 -5 1 -8 KBr ........... 36 38 40.5 1 *13 K I ............ 20.5 24 31 '0 1.5 PI - p,. P3. P3 Pi. With fused mixtures of salts the effect of variation of the rate of cooling is more complex; with KC1-KI for example this effect is opposite in direction to that obtained with the pure salts.I n tho following experiments '' rapid " cooling was always employed. The maximum pressure required t o cause flow corresponds in the case of KQl-KBr with 40% mol. KBr and in that of KBr-KI with 60% mol. KBr. I n both systems this pressure is more than double t h a t required by the constituent salt which flows t h e less readily. After the cooled masses have been kept at the ordinary temperature for five weeks tho pressures required are considerably less than the original ones. It is hence evident that as the process of decomposi-GENERAL AND PHYSICAL CHEMISTRY ii. 139 tion of the solid solutions approaches its 6nal state the curve of pressures approaches the straight line joining the pressures for the two constituent salts. With KCI-KJ the pressure of flow curve shows two maxima and a minimum.The stibility and slight amount of decomposition of t h e solid solutions of low concentrations condit,ion a very rapid rise in the pressure in passing from 0 to 10% mol. KI and from 100 to 90% mol. KI. When the concentration of the solid solutions is increased con- tinuous decomposition of these solutions occurs until a state of mechanical mixture is approached this corresponding with a rapid fall in the pressure of flow. It is remarkable that the minimum pressure which corresponds with about 50% mol. is lower than the pressure for either of the two salts. With KCNS-NH,CNS the pressure of flow curve consists of two branches (1) that from 0 to 20% mol. KCNS being almost a straight line which corresponds with mechanical mixtures of y-NH,CNS with a definite concentration of the P-solid solution (20% mol.KCNS) ; (2) that from 20 t o 100% mol. KCNS showing a maximum a t about 70% mol. KUNS and corresponding with solid solutions of the P-modifications of the two components. With KCNS-RbCNS the curve shows a maximum a t 20% mol. KCNS which greatly exceeds the pressure for either component. Owing t o the occurrence of the new solid phase mentioned above the pressure diagram of the system KCNS-NnCNS is complicated by the presence of a maximum at 20% mol. NaCNS. Further increase of the latter results in the fall of the pressure to a minimum (intermediate to the values for the separate conitituents) a t 40% mol. NaCNS. T. H. P. Chief Law of Adsorption Phenomena. SVANTE ARREIENIUS (Medd.K . Vetensk. Nobelinst. 1911 2 No. 7 1-44).-The recent work of Homfray (Abstr. 1910 ii 771 1041) and Titoff (ibid. ii 1041) 0x1 the adsorption of gases by charcoal and that of G. C. Schmidt (ibid. ii 1041) on the adsorption of acetic acid from its aqueous solutions by charcoal indicates that the ordinary exponential formula a = k@ cannot be valid between wide limits. The formula is devoid of theoretical foundation and should be regarded at most as an empirical formula the applicability of which is limited to a comparatively narrow range of concentrations. By taking into account the fact that the quantity of adsorbed substance reaches a maximum value the adsorp- tion is supposed to be determined by the equation k.dx/dc = (s - x)/x which on integration gives log s/(s - x) - 0-4343.x/s = l/k.c where x represents the quantity of adsorbed substance per gram of charcoal s the maximum value of x c the pressure of the gas or the osmotic pressure of the dissolved substance and k a constant.For a given kind of charcoal the values of s (expressed in gram-molecules) are of the same order of magnitude for different substances. It is shown that the above formula affords a satisfactory representa- tion of the adsorption of gases and of acetic acid by charcoal within wide limits of concentration. At high temperatures and small con- centrations deviations are found between the calculated and observedii. 140 ABSTRAC'I'S OF CHEMICAL PAPERS. results which are ascribed to the variation in the heat of adsorption with the quantity of substance adsorbed.From the fact that the adsorption of a gas increases with increase in the attraction between the gas molecules the author infers that the phenomenon of adsorption is essentially determined by the attractive forces between the molecules of the adsorbed substance. In support of this view it is shown t h a t the changes in volume which occur when a liquid is compressed can be represented by the above adsorption formula. I n this case x represents the density of the liquid c the pressure (internal plus external) and s the maximum density. By reference to the compressibility data for ethyl ether and ethyl alcohol it is shown that experimental observations can be adequately represented by this formula. I n comparison with the exponential formula which contains two arbitrary constants the author points out that this formula contains only one for the quantity s is in every case directly determinable from experimental observations.H. M. D. Absorption of Gases by Porous Materials. JACQUES DUCLAUX (Conzpt. rend. 1911 153 1217-1230. Compare Abstr. 