89 General and P h y s i c a l Chemistry. DEWAR formed by the Spectrum of Magnesium. By G. D. LIVEING and J. (Proc. &t~. ,!!oc., 44, 241-252) -When an electric arc is between magnesium electrodes,' most of the lines produced spark discharge are observed. The larger number of lines with an arc discharge may be due not to lowness of temperature, but to the greater mass of incandwcent matter, and to a wider range of tem- perature a t different portions of the discharge, recombinations occur- ing a t its edge. The electiic discharge itself may also give rise to vibrations distinct from those due to heat. The seven bands in the green are due to the oxide, as they are only prodiiced in the presence of oxygen or its compounds. If a piece of hurnt magnesium wire be heated in the oxphydrogen flame, the spectrum of magnesium is produced, the met:illic lines appearing if the hydrogen is in excess.The triplet near M which is produced when magnesium is burnt, is found t o be produced in the arc befween rnagnesium electrodes and in many other cahes when oxygen is present, but not in an atmosphere of nitrogeu o r hydrogen, hence it is due to the oxide. Vacuous tube8 are found to be very untrustworthy for the ultra-violet spectra, as the water-spectrum and lines of nitrogen are nearly always present, and the spectra sometinres vary unaccountably. A pump is described in which rubber connections and free contact of the mercury with air are avoided. H. K. T. Ultra-violet Spectra of Nickel and Cobalt. By G. D. LIVEING and J. DEWAR (Proc. Rot/.Xoc., 43,43O).-A comparison is made between a plane Rowland's grating with a goniometer and the concave grating (20 feet focal length) used by Bell. The results agree very closely, the concave grating gives more light, a i d is more suitable for com- plicated spectra, as the overlapping spectra of different orders are not all in focus at once. The coinci- dences are not greater than the theory of chances would allow, aud do not correspond with their chemical relationship. H. R. T. Ultra-violet lines of cobalt and nickel are compared. Two-fluid Cells. By C. R. A . WRIGHT and C. THOMPSON (Proc. Roy. Soc., 43, 489-493).-Cells are set up consisting of platinum plates in acid and alkaline solutions, with the further addition either of oxidising agents to the acid solution or of reducing agents to the alkaline solution.Currents are produced, in the first case with evolution of oxygen from the alkaline solution, in the second with evolution of hydrogen from the acid solution. The quantity of gas evolved was equivalent to the current. The acid and alkali were sul- phuric acid and potassium hydroxide respectively ; the oxidising agents being potassium permanganate, dichromatc, and ferricyanide, ferric chloride, and solutions of chlorine and bromine and the reducing VOL. LYI. 7b90 ABSTRACTS OF OHEMTCAL PAPERS. agents sodium hyposulphite, pyrogallol, cuprous chloride, and ferrous sulphate and ammonium chloride in nmmouiacal solution. Hydrogen was not evolved with sodium sulphite or hypophosphite, potassium ferrocyanide, or manganous hydroxide in ammoniacal ammonium chloride, nor was oxygen with barium dioxide and sulphuric acid, o r with hydrochloric acid and iodine.On the other hand, an aeration plate of platinum sponge gave a current four times as great. Plates of oxidisable metals in alkaline solution could be substituted for the reducing substance, hydrogen being evolved in the acid solution ; this was particularly the case when potassium cyanide was used. Gold, silver, and palladium in cyanide solution gave hydrogen, but platinum and iron were ineffective. When both oxidising and reducing agents are used, comparatively powerful currents are produced. Effect of Chlorine on the Electromotive Force of a Voltaic Couple. By G. GORE (Proc. Roy. Soc., 44, 151--152).-1f the electromotive force of a small magnesium-platinum couple in distilled water is balanced through a galvanometer and dilute chlorine-water is gradually added, the electromotive force does not alter a t first, but after a cert,aiii point has been reached (1 in 17,000 millions) it begins to increase rapidly.I n t h i s way, the one ten-thousand-millionth of a grain of chlorine in 0.1 C.U. of water can be detected. Other electrolytes give the same reaction, b u t require a larger quantity of dissolved substance. H. K. T. H. K. T. Development of Voltaic Electricity by Atmospheric Oxida- tion. By C. R. A. WRIGHT and C. THOMPSON (PYOC. RUy. Xoc., 44, 182- 2OO).-The electromotive force of cells in which aeration plates are used, falls off very rapidly if the current density exceeds a certain amount.When oxidisable liquids are used, it is difficult to determine, as it appears to vary with the nature of the aeration plate, and also with the incorrodible plate in the liquid to be oxidised. For determin- ing these electromotive forces, an arrangement is used in which t5e asration plate can be kept undisturbed, and in which the oxidisable substances are protected from alterations of temperature, impurities from the air, &c. After a few hours or days, the currents become constant. In these cells, variation of the asration plate produces a difference in the electromotive force independent of the oxidisable plate used ; similarly the effect of varying the metal is independent of the asration plate. The nature and strength of the liquid affects the results to some extent.The electro- motive force actually generated falls very considerably short of that corresponding with the chemical changa, especially when the current density is large. With silver as the oxidisable plate, however, the electromotive force is higher than the theoretical, this being due to the high negative value of the thermovoltaic constant of silver in contact with sulphurib acid. When oxygen was substituted for air over the aGration plate, a slight rise in the electromotive force was observed. With aeration plates immersed in coal-gas or hydrogen, and opposed to a platinum plate in alkaline permanganate or in sulph uric acid and potassium dichromate, very weak and variable Tables of results are given.GESERAL AND PHYSICAL CHEMISTRY.9 1 currents were observed. aGration plates in hydrogen and air respectively. The same was the case with cells formed of H. K. T. Electrolytic Conductivity of Rock Crystal. By E. WARBURG and I?. TEGETMEIER (Ann. Phys. Chem. [a], 35, 455-467).-Jn a, former paper (ibid. [2], 32, 447), the authors showed that a slice of rock crystal cut perpendicularly to the principal axis, and having its ends covered with layers of gold or plumbago, when subjected at a temperature of about 230" to a long-continued E.M.F. of considerable intensity, had its conductivity permanently reduced to a small Rztctioti of its original amoiint. In directions perpendicular to the axis, rock crystal, even at higher temperatures, has little or no conductivity. As the result of their further investigations, the authors have arrived at the conclusions that- (1.) The electrolytic conductivity of rock crystal in bhe direckion of the principal axis is, at high temperatures, about the same as that of ordinary glass.(2.) When a slice cut perpendicularly to the axis is electrolysed, sodium-amalgam being used as the anode, sodium migrates through the slice, its amount being in accordance with Faraday's law, and the weight of the slice remains unchanged. (3.) Even at high temperatures, rock crystal acts as a good insulator with respect to an E.M.F. in a direction perpedicular to the prin- cipal axis. When sodium-amalgam was used as the anode in am experiment lasting for three days, at a temperature of 250°, 88 milligrams of silver were deposited in a silver voltameter in the circuit, and the only substance which could be detected at the cathode was sodium.When potassium was used in the place'of sodium, it was found that after 40 hours the current had sunk to about the hundredth part of its original value, only 2 milligrams of silver were separated, and no potassium could be detected ah the cathode, even by means of the spectroscope. The authors therefore conclude that the conductivity is due to the presence of sodium, in the form of Na,SiO,, which was shown, by an analysis specially made by Baurnann, to be present in the crystal employed in the proportion of 1 part in 2300, so that the crystal might he regarded as a ver,y dilute solution of this salt. The electrolytic character of the conductivity was further COU- firmed by the fact that a cell giving an E.M.F.of from 1.2 to 2 volts could be formed of mercury, a slice of quartz at a temperature of 225", cut perpendicularly to the axis, and sodium-amalgam. According to Clausius's theory of electrolysis, the fact of electrolytic conduction only taking place in the directioa of the principal axie would tend to the inference that in