1911 ii 479).-An attempt to calculate the absorption coefficient of carbon for carbon dioxide on the basis of the hypothesis put forward in an earlier communication. Comparison of the result with the known value suggests that local differences in temperature have a greater effect ou absorption than local differences in pressure within the minute cavities of which the porous material is composed. w. 0. w. Adsorption in Solution. 11. Dualistic Nature of Adsorp- tion Phenomena.GEOHG vON GEORGIEVICS (Monatsh. 191 1 32 1075-1087. Compare Georgievics and Pollak Abstr. 1911 ii 1070). -The adsorption of acids from solution by wool also includes the dis- solution of acid in the wool arialogous to that shown in the adsorption of gases by charcoal. Once equilibrium between acid and wool is attained there is no slow further retention of acid. The velocity with which equilibrium is attained depends not only on the nature of the adsorbed and adsorbing substances but also on the conceutration of the solutions employed. Until a concentration of roughly 0.05 gram of acid per 250 c.c. hydrochloric and sulphuric acids are dissolved by wool ; in more concentrated solutions the phenomenon is one of adsorption i t being impossible to say whether there is any diffusion into the interior of the thread. Sulphuric acid is somewhat more easily soluble in wool than hydrochloric acid the ratio being approximately 9 7.The statement previously made that sulphuric acid is adsorbed more than hydrochloric acid in dilute solution and less in concentrated solution is now explained as owing to the excess of solubility of the sulphuric acid actually hydrochloric acid is the more adsorbed. Tha speed with which a substance is adsorbed is no criterion of adsorption as it is shown that from concentrated solutions very rapid diffusion into the inside of the solid material takes place. The facts mentioned afford an explanation of some anomalies in theGENEKAL AND PHYSICAL CHEMISTRY. ii. 141 results obtained on dyeing silk with picric acid (Georgievics Abstr.1911 i 537). I n dilute solutions containing 0.01 gram of picric acid per 100 C.C. and less the retention of picric acid is due t o solution; in more concentrated solutions of picric acid the retention is due to adsorption. I n concentrations of 0.01 to 0.02 gram per 100 c.c. there is a rapid increase of the retention of acid which is regarded as due to chemical processes. Picric acid and silk afford a case of retention due t o solution chemical action and adsorption in turn. Similarly solution of the dye in the fabric takes place in the case of dyeing in dilute solution ; there is some chemical action between dye and fabric but in stronger dye solutions tho retention is chiefly caused by adsorption. E. F. A. The Nature of S o l v a t e s and the Relationships between Adsorption and Dissociation. KURT GEBHAED (Zeitsch.Chenz. Ind. Kolloide 191 1 9 263-265. Compare Ostwald Abstr. 191 1 ii 1068).-The relationships between adsorption comporinds and solvates are discussed from the point of view of a theory of partial valencies with special reference to the explanation of the formation of complex compounds between dyes and fibres. H M. D. Mechanism of Osmosis. IEUGBNE FOUARD (Compt. rend. 19 1 1 153 1152-1155. Compare Girard Abstr. 1909 ii 537 ; 1911 ii 861).-A polemical paper against Girard and Henri drawing attention to the importance o€ absorption by the membrane in the mechanism of osmosis. Osmotic pressure can only be measured when this absorption is complete and when equilibrium has been established the osmotic effect is independent of the nature of the membrane.Objection is raised against De Vries’ definition of isotonic solutions on the ground that the protoplasm of a vegetable cell maintaining constaut volume in a solution is in osmotic equilibrium not with the external solution but with a liquid between the two cell membranes this liquid differing in composition from the first owing to the absorption of dissolved substance by the exterior envelope. w. 0. w. Behaviour of Edges and Corners in C e r t a i n Diffusion Experiments. KAPHAEL E. LIESEGANG (Zeitsch. Chem. I n d . Xolloide 1911 9 296-298).