the case of rock crystal not traversed by an electric current, the interchange of atoms between the molecules can only take place, at any rate to a sensible extent, in the direction of the principal axis. A confirmation of this inference is found in the fact first noted by Salm-Horstmar (Ann. Phys. Chem., 120, 334), that the action of hydrofluoric acid on rock crystal is much greater in the direction of the axis than perpendicular to this axis.The authors have themselves h 292 ABSTRACTS OF CHEMICAL PAPERS. made experiments to test the truth of this statement, and the results are in agreement with those of Salm-Horstmar. It would appear from the results obtained in the paper, tllnt the silicate Na2Si03 contained in the crystal must partake of its crystal- line structure. G. W. T. Effect of Occluded Gases on the Thermoelectric Properties of compounds. By J. MONCKMAN (Proc. Ro?y. SOC., 44, 220-236). -When a portion of platinum o r palladium wire is charged with hydrogen by electrolysis, and the wire afterwards heated, a cui-rent passes from ithe protected to the unprotected part. The same occurs with rods of carbon after charging and pressing together, the current passing from the hydrogen to the oxygen.The wires and rods are found to have an increased resistance, that of the oxygen rod being the greatest. The effect disappears after short circuiting. If the wires or rods be charged twice in opposite directions, the effect dis- appears, unless thc second charging is of very short durafion; in fhis case,'% reversal takes place. With carbon rods a t different tem- peratures in contact, reversal occurs a t 250" ; with a thermoelectric couple of carbon and pla,tinum, the thermoelectric line rises below 250', and falls above that temperature. The rate of decrease of resist- ance of carbon diminishes as the temperature rises to 250", but increases afterwards. The rate of evpansion increases as the tem- peratiirc rises to 250", but afterwards decreases.The specific hest increases fairly regularly up to 250°, but above that temperature falls to half. H. K. T. Electrochemical Effects on Magnetising Iron. By T. ANDREWS (Yroc. Roy. SOC., 44, 152-168).-A niagnetised and an nnmagne- tised bar of iron or steel are immersed in different reagents, and the current produced noted. The amount varies considerably, but is large in the case of bromine, salts of copper, and nitric acid. The result is dependent both on the strength of the solution and the degree of magnetisation. With powerful oxidisers, the magnetised bar is generally electropositive, but becomes electronegative with sulphuric acid, dilute hydrocbloric acid, and ferric chloride and chlorine. In the laut-named instances, the effect may be due to the diamagnetic properties of the solutions, or of the gases evolved. With ferric chloride alone, the magnetised bar is electropositive, with chlorine electronegative, with the two together, electronegative until the chlorine is exhausted, when it becomes electropositive.In the same bar, local currents are produced from the more magnetised portions to the less. These may cause the magnetised bar to be acted on to a greater extent than the unniagnetised. In strong nitric acid, a current is produced from the magnetised to the unmagnetised bar. Specific Heat of some Solid Organic Compounds. By H. HESS (Ann. Yhys. Chem. [Z], 35, 410--429).-The author states that, with the exception of some investigations by De Heen (Bull acnd. roy. brlg., 5) and A.Batt'elli (Atti R. Id. Veneto [GI, 3), he has not been able t o find any account of investigations of the specific heats of solid Experiments were also made with graphite rods. H. I(. T.GENERAL -4hTI.l PHYSICAL CHEMISTKY. 93 organic compounds, and he therefore undertook the present inresti- gation with a view especially of determining t>he manner in which the specific heats of solid organic substances depend on temperature. The author gives a number of curves showing the relation between specific heat and temperature in the substances exauiined, tempera- tiires being taken as ordinates, and the corresponding specific heats as abscissae. The curves he fiuds to be sensibly straight lines intersecting the specific heat axis above the zero point, so that the specific heat Name of substance.{ Oxalic acid.. . . Malonic acid . . Succinic acid . . Isosuccinic acid Glutaric acid (solid) (liquid) { Gtlutaric acid Py rot artaric acid nic acid { Dimethylmalo- { Sugar .. .. .. .. Benzoic acid (solid) (liquid) { Benzoic acid Phthalic acid . . {- ~~~ Mean specific heat. Tempera- ture limits. -- 0" to 50" 0 )7 75 0 7 7 94 0 77 50 0 7 7 94 0 7) 110 0 7 7 50 0 Y, 94 0 97 75 0 77 94 0 7, 50 0 7 ) 75 0 )) 94 0 ,) 99.3 0 7 , 50 0 >, 75 0 ,7 94 0 ), 105 0 7 7 50 0 7) 94 0 ,7 75 0 7, 94 0 ,, 150 J > 50 0 ,) 130 0 ), 113 0 ,) 130 0 9 , 50 0 ,, 94 0 7 7 130 0 ,) 122 0 ), 136 0 ., 75 0 7 . 119 0 ,7 150 C. 0 -3359 0 -3575 0 -3728 0 *2832 0 -3131 0.3262 0.2898 0 -3252 0 -3650 0 -3378 0.3500 0 *3ti36 0 *3081 0 *3207 0 *3461 0 * 7503 J 0 *3098 0 *3267 0 -3548 0-3575 0 -3996 I 0 34741 0 -3037 0 * 3197 0.3337 0.3511 0 *25? 1 0 -3118 0 *3319 0 '50'72 I 0-5256j 0 -2559 0 -2862 0 *3099 0 9285 1 Tempemtm coefficient.Cempeiature limits. 50" to 75" 50 77 94 50 ,) 94 94 ,, 110 50 ), 110 50 94 94 7 7 150 50 ,) 150 75 ,, 94 50 )) 75 75 ,7 94 75 y 7 94 50 7 7 94 50 ,, 75 50 7 7 94 99 7 7 130 50 7 7 75 75 ,) 105 50 7 7 94 50 7 7 106 50 77 94 75 7 ) 94 75 ,) 113 94 ,) 130 75 ), 130 50 7 7 94 50 ,) 110 122 ,) 136 94 ,) 110 75 ) ) 119 119 ,) 150 75 ,, 150 6. 0 *000864 0 -00801 0 *000839 0 *000680 0 *000771 0.000705 0 -000805 0 ~000711 0.0007ti2 0 *000512 0 -000716 0 *000600 0 -000504 0 '0913.3'7 0-000864 O~OoO7lO 0 so0676 0 *001027 0 - 000796 0 *000867 0 -000859 0 - 000842 0 *OW789 0 *OW872 0.000862 0 -00124 0.00126 0 -00165 0 -00131 0 -000689 0.000764 0 *000720 Means.b = 0 '000835 6 = 0 *000719 b = 0 *000759 i i 1 6 = 0 400609 b = 0 .000901 b = 0 *000;10 b = 0 '000842 1 b = 0*000859 I 6 = 0 *000841 I b = 0 '00125 b = 0 '00131 6 = 0 '000724 194 ABSTRACTS OF CHEMICAL PAPERS. can be represented by a formula of the form n + bt. The results obtained are given in tabular form (p. 93), c representing the mean spezific heat between tihe temperature limits indicated, and b the temperature coefficient. The values obtained by assuming the true specific heat to be repre- sented by a forniula of the form cc + bt are given in the second table, under the head of "observed specific heat," the column headed " calculated specific heat " being calculated from Kopp's law, that the molecular heat of a body i s equal to the sum of the atomic heats of its constituents.The atomic heats of carbon, hydrogen, and oxygen respectively are taken as 1.8, 2.3, and 4.0. The column headed t gives the temperature a t which the observed and calculated specific heats are equal, and it will be seen that with the exception of oxalic and isosuccinic acids, the different substances obey Kopp's law for some temperature within the limits 3-5" and 50". Kopp's law might be generalised if we could assume the specific heats of carbon, hydrogen, and oxygen to be functions of the tem- perature, but this would not lead to correct general formulae, for Regnault (Compt. rend., 26, 311) and E, Wiedemann ( A m . P h y s . Chem. 15 7, 1) have shown that the specific heats of hydrogen and oxygen are sensibly iudependeut of the temperature, and although H.P. Weber has shown (Ann. Phys. Chem., 147,362) that the specific: heat of carbon increases considerably with the temperature, this increase would not be sufficient to account for the observed increase in the temperature coefflcient. Name of substance. Oxalic acid.. ................. Malonic acid.. ................ Succinic acid ................. Isosuccinic acid ............... Gtlutaric acid (solid) ........... I? yrot art aric acid. ............. Dimethylmalonic acid ......... Sugar.. ...................... Rerizoic acid.. ............... Plithalic acid.. ............... Benzoic acid.. ................ Glutaric acid (liquid) .......... Specific heat. Calculated.0 *2689 0.2942 0 -3136 0 - 3 h 8 0 *3!!40 0 -2820 0 -2602 ¶) - - 0 bserved. 0-2941 + 0-00167t 0'2473 + 0.001445 0.2518 + 0-00152f 0.3067 + 0-00122t 0 -2620 + 0.00180t 0.2677 + 0'00168t 0.2666 + 0'001725 0.2387 + O'OCI173t 0.1946 + 0.002505 0.2016 + 0-00145~ 0 -6580 + 0 -00142t 0.3474 + 0.002625 t. - 15 -1' + 32 -6 + 40.7 + 5-7 f 37 -1 + 36 *4 + 36 *2 + 49 -3 + 35-0 +40*4 - - - The author's results show that there are often considerable differ- ences in the specific heats of different isomeric compounds. Evolution of Gases from Homogeneous Liquids. Ry V. H. VELEY (Proc. B o y . Xoc., 44, 239--240).