-Tf irregularly shaped pieces of jelly obtained by the solidification of a 10% solution of gelatin containing a small quantity of silver nitrate are placed in a saturated solution of potassium dichromate it is found that silver chromate is preferentially formed on those parts of the surface which are approximately plane.The portions of the surface in the neighbourhood of hollows and pro- tuberances are unaffected by the surrounding reagent. The importance of this observation in connexion with certain processes employed by histologists is referred to. H. 31. D. Diffusion of Oxygen and Carbon Dioxide in Water. TOR CARLSON (Medd. K . Vetensk. ivobelinst. 1911 2 No. 6 1-8).-The coefficients of diffusion of carbon dioxide and oxygen in water a t 18.2O VOL. CII. ii. 10ii. 142 ABSTRACTS OF CEEMICAL PAPERS. have been found to be 1.479 and 1.720 respectively. These values when corrected for difference in temperature are in fairly good agree- ment with the coefficients determined at 16” by Stefan and Hiifner.The ratio of the coeflicients is 1.166 which approximates closely to t h a t calculated from measurements of the rates of dissolution of the gases. The rates of diffusion are very nearly in t h e inverse ratio of the square- roots of the densities which requires 1.1 73 instead of the observed ratio of 1.166. H. 11. D. Velocity of Diffusion and Size of the Particles in Disperse Eystems. THE SVEDBEKG (Arkiv. Kern. Min. Geol. 1911 4 No. 12 1-7).-The author has carried out further experiments to decide between the Sutherland-Einstein and the Smoluchowski formulae for the rate of diffusion of a particle using a colloidal gold solution for the purposes of experiment.The velocity of diffusion of the gold particles was measured by means of the apparatus used by Svedberg and Andreen-Svedberg (Abstr. 191 1 ii 375) the concentrations being determined by measuring the extsinction-coeficients of the solutions. From the velocities of diffusion so determined the radii of the diffusing particles were then calculated by mems of the above-mentioned formulae and found t o be 1 . 2 5 ~ ~ and 297pp respectively. The size of the particles was then determined by Zligmondy’s ‘‘ nucleus method ” (Keimmebhode) and found to be 1*33pp which agrees with the first of the above two values. The Sutherland-Einstein formula may there- fore be taken as the correct one. 11. T. S. P. The Alleged Colour of the Ions. GAETANO MAGNANINI (Gaxzetta 1911 41 ii 425-429).-The author re-asserts his views on this subject (see Abstr.1893 ii 570 ; 1894 ii 226 ; 1897 ii 14). S y s t e m a t i c s of the Aggregated States of Matter. R. V. S. P. P. YON WEIMARN (J. ~ U S S . Yhys. Chan~ soc. 1911 43 1477-1502).- Principally a discussion of the author’s orientation or vectorial theory of matter (compare Abstr. 1910 ii 1048 and elsewhere). T. H. P. The Dependence of the Brown-Zsigmondy Movement on Temperature. 31. SEDDIG (Zeitsch. anorg. Chem. 1912 73 360- 384).-If the molecular explanation of the Brownian movement is correct the displacements should be proportional to the square-root of the absolute temperature and inversely proportional to the square- root of the internal friction of the liquid used (Einstein Ann.Physik 1905 [iv] 1’7 549). The values hitherto obtained from visual observation dift’er widely from theory owing to errors of measure- ment and t o the relatively long time of observation which allows currents in the liquid to form. An arrangement for preparing kinematographic records at a magnification of 3000 is described but quantitative measurements have not been made satisfactorily. The best results are obtained by using an ultra-microscope in which the rays are focussed in the object by a n oblique concave mirror and are prevented from reaching the eye-piece by a small lateral stop.GENERAL AND PHYSICAL CHEMISTRY. ii. 143 Two successive short exposures are made by ineans of a special shutter and the distance separating pairs of images is measured.The stage is electrically heated to a constant temperature which is measured by a thermo-couple placed in a layer of paraffin immediately below the thin glass carrying the liquid under examination. The suspensions used are cinnabar and freshly-prepared lamp black both of which give good contrasts and are uniform in size. As the displacements thus observed are only projections of the actual displacements on a horizoutal plane the absolute values are only found with difficulty but relative values for a range of tem- perature from 5.5” to 90” are obtained and show divergences which do not differ from those calculated by Einstein’s formula by more than 6% being always too high probably owing to absorption of heat from the source of light during the short exposure. The Brownian Movement of Particles in Colloidal Solu- tions 111.THE SVEDBERG and KATSUJI INOUYE (Arkiv. Kern. Mzn. Geol. 1911 4 No. 19 1-20).-!I!he method of measuring the Brownian motion of colloidal particles referred to in a previous paper (compare Abstr. 1910 ii 772 1017) is now described in detail. It consists in tracing the movements of individual particles over a considerable period of time (five to twenty seconds) and registering the positions occupied at successive times during such periods on R moving photographic plate. From the records obtained for a number of particles the displacements which occur as the result of the Brownian motion can be calculated. Experiments made with gold hydrosols prepared by direct reduction of gold chloride solutions with hydrazine show that the movements are in agreement with Einstein’s molecular kinetic formula.For Avogadro’s constant the data obtained with such hydrosols consisting of small sized particles give the value 6.2 x 1023. Determination of the Dimensional Distribution of the Colloidal Particles in a Disperse System. THE SvEDBERa and KNUD EsmuP (Zeitsch. Chern. fnd. Kolloide 1911 9 259-261).- The degree of uniforn-ity in the size of the colloidal particles in various dispersoid systems has been investigated by microscopic observations of the changes which occur in the spacial distribution of the particles when these are allowed to fall under the influence of gravity. Experiments were made with aqueous suspensions of mercury and gamboge with the milk juice of Chelidonium Eaciniatum cow’s milk cream from cow’s milk and cocoanut milk.Curves are plotted which show the numerical distribution of the particles accord- ing to their dimensions. I n this way curves are obtained which are more or less symmetrical and convex to the dimensional axis and the closeness with which the symmetrical branches of the curve approximate affords a measure of the degree of uniformity of the size of the particles. Coagulation of Prussian-Blue. NIGOLA PAPPAD~ (Gaxxetta 1911 41 ii 454-460. Compare PappadA and Sadowski Abstr. 1910 ii 593).-Ferric ferrocyanide differs from silicic acid in that C. H. D. H. M. D. H. M. D. 10-2ii. 144 ABSTRACTS OF CHEMICAL PAPERS. its coagulation is instantaneous so that the effect of different coagulating agents has to be measured by comparing the minimum quantities of them required to produce coagulation.The results obtained are identical with those for silicic acid. Undissociated organic substances do not cause coagulation. Coagulation is produced by electrolytes and the colloid has a negative charge. The coagulat- ing action of the salts of univalent cations increases with increase in the atomic weight of the cation whilst a comparison of univalent bivalent and tervalent cations shows that the coagulating power increases with the electrical charge of the cation. R. V. S. Stability of Emulsions of Water in Hydrocarbon Oils. ERICH GROSCHUFF (Zeitsch. Chem. Znd. Rolloide 19 11 9 257-259)- Observations relative to the stability of emulsions of water in benzene petroleum (D 0.792) machine oil (D 0-864) and paraffin oil (D OW32) have shown that the stability is not determined by the difference in density between the disperse phase and the dispersive medium or by the solubility of water in the hydrocarbon.It is more probable that the stability is to a large extent dependent on the surface tension at the surface of separation of t-he two phases increasing as the surface tension diminishes. The water emulsions examined increase in stability in the order benzene petroleum paraffin oil machine oil and the pronounced difference in the stability of the two last emulsioris is attributed to the presence of colloidal impurities i n the machine oil. H. M. D. Plant Colloids. I. The Solution Swelling of Starch in Presence of Crystalloids. MAX SAMEC (Koll.Chem. Beihefte 19 11 3 123-160).-Measurernents have been made of the temperatures at which starch granules lose their characteristic form by a process of swelling in which highly viscous globules of a starch solution are formed when the starch is heated in contact with water or aqueous solutions. An optical method is described by means of which the accompanying change in the transparency and refractive index of the granules can be employed for the accurate determination of this temperature. For a given solution the separlite observations agree to about 0*2-0.4°. I n all experiments the same sample of starch was used the swelling solution temperature of this being 59.7'. This temperature is altered in presence of salts acids bases and non-electrolytes and the effect of these different substances is qualitatively the same as in the case of the swelling of gelatin.Whether the temperature is raised or lowered is essentially deter- mined in the case of salts by the anion the cation having a much smaller influence. A t low concentrations bases exhibit the largest effect in lowering the swelling temperahre and in accordance with this it is found that kalts formed by a strong base and a weak acid have in general a depressant effect at moderate concentrations. I n certain cases such as chloral hydrate and carbamide the influence of non-electrolytes on the sweiling temperature is very considerable. H. M. D.GENERAL AND PRYSICAT CHEMISTRY. ii. 145 The Ultramicroscopic Structure of Jellies WILHELM BACH- MANN (Zeitsch.aqzorg. Chem. 1911 73 125-172).