-The addition of fineiy divided substances is found to increase the rate of evolution of gases from liquids in which they are formed. When the temperature G. W. T.GENERAL AND PHYSICAL CHEMISTRY.95 remains the same, the rake of evolution rises slowly until a maximum is reached, which is maintained for some time. The rate then decreases proportionally to the diminution in mass. The phenomenon repeats itself when the temperature is lowered and then raised t o its former point, and also when the pressure is suddenly increased. The reduction of the pressure to a fraction of an atmosphere produces no permanent effect. The rate of decomposition of formic acid into carbonic anhydride and water is also examined, and is found t o agree with the equation log (T + t ) + log r = log c, where T is the time from the Commencement of the observations, t the interval of time from the moment of commencement up to the moment at which the time required for unit change is mil, r the mass at the end of each observation, and c a constant.The curve of rate of change conforms with the law drld7 = - r2/c, which expresses the rate at which equivalent masses react on one another. Hence it is presumable that equivalent masses react, and that the change is represented by the equations HCOsOH + HCO*OH = HCO-OCHO + H,O and HCO.O*CHO = 2CO + H20, a reaction similar to the production of ethyl formate from formic acid and alcohol. Properties of Matter in the Gaseous and Liquid State under Various Conditions of Temperatura and Pressure. By the late T. ANDREWS (Ann. Chirn. Pl~ys. [GI, 13, 411-432).-Regnault (Xem. Acad. Sci., 26, 680-696) made a series of experiments to determine the tension of a mixture of a gas and a vapour, such as nitrogen or air, and the vapour of water or some more volatile liquid, and came to the conclusion that Dalton’s law of partial pressures may be con- sidered theoretically correct in the case of such mixtures, and that probably this law could be proved to be correct experimentally if the mixture of gas and vapour could be enclosed i n a vessel the interior surface of which was composed of the volatile liquid.He also found that, under pressures varying from + to 2 atmospheres, the com- pressibility of a mixture of ordinary gases, such as air and carbonio anhydride, hydrogen and sulphurous anhydride, was intermediate between that of each gas separately for the same variations of pres- sure (ibid., 258). The results of all experiments which had been carried out up to the time when the author’s investigations were commenced, had been to show that, with one exception, Dalton’s law i s true in all cases for mixtures of gases or vapours, or at any rate in the case of gases and vapours which exert no chemical action on one another.A mixture of the vapours of two mutually soluble liquids, in presence of the two lcquids mixed or dissolved, constitutes, however, an important exception to this law, because of the disturbing influence of the chemical affinity of the liquids. But as, up to this time, no experi- ments had been carried out, to prove the truth of Dalton’s law under pressures greater than 2 atmospheres, the author investigated the change in volume of a mixture of 3 vols. of pure carbonic anhydride and 4-05 1-01s.of nitrogen at temperatures above and below the critical temperature of carbonic anhydride, the pressure employeci vcrying between about 40 and 300 atmospheres. H. K. T.9 8 ABSTRACTS OF OH'EJllCAL PAPERS. From the results, which are given in tabular form, curves are drawn showing the volume of tlie mixture at the various temperatures and pressures. These curves are all very similar, showing no differ- ence in character for temperatures above or below 31". If it be granted t h a t Dalton's and Boyle's laws are true in the case of nitrogen under the pressures employed, the curves showing the change in volume of the carbonic anhydride in the mixture under tlie various conditions of temperature and pressure prove that below 31" t h i s gas tends to occupy the volume corresponding with the liquid state, although the curves are quite different from those of carbonic anhydride alone.It follows, therefore, that Dalton's l a w is no longer applicable in this case, and is only strictly true of a perfect gas. As no liquefaction took place in any of the above experiments, showing that the presence of nitrogen lowered the critical point OF the carbonic anhydride, the author investigated this phenomenon more fully. A mixture of 6.2 vols. of carbonic anhydride and 1 vol. of nitrogen was placed under a pressure of 48.3 atmospheres ; no con- densation occurred until the temperature was lowered to 3.5". As the pressure was increased the volume of the liquid augmented, and after each increase of pressure, the volume continued to augment slowly for some time; for example, under a pressure of 82 atmo- spheres the relative volumes of the gas and liquid were at first 8.5 and 5.8, but, the apparatus having been left for some time, thcb volume of the liquid slowly increased.The pressure having been then raised to 102 atmospheres, the volume of the gas which was a t first 1.7 diminished gradually until only a small globule remained, which finally disappeared entirely, the nitrogen dissolving in the liquid carbonic anhydride. In a second experiment, with the same mixture at a higher and constant temperature, the liquid had a t first its usual concave surface, and as the pressure was increased, the volume of the liquid also augmented without any noticeable change in the appearance of the concave surface ; on further increasing the pressure, the surfaca of separation appeared in section as a fine line, but when the pressure was again increased, i t disappeared entirely, the whole becoming homogeneous.The position in the tube, occupied by the surface of separation, depended on the temperature a t which the observation was made ; a t 14" the liquid filled about, two-thirds of the entire space a t the very moment when the surface of separation was about to disappear, The critical temperature of a mixture of 1 vol. of nitrogen and 3.43 vols. of carbonic anhydride was found to be 14", and the cor- responding pressure 98 atmospheres. Experiments with this mixture showed that at 6.3" no condensation took place until the pl:essure reached 68.7 atmospheres ; the liquid then disappeared under in- creased pressure, but reappeared when the pressure reached 113.2 atmospheres.At 9.Y0, the liquid first appeared when the pressure reached 77.6 atmospheres ; after having disappeared i t was again formed under a pressure of 107.8 atmospheres. At 13*2", the liquid appeared under a pressure of 91.6 atmospheres, disappeared as the pressure was increased, and reappeared when it attained 103.2 atmo- spheres, If the mean of the two pressures for each of the aboveGESERAL AXD PHYSICAL CHEMISTRY. 97 temperatures is taken, the critical pressure a t 6*3", 9*9", 13 2", and 14" is found to be 90.9, 92.7, 944, and 98 atmospheres respectively. I n the course of these experiments, the author found it convwient to employ a tube bent twice at right angles.When the gaseous mixture was compressed below the critical point, the liquid carbonic: anhydride collected in the lower portion of the tube, although part of the liquid was first formed a t the surface of the mercury ; but the whole of the liquid soon collected at the bottom of the tube. In some experiments, the carbonic anhydride liquefied a t temperatures above 20°, and sometimes no condensation took place even a few degrees below this temperature. This phenomenon was found to be owing to the fact that when liquefaction had taken place, if the pressure was diminished so that the mixture could become completely gaseous, the liquid separated into two portions, one rich, the other poor, in carbonic anhydride. The portions of the tube which had been previously occupied by the liquid then contained a large excess of carbonic anhydride, especially when the tube had been previously cooled to -lo", so that almost the whole of the carbonic anhydride had been liquefied. If the pressure was reduced so as to bring the whole of the liquid to the gaseous state, the temperature being a t the same time raised to 26", it was found that the carbonic anhydride could be liquefied by pressure alone (at 26"), provided that, the experiment was performed without loss of time.When, however, the mixture was left for some time in the gaseous state, diffusion gradually took place, and the temperature at which liquefaction wa9 possible decreased accordingly. Diffusion was not complete until after some hours, and then increased pressure caused no liquefaction until the temperature was reduced to 14".This method of separating the gases was employed to shorn the effect of diffusion as follows :-A mixture of carbonic anhydride and nitrogen was kept at 8.5" under a pressure of 46.4 atmospheres until diffnsion was complete ; the volume of the mixture was then 162.2, After liquefying the carbonic anhydride by employing great pressure and lowering the temperature to -12", the temperature was again raised to 8*5", and the pressure brouqht back to 46.4 atmospheres; the volume was then found to be 159.5, showing that a contraction of 2.7 ~01s. had taken place owing to the separation of the mixed gases. A t the end of 1+ hours the volume had increased to 161.5 in conse- quence of partial diffusion. In a second expeiiment at 16", under a pressure of 47.9 atmospheres, the original volume of the mixture was 164.6, but, after liquefaction, only 161.9 when brought back to the initial temperature and pres- sure ; after 1; hours the volume had increased to 164.1.I n a third experiment a t 20", under a pressure of 46.4 atmospheres, the volume decreased from 175.8 to 173.5 after the separation of the gases. These results show that when the two gases d i f i s e into one another under great pressure, an increase in volume ocmrs, and when they are separated the volume is diminished. This change in volume undoubtedly occurs also under ordinary pressures, but the variation would probably be so small that it would be dif€icult to detect experi- mentally. F. S. K.98 ABSTRACTS OF CHEMICAL PAPERS.The Behaviour in Relation to Boyle’s Law of certain Gases at Low Pressures. By F. FUCHS (AN%. Phys. Chem. [‘L], 35, 430- 450).