-Tbe internal structure of organic jellies may be studied by means of the cardioid condenser whilst t h e finer structure of silicic acid gels requires the ultramicroscope. The process of gel-formition is the same in gelatin agar-agar and silicic acid gels. The amplitude of the movements of the ultramicro- scopic particles diminishes and the size of the particles increases during the ageing of the gel. The process resembles that of the separation of two liquid phases from a solution (Lepkowski Abstr. 1911 ii 95). Solutions containing less than 1% of gelatin which do not set on cooling show a gradual formationof separate flocculent particles or aggregates of su bmicrons. With increasing concentration the structure becomes less distinct and at last 1x1 tramicroscopically homogeneous owing to the closeness of packing of the “ jelly-elements.” There is no evidence of a honeycomb or foam structure and these structures where previously observed are due to diffraction effects in the micro- scope and are much coarser than the submicronic structure.After treating with alcohol or chromic acid a new coarser structure is obtained due to the formation of capillary cracks following on partial coagulation. The clearest colloidml silica is obtained by dialysis with a collodion membrane. The mass dried over sulphuric acid polarises light when viewed in the ultramicroscope indicating a very minute heterogeneity. The changes observed during the passage through the opaque stage after immersion in benzene correspond with those described by Zsigruondy (Abstr.191 1 ii 880). C. H. D. Application of the Kinetic Theory of Gases to Chemical Problems. OTTO SACKUR (Ann. Physik 191 1 [iv] 36 958-980).- A theoretical paper in which the relationship betmeeu the entropy of a system in a given Condition and the ‘‘ probability ” of this condition is applied to the consideration of irreversible chemical changes. This leads t o the conclusion that molecules of different substances which are contained in an adiabatically closed space will only react together if the reaction leads to an increase in the “probability” of the system. The condition of equiiibrium is characterised by a maximum value of the “ probability.” H. M. D. Equilibrium in the Adsorption by Graham’s Ferric Oxide Hydrosol.P. MAFFIA (KolL. C‘hem. Beihefte 1911 3 85-122) - The relationship between the quantity of chloride adsorbed by colloidal ferric hydroxide prepared by Graham’s method and the chloride con- centration of the aqueous solution in equilibrium with it has been investigated by two series of experimenth in one of which the colloidal solution was filtered through a membrane of collodion and in the other was subjected to dialysis. The results obtained by each method indicate that the adsorbed chloride varies with the chloride concentra- tion of the dispersive medium in accordance with the requirements of the exponential equation q’m = p.cl/p. H. M. D.ii. 146 ABSTRACTS OF CHEMICAL PAPERS Reduction of Mercuric Chloride by Phosphorous Acid and the Law of Mass Action.JAMES B. GARNER (Amer. Chern. J. 1911. 46 648).-Garner Foglesong and Wilson's interpretation of the data obtained in their study of the reduction of mercurous chloride by phosphorous acid (Abstr. 1911 ii 972) was erroneous. New calcula- tions are being made and will be presented in a subsequent communication. E. G . A Special Case of Heterogeneous Equilibrium. P. P. FEDOTPEFF (Zeitsch. anoyg. Chem. 19 11 73 173-199).-Following on the study of the cuprous iodide equilibrium (Abstr. 