-The author, from the results of a series of experiments on atmospheric air, carbonic and sulphurous anhydrides and hydrogen, arrives a t the following conclusions :- (1.) At ordinary temperatures, Boyle’s law does not represent a limiting state towards which a gas approaches indefinitely with increasing rarefaction, but a t pressures respectively above and below a certain amount, the deviations from Boyle’s law are respectively positive and negative. The limits of pressure within which Boyle’s law holds are indefinitely small, as any finite change in volume will alter the forces betweeii the gaseous molecules. (2.) I n the case of atmospheric air a t the temperature 0”’ a change of sign of this kind takes place at a pressure very slightly below the ordinary atmospheric pressure.If any similar change of sign occurs witb carbonic and sulphurous anhydrides, it must be a t pressures less than any at which the author’s observations were made. (3.) The deviations from Boyle’s law in the case of hydrogen at low pressures are so small that hydrogen under these circumstances may, without, sensible error, be regarded as a perfect gas. G. W. T. Constitution of Solutions. By F. R~DORFF (Ber., 21, 3044- 3050).-Snlts of the composition R2S04 t R“S04 + 6H20 and R,SOd + R,“’(S04)3 + 24H20 are partially decomposed into their constituents when dissolved in water (compare Abstr., 1888, 342).Hydrogen potassium sulphate behaves similarly, but hydrogen ethyl sulphate diffuses unchanged. 3KZCzO4,Fe2(C2O4), + 6H,O ; 3Na2CzOd,Fe,( C204), + 6Hz0 ; 3K,C,Oa,Cr,(C2O4), + 6H20 ; 2(NH4)HC2O4 + H,O ; PU’nHC4H406 + H20, and 2K(SbO)C4H40, + H20, diffuse unchanged, but (NHa)HC,04,C,H204 + 2H,O is partially de- composed into oxalic acid and hydrogen ammonium oxalate. Solutions of potassium chromate, potassium dichromate, and sodium dichromate diffuse unchanged, but the salt (NH4)2Cr04,MgCrOl + 6H20 The following salts :- is partially decomposed when dissolved in water. ‘l’he following salts :-2NaCl,PtCI, + 8H20 ; 2KCl,PtCIz ; 2NH4C1,HgC12 ; Ba(CN),,Pt(CN), + 4H20, and all double cyanides are true molecular compounds, but KCl,Hg( CN), is partially resolved into its constituents when dissolved in water (Zoc.cit.). NaH2POa and Na2HP04 diffuse unchanged: Na,P04, on the con- trary, is partially decomposed. The three sodium salts of citric acid are not decomposed in aqueous solution. F. S. K. Physical Properties of Colloi’d Solutions. By C. L~DEKING (Ann. Phys. Chenz. [2], 35, 552--557).--In a paper with Wiedemann (Abstr., 1885, 1032) it, was shown that the vapour-pressure of aGENERAL AND PHYSICAL CHEJIISTRY. 99 40 per cent. aqueous solution of gelatin was less a t a temperature of 40" than that of pure water. According to Guthrie (this Journal, 1877, i, 36), a 40 per cent. solut'ion of gnm boiled at 98", and a 45 per cent. solution of gelatin at 97.5": results which were in contra- diction to those above mentioned.With a view of discovering the reason of the discrepancy, the author made experiments on solutions of gum arabic, gum trngacan th, dextrin, starch, and agar-agar. He finds that a 40 per cent. solution of gum arabic boils at IOO", but carbonic anhydride begins to be given off a t a temperature of about go", and a t a somewhat higher temperature gives the appearance of boiling to the solution. The other solutions also boiled at loo", although in the case of gelatin boiling began with the thermometer a t 98", which, however, the author attributes to the liquid not rapidly assuming the same temperature throughout, owing to its viscidity preventing the forma- tion of convection currents, This opinion was based on the fact that the thermometer did not remain at 98", butpadually rose to 99.8", where it remained constant.The author found that the addition of the colloid in every case slightly lowered the vapour-pressure, and, as he points out, the presence of a solid i n solution could not possibly increase the vapour-pressure. For example, if the steam given off at 98" from a gelatin solution had a pressure of 760 mm., i t would necessarily re- condense to water and mix again with the solution. When solutions of gum or gelatin are cooled considerably below zero, the author finds that they do not solidify as a whole, as stated by Guthrie, but ice crystals gradually separate out. He finds that gelatin has a strong condensing action on the water of solution. G. W. T. Precipitation of Colloid Substances by Salts. By 0.NASSE (Pjiiyer's Archiv, 41, 504--514).-A11 prote'ids except peptone can be precipitat!ed by saturating a neutral solution with ammonium sul- phate, some more easily than others, for instance, globulins more easily than albumins. Other salts bave the same power, but none act so readily as ammonium sulphate. This property, however, is not characteristic of proteids : soaps, gelatin, and certain fioluble carbo- hydrates (glycogen, amidulin, inulin, &c.>, are similarly precipitated ; it in fact seems to be a property common to colloid substances. The question arises, on what does the difference in the concentra- tion of the salt necessary to produce precipitation depend ? Is the action of the salt simply due to a struggle of the molecule of proteid, gelatin, &c., with that of the salt for water, and t h a t the pre- cipitation of the colloid substance occurs as soon as its water-attract- ing power is exceeded by that of the salt? In order to determine whether this is the case, one m u d ascertain the amount of two or more different salts necessary to precipitat'e the same aniount of one colloid substance, the necessary concentration of the salt solutions corresponding with a : b : c, &c. The same ques- tion is then invebtigated €or another colloid substance, and the ratio100 ABSTRACTS OF CHEMICAL PAPERS.a' : b' : c', &c., found. for water, a : b : c, &c., will be found = a' : b' : c', &c. Tf hhen we have only to deal with attraction The following table illustrates the results obtained :- Collo'id substances.Gplatin ......... The solution contained in 100 C.C. the following amounts of salt w18cn precipitation began. L--v---J a. 6 . Ammonium Magnesium sulphate. sulphat e. a : b. 12.4 14.8 0.84 White of egg.. .... 20.2 19.6 1-03 9 3 ...... 18.5 19.3 0.95 Serum prote'rds .... 17.4 18.5 0.94 Albumose. ........ 12.7 13.6 0.9:3 ,, ......... 14.9 17.6 0.85 Amidulin. ........ 20.9 10.5 1-99 Glycogen-dextrin . . 44.7 23.7 1-99 The differences of the numbers in the last column show that water-attracting power is not the only influence at work, but some other relation must exist between the colloid and the salt. Still it is possible that it may explain some of the precipitations, especially that of gelatin. Gelatin loses many of its characteristic properties after the prolonged heating of its solutions; it, for instance, no longer gelatinises on cooling, and its water-holding power is greatly in- creased, yet, the ratio a : b = 0.84 remains constant for the gelatin in all the different stages of this change.With regard to the proteids, in which such wide differences occur, it is thought probable that the explanation lies in the fact that loose compounds with the salts are formed. The paper concludes with some remarks on the influence of tern- perature in determining precipitation by salts. New Formula for Calculating the Molecular Volumes of Chemical Compounds at the Boiling Points. By J. A. GROSHANS (Rac. l'rav. Chim., 7, 220--225).