1911 ii 42) the equilibrium Cu(NO,) + P b I Z? Pb(NO,! + CuI + I has been investigated. Neither cuprous iodide nor iodine reacts with lead nitrate but a mixture of the two readily forms lead iodide. The solubility of lead iodide in pure water a t 20' is 0.015 mol. per litre and this is increased to 0.216 in a saturated solution of iodine owing to the formation of the tetraiodide.I n a solution saturated with copper and lead nitrates the concentration of copper nitrate is almost the same as in pure water but that of the lead nitrate is reduced from 1.523 t o 0.052 mol. per litre. When Cu(N0,),,6H20 and CuI itre both present in the solid phase an irreversible reaction sets in with the formation of oxides of nitrogen and a basic salt. P b( N 0,) 2 P b I C u I I is realisable. The value of K= SCu/8Pb is not constant but varies with the total concentration 8(NO,) increasiDg with the dilution. The equilibrium has also been studied electrolytically by means of the cell Pt I solid PbI,,I I Pb(NO,) solution 1 KNO solution I CU(NO,)~ solution I solid Cu1,I I Pt the solutions used being in different experiments 3N/2 N / 1 and N/2 as regards S(NO,),.The observed E.M.F. is in good agreement with the calculated. The results are applied to the theory of fractional precipitation as in tho addition of a soluble iodide to a mixture of lead and copper nitrates C. H. D. The system with solid phase Colorimetric Investigation of Neutral Salt Action. BOHDAN VON SZYSZKOWSKI (Zeitsch. physiknl. Chem. 1912 78 426-489. Compare Abstr. 1908 ii 761).-The measurements were not made in a coloi-imeter but with two glass cyliuders which for comparison were held in the hands and examined with a background of white paper some cms. away. Methyl-orange was exclusively used its indicator. When solutions of acetic acid and of hydrochloric acid are compared the former acid gives the deeper tint with methyl-orange up to a H' ion concentration of 1 x 10-3 mols.per litre but with higher concentrations hydrochloric acid gives the deeper colour. The tints produced by the two acids are not quite identical. Neutral salts enter into reaction with methyl-orange and render i t more sensitive t o H' ions. This effect is mainly due to increased dissociation of the indicator as an acid and partly to the formationGENERAL AND PHYSICAl. CHEMIS’I’RY. ii. 147 of complex compounds with the salts. That the latter is not the sole factor is shown by the observation that sodiiim chloride is much more effective than potassium chloride in intensifying the colour. The interpretation of the results is complicated by the fact that methyl-orange is an amphoteric electrolyte.The effect of salts i n deepening the colour of solutions containing acetic acid and methyl-orange is due partly to an increase in the dissociation of the acid and partly to increased sensitiveness of methyl- orange. The “neutral salt action” of sodium salts is greater than that of potassium salts and that of bromides greater than that of chlorides. Non-electrolytes are practically without influence on the sensitiveness of methyl-orange. An estimate of the influence of potassium and sodium chlorides on the degree of dissociation of hydrochloric acid is given but the results are only approximate. Aqueous solutions of methyl-orange alter slowly in sensitiveness for some hours after preparation b u t then remain constant for weeks.Solutions of methyl-orange containing acids and salts slowly undergo an irreversible chemical change whereby the colour becomes much paler. G. S. Hydrolysis of Esters of Substituted Fatty Acids. WILLIAM A. DRUSHEL (Amer. J. Sci. 1912 [iv] 33 27-31).-From measurements of the rate of hydrolysis of ethyl acetate chloroacetate and cyano- acetate in presence of O*lN-hydrochloric acid a t 2 5 O i t is found that the velocity coefficients are in the ratio 67.7 45.8 10.25. The rate of hydrolysis of the cyanoacetate is smaller than would be expected from the relative ionisation constants of the three acid,.. The relative amounts of hydrolysis for a given time interval were also compared in a series of experiments with 0.1 molar solutions of the three esters in the absence of acid.The data show that the chloroncetate and cyano- acetate are hydrolymd much more quickly than ethyl acetate and the velocity for the chloroacetate is considerably greater than for the cyanoacetnte although cyanoacetic acid is a stronger acid than chloro- acetic. It is suggested that the observations may be explained if it is assumed that the cyanoacetic acid is to some extent polymerised in its aqueous solutions. H. 1%. D. Conflgurtttion of Ring Systems. JACOB BOESEKEN and A. VAN ROSSEM (Rec. trav. chim. 1911 30 392-406. Compare Abstr. 1911 ii 197).-Magnani (Abstr. 1890 1357 ; 1891 251) has studied the behaviour of boric acid i n solution towards mannitol and dulcitol by determining the conductivities of solutions of various strengths OF mixtures.The authors have extended the research to nine other alcohols and phenols. Glycerol pentaerythritol catechol and pyrogallol are found t o increase the molecular conductivity of boric acid whilst the other alcohols and phenols studied either have no action or else cause a diminution of the conductivity. Comparison of the enormous influence of catechol with the indifference of ethylene glycol shows that it is not sufficient that the hydroxyl groups should be in the ortho- or a-position to one another but that these two hydroxyl groups should be in theii. 