-The molecular volume of a sut)stance, CPH4Or, at the boiling point may be represented by t7, = a + 1O(p + q ) - 7.28 B for a fatty compound, or by the same expression miiius 15 for an aromatic compound, c1 being the number of' O.C.equal to the molecular weight of the compound, and B = p + q + Y. Both these formulte obey Kopp's rules, that homologous compounds differing in their composition by CH, should differ i u their molecular volumes by 22, and that a fatty compound should have approximately the same volume as an aromatic compound which differs in its formula by C2 - Hq. With hydrocarbons, since 7 = 0, the formula may also be written (us - a)/B = 2-72 for fatty compounds, or (u, - a -+- 15)/B = 2-72 for aromatic, both of which are fonnd t o agree well with experiment. For halogen-deiaivatives, an addition of 15 must be made for each atom of halogen oontained.H. C. W. D. H.ISORGANIC CHEMISTRY. 101 Molecular Lowering of the Freezing Point of Benzene by Phenols. By E. PAT ERN^ (Ber., 21, 3178-3180).--The author made a, number of experiments to ascertain whether the fact that certain substances containing the hydroxyl-group produce an abnormal lower- ing of the freezing point of benzene WRS true of all substances, and whether this abnormal behaviour was sufficient proof of the pi*esence of the hydroxyl-group (compare Raoult, Abstr., 1884, 959). The re- sults showed that although phenol behaves in an abnormal manner, the following compounds : ethyl phenol, acetylphenol, two isomeric nitrophenols, tribromophenol, picric acid, paracresol, methyl salicylate, thymol, nitrothymol, nitrosothymol, a-naphthol, /-3-naphthol, and benzylphenol, all produce the normal lowering of the freezing point of benzene and of acetic acid, either in dilute or moderately concentrated solutions, the variations caused Ey change in concentration of course being taken into consideration.The molecular weight of water determined by Raoult's method in acetic acid solution was found to be 18 (compare Rxonlt, lor. cit.), but the author points out that tlhis result is not by any means conclusive, as, even if the molecules were originally more complex, they would be simpli6ed by the act of solution. Hentschel's experiments with acetic acid (Zeit. phys. Chenz., 2, 308) seem to point to an opposite conclusion, but in this case the freezing point of the benzene solution was considerably below that of acetic acid.The fact that the molecular weight of water is found to be 36 when the freezing point of the solution lies below 0" shows that the temperature a t which the mixture freezes is a most important factor in the case. F. S. K.89General and P h y s i c a l Chemistry.DEWARformedby theSpectrum of Magnesium. By G. D. LIVEING and J.(Proc. &t~. ,!!oc., 44, 241-252) -When an electric arc isbetween magnesium electrodes,' most of the lines producedspark discharge are observed. The larger number of lines with anarc discharge may be due not to lowness of temperature, but to thegreater mass of incandwcent matter, and to a wider range of tem-perature a t different portions of the discharge, recombinations occur-ing a t its edge.The electiic discharge itself may also give rise tovibrations distinct from those due to heat. The seven bands in thegreen are due to the oxide, as they are only prodiiced in the presenceof oxygen or its compounds. If a piece of hurnt magnesium wire beheated in the oxphydrogen flame, the spectrum of magnesium isproduced, the met:illic lines appearing if the hydrogen is in excess.The triplet near M which is produced when magnesium is burnt, isfound t o be produced in the arc befween rnagnesium electrodes and inmany other cahes when oxygen is present, but not in an atmosphereof nitrogeu o r hydrogen, hence it is due to the oxide. Vacuous tube8are found to be very untrustworthy for the ultra-violet spectra, as thewater-spectrum and lines of nitrogen are nearly always present, andthe spectra sometinres vary unaccountably.A pump is described inwhich rubber connections and free contact of the mercury with airare avoided. H. K. T.Ultra-violet Spectra of Nickel and Cobalt. By G. D. LIVEING andJ. DEWAR (Proc. Rot/. Xoc., 43,43O).-A comparison is made betweena plane Rowland's grating with a goniometer and the concave grating(20 feet focal length) used by Bell. The results agree very closely,the concave grating gives more light, a i d is more suitable for com-plicated spectra, as the overlapping spectra of different orders are notall in focus at once.The coinci-dences are not greater than the theory of chances would allow, auddo not correspond with their chemical relationship. H.R. T.Ultra-violet lines of cobalt and nickel are compared.Two-fluid Cells. By C. R. A . WRIGHT and C. THOMPSON (Proc.Roy. Soc., 43, 489-493).-Cells are set up consisting of platinumplates in acid and alkaline solutions, with the further addition eitherof oxidising agents to the acid solution or of reducing agents to thealkaline solution. Currents are produced, in the first case withevolution of oxygen from the alkaline solution, in the second withevolution of hydrogen from the acid solution. The quantity of gasevolved was equivalent to the current. The acid and alkali were sul-phuric acid and potassium hydroxide respectively ; the oxidisingagents being potassium permanganate, dichromatc, and ferricyanide,ferric chloride, and solutions of chlorine and bromine and the reducingVOL. LYI.790 ABSTRACTS OF OHEMTCAL PAPERS.agents sodium hyposulphite, pyrogallol, cuprous chloride, and ferroussulphate and ammonium chloride in nmmouiacal solution. Hydrogenwas not evolved with sodium sulphite or hypophosphite, potassiumferrocyanide, or manganous hydroxide in ammoniacal ammoniumchloride, nor was oxygen with barium dioxide and sulphuric acid, o rwith hydrochloric acid and iodine. On the other hand, an aerationplate of platinum sponge gave a current four times as great. Platesof oxidisable metals in alkaline solution could be substituted for thereducing substance, hydrogen being evolved in the acid solution ; thiswas particularly the case when potassium cyanide was used. Gold,silver, and palladium in cyanide solution gave hydrogen, but platinumand iron were ineffective.When both oxidising and reducing agentsare used, comparatively powerful currents are produced.Effect of Chlorine on the Electromotive Force of a VoltaicCouple. By G. GORE (Proc. Roy. Soc., 44, 151--152).-1f theelectromotive force of a small magnesium-platinum couple in distilledwater is balanced through a galvanometer and dilute chlorine-wateris gradually added, the electromotive force does not alter a t first, butafter a cert,aiii point has been reached (1 in 17,000 millions) it beginsto increase rapidly. I n t h i s way, the one ten-thousand-millionth ofa grain of chlorine in 0.1 C.U. of water can be detected. Otherelectrolytes give the same reaction, b u t require a larger quantity ofdissolved substance. H.K. T.H. K. T.Development of Voltaic Electricity by Atmospheric Oxida-tion. By C. R. A. WRIGHT and C. THOMPSON (PYOC. RUy. Xoc., 44,182- 2OO).-The electromotive force of cells in which aeration platesare used, falls off very rapidly if the current density exceeds a certainamount. When oxidisable liquids are used, it is difficult to determine,as it appears to vary with the nature of the aeration plate, and alsowith the incorrodible plate in the liquid to be oxidised. For determin-ing these electromotive forces, an arrangement is used in which t5easration plate can be kept undisturbed, and in which the oxidisablesubstances are protected from alterations of temperature, impuritiesfrom the air, &c.After a few hours or days, the currents becomeconstant. In these cells, variation of the asration plate produces adifference in the electromotive force independent of the oxidisableplate used ; similarly the effect of varying the metal is independent ofthe asration plate. The nature and strength of the liquid affects theresults to some extent. The electro-motive force actually generated falls very considerably short of thatcorresponding with the chemical changa, especially when the currentdensity is large. With silver as the oxidisable plate, however, theelectromotive force is higher than the theoretical, this being due tothe high negative value of the thermovoltaic constant of silver incontact with sulphurib acid. When oxygen was substituted for airover the aGration plate, a slight rise in the electromotive force wasobserved.With aeration plates immersed in coal-gas or hydrogen,and opposed to a platinum plate in alkaline permanganate or insulph uric acid and potassium dichromate, very weak and variableTables of results are givenGESERAL AND PHYSICAL CHEMISTRY. 9 1currents were observed.aGration plates in hydrogen and air respectively.The same was the case with cells formed ofH. K. T.Electrolytic Conductivity of Rock Crystal. By E. WARBURGand I?. TEGETMEIER (Ann. Phys. Chem. [a], 35, 455-467).