148 ARSTRAC rs OF CHEMICAL PAPERS. same plane thus permitting the formation of ring combinations with the boric acid W. G. Velocity of Saccharification of Starch. VI. HENRI VAN LAER (Bull.Acad. Boy. Relg. 1911 795-830. Compare Abstr. 1910 ii 839; 1911 ii 28 478).-This part deals with Kjeldahl’s law of pro- portionality and the retarding action of the products of reaction on the activity of diastase. The results obtained by previous investigators are first critically considered notably those of Kjeldahl (Med. Cn~lsberg Lab. 1878 French edition) Brown and Heron (Trans. 1879,35,596) Ling (Abstr. 1902 ii 636) Ford (Trans. 1904 85 980; 1906 89 76) and Wohl and Glimm (Abstr. 1910 i 799) and a large number of new results are recorded. From all the results available the following conclusions are drawn. Even in feeble concentrations maltose exerts a small but nearly negligible inhibiting action on the activity of diastase but this retardation does not mask the appreciable rise in the coefficient of velocity of saccharification.The adsorption of diastase by maltose appears to result in the combination of minute quantities of the enzyme with a large excess of the sugar. All the peculiarities of the reaction appear to be explained by the adsorption compounds formed by the ferment with ( a ) starch ( b ) the dextrins and ( c ) maltose; thus the increase or decrease in the velocity of saccharification depends on the adsorption or release of new quantities of ferment by non-hydrolysed material. When the conditions are such that the enzyme is all adsorbed by the maltose a t the moment of its formation the reaction follows the logarithmic law. Observations on the reaction should not be made near the limits of saccharification.Kjeldahl’s law is followed in t h e reaction below and above the limit of 45% of maltose laid down by its author. Apart from the inhibiting action of maltose there are a series of retarding and accelerating influences capable of altering the limits within which the law holds; of these the most important is that due t o the rapid lowering of concentration of starch in liquids containing much diastase. The effect of the quantity of diastase used on the velocity of sacchari- fication is expressed by the equation X=nFm. In reactions in which the logarithmic law for unimolecular reactions is followed m = I but it becomes greater although always remaining nearly 1 when any departure from this law is made. Such increases in the value of m are not occasioned by greater adsorption of enzyme in solutions rich in diastase. I n hydrolysis of starch by acids the velocity depends on the number of ions per unit of volume whilst in hydrolysis by diastase it depends at each instant on the relation between the mass of enzyme adsorbed and that of the maltose which remains to be formed.T. A. H. Dependence of the Influence of Neutral Salts on the Concen- tration of the Acid in Catalytic Reactions. HARALD LUNDEN (Medd. K. Vetensk. Nobelinst. 1911 2 No. 3 1-5).-To express the influence of neutral chIorides on the catalytic action of hydrochloric acid in the inversion of sucrose the formula k = A.[H](l+ a [ H ] + b[K] +c[ClJ> is suggested. I n this ?c is the velocity coeficient [HI [R],GENERAL AND PHYSICAL CHEMISTRY. ii. 149 [Cl] the concentration of the hydrogen salt cation and chlorine ion respectively and A a b and c are constants.It is shown that the experimental data obtained by Trey for the influence of sodium and potassium chloride can be satisfactorily represented by means of this formula. The constants cb + b and b + c represent respectively the influence of the acid and of the neutral salt on the catalytic activity of the hydrogen ions. Since these have the same value for varying concentrations of the catalysing acid it follows that the relative neutral salt action is independent of the concentration of the acid. Since this is also independent of the concentration of the hydrolyte it seems probable that the neutral salt action is dependent on physical rather than on chemical factors.H. M. D. A New and Simple Method f o r Comparing Molecular Weights. I. PHILIP BLACKMAN (J. Physiccd Chem. 1911 15 866-868. Compare Thovert Compt. rend. 1902 134 564).-The method is based on Riecke’s theory which requires that the rate of diffusion should be inversely proportional to the square-root of the molecular weight. When solutions of equal (weight) concentration of the two substances the molecular weights of which are to be compared are connected by a tube of pure solvent the position in the tube a t which the diffusing solutes meet enables their relative diffusion velocities and hence their relative molecular weights to be calculated. A suit- able apparatus is described for use with pairs of solutes which give a coloration or precipitate when they meet in the tube.The result is independent of the state of ionisation of the solutes. R. J. C. Determination of the Molecular Weight of Crystalline Substances GUSTAV TAMMANN (Bey. 191 1,44 3618-3628).