-Jn a,former paper (ibid. [2], 32, 447), the authors showed that a slice ofrock crystal cut perpendicularly to the principal axis, and having itsends covered with layers of gold or plumbago, when subjected at atemperature of about 230" to a long-continued E.M.F.of considerableintensity, had its conductivity permanently reduced to a small Rztctiotiof its original amoiint. In directions perpendicular to the axis, rockcrystal, even at higher temperatures, has little or no conductivity.As the result of their further investigations, the authors havearrived at the conclusions that-(1.) The electrolytic conductivity of rock crystal in bhe direckion ofthe principal axis is, at high temperatures, about the same as that ofordinary glass.(2.) When a slice cut perpendicularly to the axis is electrolysed,sodium-amalgam being used as the anode, sodium migrates throughthe slice, its amount being in accordance with Faraday's law, and theweight of the slice remains unchanged.(3.) Even at high temperatures, rock crystal acts as a good insulatorwith respect to an E.M.F. in a direction perpedicular to the prin-cipal axis.When sodium-amalgam was used as the anode in am experimentlasting for three days, at a temperature of 250°, 88 milligrams of silverwere deposited in a silver voltameter in the circuit, and the onlysubstance which could be detected at the cathode was sodium.When potassium was used in the place'of sodium, it was found thatafter 40 hours the current had sunk to about the hundredth part ofits original value, only 2 milligrams of silver were separated, and nopotassium could be detected ah the cathode, even by means of thespectroscope.The authors therefore conclude that the conductivityis due to the presence of sodium, in the form of Na,SiO,, which wasshown, by an analysis specially made by Baurnann, to be present inthe crystal employed in the proportion of 1 part in 2300, so that thecrystal might he regarded as a ver,y dilute solution of this salt.The electrolytic character of the conductivity was further COU-firmed by the fact that a cell giving an E.M.F.of from 1.2 to 2 voltscould be formed of mercury, a slice of quartz at a temperature of225", cut perpendicularly to the axis, and sodium-amalgam.According to Clausius's theory of electrolysis, the fact of electrolyticconduction only taking place in the directioa of the principal axiewould tend to the inference that in the case of rock crystal nottraversed by an electric current, the interchange of atoms betweenthe molecules can only take place, at any rate to a sensible extent, inthe direction of the principal axis.A confirmation of this inference is found in the fact first noted bySalm-Horstmar (Ann.Phys. Chem., 120, 334), that the action ofhydrofluoric acid on rock crystal is much greater in the direction ofthe axis than perpendicular to this axis. The authors have themselvesh 92 ABSTRACTS OF CHEMICAL PAPERS.made experiments to test the truth of this statement, and the resultsare in agreement with those of Salm-Horstmar.It would appear from the results obtained in the paper, tllnt thesilicate Na2Si03 contained in the crystal must partake of its crystal-line structure.G. W. T.Effect of Occluded Gases on the Thermoelectric Propertiesof compounds. By J. MONCKMAN (Proc. Ro?y. SOC., 44, 220-236).-When a portion of platinum o r palladium wire is charged withhydrogen by electrolysis, and the wire afterwards heated, a cui-rentpasses from ithe protected to the unprotected part. The same occurswith rods of carbon after charging and pressing together, the currentpassing from the hydrogen to the oxygen. The wires and rods arefound to have an increased resistance, that of the oxygen rod beingthe greatest. The effect disappears after short circuiting. If thewires or rods be charged twice in opposite directions, the effect dis-appears, unless thc second charging is of very short durafion; infhis case,'% reversal takes place.With carbon rods a t different tem-peratures in contact, reversal occurs a t 250" ; with a thermoelectriccouple of carbon and pla,tinum, the thermoelectric line rises below250', and falls above that temperature. The rate of decrease of resist-ance of carbon diminishes as the temperature rises to 250", butincreases afterwards. The rate of evpansion increases as the tem-peratiirc rises to 250", but afterwards decreases. The specific hestincreases fairly regularly up to 250°, but above that temperature fallsto half.H. K. T.Electrochemical Effects on Magnetising Iron. By T. ANDREWS(Yroc. Roy. SOC., 44, 152-168).-A niagnetised and an nnmagne-tised bar of iron or steel are immersed in different reagents, and thecurrent produced noted.The amount varies considerably, but islarge in the case of bromine, salts of copper, and nitric acid. Theresult is dependent both on the strength of the solution and the degreeof magnetisation. With powerful oxidisers, the magnetised bar isgenerally electropositive, but becomes electronegative with sulphuricacid, dilute hydrocbloric acid, and ferric chloride and chlorine. In thelaut-named instances, the effect may be due to the diamagnetic propertiesof the solutions, or of the gases evolved. With ferric chloride alone,the magnetised bar is electropositive, with chlorine electronegative,with the two together, electronegative until the chlorine is exhausted,when it becomes electropositive. In the same bar, local currents areproduced from the more magnetised portions to the less.These maycause the magnetised bar to be acted on to a greater extent than theunniagnetised. In strong nitric acid, a current is produced from themagnetised to the unmagnetised bar.Specific Heat of some Solid Organic Compounds. By H.HESS (Ann. Yhys. Chem. [Z], 35, 410--429).-The author states that,with the exception of some investigations by De Heen (Bull acnd. roy.brlg., 5) and A. Batt'elli (Atti R. Id. Veneto [GI, 3), he has not beenable t o find any account of investigations of the specific heats of solidExperiments were also made with graphite rods.H. I(. TGENERAL -4hTI.l PHYSICAL CHEMISTKY. 93organic compounds, and he therefore undertook the present inresti-gation with a view especially of determining t>he manner in which thespecific heats of solid organic substances depend on temperature.The author gives a number of curves showing the relation betweenspecific heat and temperature in the substances exauiined, tempera-tiires being taken as ordinates, and the corresponding specific heats asabscissae.The curves he fiuds to be sensibly straight lines intersectingthe specific heat axis above the zero point, so that the specific heatName ofsubstance.{ Oxalic acid.. . .Malonic acid . .Succinic acid . .Isosuccinic acidGlutaric acid(solid)(liquid) { Gtlutaric acidPy rot artaricacidnic acid { Dimethylmalo-{ Sugar .. .. .. ..Benzoic acid(solid)(liquid) { Benzoic acidPhthalic acid . . {-~~~Mean specific heat.Tempera-ture limits.--0" to 50"0 )7 750 7 7 940 77 500 7 7 940 7) 1100 7 7 500 Y, 940 97 750 77 940 7, 500 7 ) 750 )) 940 ,) 99.30 7 , 500 >, 750 ,7 940 ), 1050 7 7 500 7) 940 ,7 750 7, 940 ,, 150J > 500 ,) 1300 ), 1130 ,) 1300 9 , 500 ,, 940 7 7 1300 ,) 1220 ), 1360 ., 750 7 . 1190 ,7 150C.0 -33590 -35750 -37280 *28320 -31310.32620.28980 -32520 -36500 -33780.35000 *3ti360 *30810 *32070 *34610 * 7503 J0 *30980 *32670 -35480-35750 -3996 I0 347410 -30370 * 31970.33370.35110 *25? 10 -31180 *33190 '50'72 I0-5256j0 -25590 -28620 *30990 9285 1Tempemtm coefficient.Cempeiaturelimits.50" to 75"50 77 9450 ,) 9494 ,, 11050 ), 11050 9494 7 7 15050 ,) 15075 ,, 9450 )) 7575 ,7 9475 y 7 9450 7 7 9450 ,, 7550 7 7 9499 7 7 13050 7 7 7575 ,) 10550 7 7 9450 7 7 10650 77 9475 7 ) 9475 ,) 11394 ,) 13075 ), 13050 7 7 9450 ,) 110122 ,) 13694 ,) 11075 ) ) 119119 ,) 15075 ,, 1506.0 *0008640 -008010 *0008390 *0006800 *0007710.0007050 -0008050 ~0007110.0007ti20 *0005120 -0007160 *0006000 -0005040 '0913.3'70-000864O~OoO7lO0 so06760 *0010270 - 0007960 *0008670 -0008590 - 0008420 *OW7890 *OW8720.0008620 -001240.001260 -001650 -001310 -0006890.0007640 *000720Means.b = 0 '0008356 = 0 *000719b = 0 *000759ii1 6 = 0 400609b = 0 .000901b = 0 *000;10b = 0 '000842 1b = 0*000859I 6 = 0 *000841 Ib = 0 '00125b = 0 '001316 = 0 '000724 94 ABSTRACTS OF CHEMICAL PAPERS.can be represented by a formula of the form n + bt.The resultsobtained are given in tabular form (p. 93), c representing the meanspezific heat between tihe temperature limits indicated, and b thetemperature coefficient.The values obtained by assuming the true specific heat to be repre-sented by a forniula of the form cc + bt are given in the second table,under the head of "observed specific heat," the column headed" calculated specific heat " being calculated from Kopp's law, that themolecular heat of a body i s equal to the sum of the atomic heats ofits constituents. The atomic heats of carbon, hydrogen, and oxygenrespectively are taken as 1.8, 2.3, and 4.0.The column headed tgives the temperature a t which the observed and calculated specificheats are equal, and it will be seen that with the exception of oxalicand isosuccinic acids, the different substances obey Kopp's law forsome temperature within the limits 3-5" and 50".Kopp's law might be generalised if we could assume the specificheats of carbon, hydrogen, and oxygen to be functions of the tem-perature, but this would not lead to correct general formulae, forRegnault (Compt. rend., 26, 311) and E, Wiedemann ( A m . P h y s .Chem. 15 7, 1) have shown that the specific heats of hydrogen andoxygen are sensibly iudependeut of the temperature, and althoughH. P. Weber has shown (Ann. Phys. Chem., 147,362) that the specific:heat of carbon increases considerably with the temperature, thisincrease would not be sufficient to account for the observed increase inthe temperature coefflcient.Name of substance.Oxalic acid...................Malonic acid.. ................Succinic acid .................Isosuccinic acid ...............Gtlutaric acid (solid) ...........I? yrot art aric acid. .............Dimethylmalonic acid .........Sugar.. ......................Rerizoic acid.. ...............Plithalic acid.. ...............Benzoic acid.. ................ Glutaric acid (liquid) ..........Specific heat.Calculated.0 *26890.29420 -31360 - 3 h 80 *3!!400 -28200 -2602¶)--0 bserved.0-2941 + 0-00167t0'2473 + 0.0014450.2518 + 0-00152f0.3067 + 0-00122t0 -2620 + 0.00180t0.2677 + 0'00168t0.2666 + 0'0017250.2387 + O'OCI173t0.1946 + 0.0025050.2016 + 0-00145~0 -6580 + 0 -00142t0.3474 + 0.002625t.- 15 -1' + 32 -6 + 40.7 + 5-7f 37 -1 + 36 *4 + 36 *2 + 49 -3 + 35-0+40*4 - - -The author's results show that there are often considerable differ-ences in the specific heats of different isomeric compounds.Evolution of Gases from Homogeneous Liquids.Ry V. H.VELEY (Proc. B o y . Xoc., 44, 239--240).-The addition of fineiydivided substances is found to increase the rate of evolution of gasesfrom liquids in which they are formed. When the temperatureG. W. TGENERAL AND PHYSICAL CHEMISTRY. 95remains the same, the rake of evolution rises slowly until a maximumis reached, which is maintained for some time.The rate thendecreases proportionally to the diminution in mass. The phenomenonrepeats itself when the temperature is lowered and then raised t o itsformer point, and also when the pressure is suddenly increased. Thereduction of the pressure to a fraction of an atmosphere produces nopermanent effect. The rate of decomposition of formic acid intocarbonic anhydride and water is also examined, and is found t o agreewith the equation log (T + t ) + log r = log c, where T is the timefrom the Commencement of the observations, t the interval of timefrom the moment of commencement up to the moment at which thetime required for unit change is mil, r the mass at the end of eachobservation, and c a constant.The curve of rate of change conformswith the law drld7 = - r2/c, which expresses the rate at whichequivalent masses react on one another. Hence it is presumablethat equivalent masses react, and that the change is represented bythe equations HCOsOH + HCO*OH = HCO-OCHO + H,O andHCO.O*CHO = 2CO + H20, a reaction similar to the production ofethyl formate from formic acid and alcohol.Properties of Matter in the Gaseous and Liquid State underVarious Conditions of Temperatura and Pressure. By the lateT. ANDREWS (Ann. Chirn. Pl~ys. [GI, 13, 411-432).-Regnault (Xem.Acad. Sci., 26, 680-696) made a series of experiments to determinethe tension of a mixture of a gas and a vapour, such as nitrogen orair, and the vapour of water or some more volatile liquid, and cameto the conclusion that Dalton’s law of partial pressures may be con-sidered theoretically correct in the case of such mixtures, and thatprobably this law could be proved to be correct experimentally if themixture of gas and vapour could be enclosed i n a vessel the interiorsurface of which was composed of the volatile liquid.He also foundthat, under pressures varying from + to 2 atmospheres, the com-pressibility of a mixture of ordinary gases, such as air and carbonioanhydride, hydrogen and sulphurous anhydride, was intermediatebetween that of each gas separately for the same variations of pres-sure (ibid., 258).The results of all experiments which had been carried out up tothe time when the author’s investigations were commenced, had beento show that, with one exception, Dalton’s law i s true in all cases formixtures of gases or vapours, or at any rate in the case of gases andvapours which exert no chemical action on one another.A mixtureof the vapours of two mutually soluble liquids, in presence of thetwo lcquids mixed or dissolved, constitutes, however, an importantexception to this law, because of the disturbing influence of thechemical affinity of the liquids. But as, up to this time, no experi-ments had been carried out, to prove the truth of Dalton’s law underpressures greater than 2 atmospheres, the author investigated thechange in volume of a mixture of 3 vols. of pure carbonic anhydrideand 4-05 1-01s. of nitrogen at temperatures above and below thecritical temperature of carbonic anhydride, the pressure employecivcrying between about 40 and 300 atmospheres.H. K.T9 8 ABSTRACTS OF OH'EJllCAL PAPERS.From the results, which are given in tabular form, curves aredrawn showing the volume of tlie mixture at the various temperaturesand pressures. These curves are all very similar, showing no differ-ence in character for temperatures above or below 31". If it begranted t h a t Dalton's and Boyle's laws are true in the case ofnitrogen under the pressures employed, the curves showing thechange in volume of the carbonic anhydride in the mixture under tlievarious conditions of temperature and pressure prove that below 31"t h i s gas tends to occupy the volume corresponding with the liquidstate, although the curves are quite different from those of carbonicanhydride alone.It follows, therefore, that Dalton's l a w is no longerapplicable in this case, and is only strictly true of a perfect gas.As no liquefaction took place in any of the above experiments,showing that the presence of nitrogen lowered the critical point OF thecarbonic anhydride, the author investigated this phenomenon morefully. A mixture of 6.2 vols. of carbonic anhydride and 1 vol. ofnitrogen was placed under a pressure of 48.3 atmospheres ; no con-densation occurred until the temperature was lowered to 3.5". Asthe pressure was increased the volume of the liquid augmented, andafter each increase of pressure, the volume continued to augmentslowly for some time; for example, under a pressure of 82 atmo-spheres the relative volumes of the gas and liquid were at first 8.5and 5.8, but, the apparatus having been left for some time, thcbvolume of the liquid slowly increased.The pressure having beenthen raised to 102 atmospheres, the volume of the gas which was a tfirst 1.7 diminished gradually until only a small globule remained,which finally disappeared entirely, the nitrogen dissolving in theliquid carbonic anhydride. In a second experiment, with the samemixture at a higher and constant temperature, the liquid had a t firstits usual concave surface, and as the pressure was increased, thevolume of the liquid also augmented without any noticeable changein the appearance of the concave surface ; on further increasing thepressure, the surfaca of separation appeared in section as a fine line,but when the pressure was again increased, i t disappeared entirely,the whole becoming homogeneous.The position in the tube, occupiedby the surface of separation, depended on the temperature a t whichthe observation was made ; a t 14" the liquid filled about, two-thirds ofthe entire space a t the very moment when the surface of separationwas about to disappear,The critical temperature of a mixture of 1 vol. of nitrogen and3.43 vols. of carbonic anhydride was found to be 14", and the cor-responding pressure 98 atmospheres. Experiments with this mixtureshowed that at 6.3" no condensation took place until the pl:essurereached 68.7 atmospheres ; the liquid then disappeared under in-creased pressure, but reappeared when the pressure reached 113.2atmospheres.At 9.Y0, the liquid first appeared when the pressurereached 77.6 atmospheres ; after having disappeared i t was againformed under a pressure of 107.8 atmospheres. At 13*2", the liquidappeared under a pressure of 91.6 atmospheres, disappeared as thepressure was increased, and reappeared when it attained 103.2 atmo-spheres, If the mean of the two pressures for each of the abovGESERAL AXD PHYSICAL CHEMISTRY. 