-From a consideration of the pressure-temperature ( p 5”) curves the author divides crystalline substances into two classes ~v hich are distinguished by use of the term ‘‘crystal group.” Substances of the first class which correspond with monotropic substances form only one crystal group but to each crystal group there belong a stable and a series of unstable thermally different forms. Substances belonging to the second class which correspond with enantiotropic substances form two crystal groups and each of these groups consists of a stable and a series of unstable forms.Comparison of the association of different liquids as measured by the Eotvos-Ramsay and Shields method shows that the normal liquids on crystallisation give only one crystal group whereas the associated liquids often give two crystal groups. From t h i s the author draws the conclusion that the molecular weight of normal liquids does not alter during crystallisation and consequently that gas liquid and crystals have the same molecular weight. Further considerations lead him to the conclusion that. the molecular structure of a crystal deposited from an associated liquid is simpler than that of the liquid itself. Rules are given by means of which from a consideration of the specific volumes and heats of fusion of unstable forms the crystal group to which they belong may be determined. T. s. P.ii. 150 ABSTRACTS OF CHEMICAL PAPERS. The Effect of Continued Grinding on Water of Crystal- lisation. C. E. GILLETTE (Chem. News 1922 104 313-314. Com pare Bleeker Abstr. 1910 ii 235).-0n continued grinding of 3 grams of the salt in an agate mortar barium chloride lost from 0.01 to 0.51% of water the loss being lessened by recrystallisation. The loss is chiefly due to superficial water which is again taken up when the ground salt is exposed to air. Potassium and ammonium alum lost over 2% of water on continued grinding. Strontium chloride lost 1 *5-2% borax 3*3-3*7% manganous chloride 1*92% and disodium hydrogen phosphate 28-30% of water. 0. H. D. An Electrical Laboratory Furnace Wound with a Non- noble Metal. LEO UBBELOHDE (Chem. Zeit. 1911 35 1403-1404). -The disadvantage of a furnace wound with nickel wire is that after a time the nickel changes its structure and readily oxidises. The author prevents this oxidation by imbedding in charcoal the wire which is previously protected with a fire-resisting but porous insulating layer made from kaolin alumina and asbestos. A t tbe temperature of the furnace a sn?all quantity of the charcoal burns giving a mixture of carbon monoxide and dioxide which protects the wire from oxida- tion. An arrangement is made for adding charcoal from time to time to replace that which burns away. The wire used in the furnace is an alloy (the composition is not given) the resistance of which a t 1000" is only slightly higher than a t 20'. The furnace may therefore be directly connected with the heating voltage without i t being necessary to insert rheostats and thus offers many advantages over furnaces supplied with platinum resistances (Herseus furnaces). T. S. P. Two Simple Forms of Gas-Pressure Regulators. EDGAR STANSFIELD (Trans. Fayaday Soc. 1911 7 116-118).-A glass cylinder open below is fixed in an outer vessel containing water. A beaker moves easily within this as a float and carries a glass rod at the upper end of which is in one form of apparatus a glass bulb in the other a cap to serve as a valve. The seating for the bulb is provided by an outer short glass tube with ground top whilst the cylindrical cap of the second form nearly fits the inlet tube. The delivery pressure is equal to the difference of level in the inner and outer vessels when the valve is just on the point of closing. The first form is the more sensitive but is occasionally liable to set up a continuous vibration. C. H. D. A New Funnel. 111. PHILIP BLACKMAN (Chem. News 1911 104 312. Compare Abstr. 1911 ii 796 1081).-The funnel has a wide lower aperture and no stem so that the cone of the filter-paper projects freely. C. H. D. Exact Reading Device for the Mohr-Westphal Balance. VON HEYGENDORFF (Chern. Zeit. 1912 36 5-6).-1n order to facilitate bringing the point of the swinging arm of the specific gravity balance into coincidence with the stationary point two glass microscope slides,INORGANIC CtfEMJSTRY. ii. 151 each with an etched horizontal line are attached to the stationary arm by rubber bands so that the swinging point lies between thern. C. H. D.