97temperatures is taken, the critical pressure a t 6*3", 9*9", 13 2", and14" is found to be 90.9, 92.7, 944, and 98 atmospheres respectively.I n the course of these experiments, the author found it convwientto employ a tube bent twice at right angles.When the gaseousmixture was compressed below the critical point, the liquid carbonic:anhydride collected in the lower portion of the tube, although part ofthe liquid was first formed a t the surface of the mercury ; but thewhole of the liquid soon collected at the bottom of the tube. Insome experiments, the carbonic anhydride liquefied a t temperaturesabove 20°, and sometimes no condensation took place even a fewdegrees below this temperature. This phenomenon was found to beowing to the fact that when liquefaction had taken place, if thepressure was diminished so that the mixture could become completelygaseous, the liquid separated into two portions, one rich, the otherpoor, in carbonic anhydride. The portions of the tube which hadbeen previously occupied by the liquid then contained a large excessof carbonic anhydride, especially when the tube had been previouslycooled to -lo", so that almost the whole of the carbonic anhydridehad been liquefied. If the pressure was reduced so as to bring thewhole of the liquid to the gaseous state, the temperature being a t thesame time raised to 26", it was found that the carbonic anhydridecould be liquefied by pressure alone (at 26"), provided that, theexperiment was performed without loss of time.When, however, themixture was left for some time in the gaseous state, diffusion graduallytook place, and the temperature at which liquefaction wa9 possibledecreased accordingly. Diffusion was not complete until after somehours, and then increased pressure caused no liquefaction until thetemperature was reduced to 14".This method of separating the gases was employed to shorn theeffect of diffusion as follows :-A mixture of carbonic anhydride andnitrogen was kept at 8.5" under a pressure of 46.4 atmospheres untildiffnsion was complete ; the volume of the mixture was then 162.2,After liquefying the carbonic anhydride by employing great pressureand lowering the temperature to -12", the temperature was againraised to 8*5", and the pressure brouqht back to 46.4 atmospheres;the volume was then found to be 159.5, showing that a contraction of2.7 ~01s.had taken place owing to the separation of the mixed gases.A t the end of 1+ hours the volume had increased to 161.5 in conse-quence of partial diffusion.In a second expeiiment at 16", under a pressure of 47.9 atmospheres,the original volume of the mixture was 164.6, but, after liquefaction,only 161.9 when brought back to the initial temperature and pres-sure ; after 1; hours the volume had increased to 164.1.I n a thirdexperiment a t 20", under a pressure of 46.4 atmospheres, the volumedecreased from 175.8 to 173.5 after the separation of the gases.These results show that when the two gases d i f i s e into oneanother under great pressure, an increase in volume ocmrs, and whenthey are separated the volume is diminished. This change in volumeundoubtedly occurs also under ordinary pressures, but the variationwould probably be so small that it would be dif€icult to detect experi-mentally. F.S. K98 ABSTRACTS OF CHEMICAL PAPERS.The Behaviour in Relation to Boyle’s Law of certain Gasesat Low Pressures. By F. FUCHS (AN%. Phys. Chem. [‘L], 35, 430-450).-The author, from the results of a series of experiments onatmospheric air, carbonic and sulphurous anhydrides and hydrogen,arrives a t the following conclusions :-(1.) At ordinary temperatures, Boyle’s law does not represent alimiting state towards which a gas approaches indefinitely withincreasing rarefaction, but a t pressures respectively above and belowa certain amount, the deviations from Boyle’s law are respectivelypositive and negative. The limits of pressure within which Boyle’slaw holds are indefinitely small, as any finite change in volume willalter the forces betweeii the gaseous molecules.(2.) I n the case of atmospheric air a t the temperature 0”’ a changeof sign of this kind takes place at a pressure very slightly below theordinary atmospheric pressure.If any similar change of sign occurswitb carbonic and sulphurous anhydrides, it must be a t pressures lessthan any at which the author’s observations were made.(3.) The deviations from Boyle’s law in the case of hydrogen atlow pressures are so small that hydrogen under these circumstancesmay, without, sensible error, be regarded as a perfect gas.G. W. T.Constitution of Solutions. By F. R~DORFF (Ber., 21, 3044-3050).-Snlts of the composition R2S04 t R“S04 + 6H20 andR,SOd + R,“’(S04)3 + 24H20 are partially decomposed into theirconstituents when dissolved in water (compare Abstr., 1888, 342).Hydrogen potassium sulphate behaves similarly, but hydrogen ethylsulphate diffuses unchanged.3KZCzO4,Fe2(C2O4), + 6H,O ; 3Na2CzOd,Fe,( C204), + 6Hz0 ;3K,C,Oa,Cr,(C2O4), + 6H20 ; 2(NH4)HC2O4 + H,O ;PU’nHC4H406 + H20, and 2K(SbO)C4H40, + H20,diffuse unchanged, but (NHa)HC,04,C,H204 + 2H,O is partially de-composed into oxalic acid and hydrogen ammonium oxalate.Solutions of potassium chromate, potassium dichromate, andsodium dichromate diffuse unchanged, but the salt(NH4)2Cr04,MgCrOl + 6H20The following salts :-is partially decomposed when dissolved in water.‘l’he following salts :-2NaCl,PtCI, + 8H20 ; 2KCl,PtCIz ;2NH4C1,HgC12 ; Ba(CN),,Pt(CN), + 4H20, and all double cyanidesare true molecular compounds, but KCl,Hg( CN), is partiallyresolved into its constituents when dissolved in water (Zoc.cit.).NaH2POa and Na2HP04 diffuse unchanged: Na,P04, on the con-trary, is partially decomposed. The three sodium salts of citric acidare not decomposed in aqueous solution. F. S. K.Physical Properties of Colloi’d Solutions. By C. L~DEKING(Ann. Phys. Chenz. [2], 35, 552--557).--In a paper with Wiedemann(Abstr., 1885, 1032) it, was shown that the vapour-pressure of GENERAL AND PHYSICAL CHEJIISTRY. 9940 per cent. aqueous solution of gelatin was less a t a temperature of40" than that of pure water. According to Guthrie (this Journal,1877, i, 36), a 40 per cent. solut'ion of gnm boiled at 98", and a 45 percent.solution of gelatin at 97.5": results which were in contra-diction to those above mentioned.With a view of discovering the reason of the discrepancy, theauthor made experiments on solutions of gum arabic, gum trngacan th,dextrin, starch, and agar-agar.He finds that a 40 per cent. solution of gum arabic boils at IOO",but carbonic anhydride begins to be given off a t a temperature ofabout go", and a t a somewhat higher temperature gives the appearanceof boiling to the solution.The other solutions also boiled at loo", although in the case ofgelatin boiling began with the thermometer a t 98", which, however,the author attributes to the liquid not rapidly assuming the sametemperature throughout, owing to its viscidity preventing the forma-tion of convection currents, This opinion was based on the fact thatthe thermometer did not remain at 98", butpadually rose to 99.8",where it remained constant.The author found that the addition of the colloid in every caseslightly lowered the vapour-pressure, and, as he points out, thepresence of a solid i n solution could not possibly increase thevapour-pressure. For example, if the steam given off at 98" from agelatin solution had a pressure of 760 mm., i t would necessarily re-condense to water and mix again with the solution.When solutions of gum or gelatin are cooled considerably belowzero, the author finds that they do not solidify as a whole, as statedby Guthrie, but ice crystals gradually separate out.He finds thatgelatin has a strong condensing action on the water of solution.G.W. T.Precipitation of Colloid Substances by Salts. By 0. NASSE(Pjiiyer's Archiv, 41, 504--514).-A11 prote'ids except peptone canbe precipitat!ed by saturating a neutral solution with ammonium sul-phate, some more easily than others, for instance, globulins moreeasily than albumins. Other salts bave the same power, but noneact so readily as ammonium sulphate. This property, however, is notcharacteristic of proteids : soaps, gelatin, and certain fioluble carbo-hydrates (glycogen, amidulin, inulin, &c.>, are similarly precipitated ;it in fact seems to be a property common to colloid substances.The question arises, on what does the difference in the concentra-tion of the salt necessary to produce precipitation depend ? Is theaction of the salt simply due to a struggle of the molecule ofproteid, gelatin, &c., with that of the salt for water, and t h a t the pre-cipitation of the colloid substance occurs as soon as its water-attract-ing power is exceeded by that of the salt?In order to determine whether this is the case, one m u d ascertainthe amount of two or more different salts necessary to precipitat'e thesame aniount of one colloid substance, the necessary concentrationof the salt solutions corresponding with a : b : c, &c.The same ques-tion is then invebtigated €or another colloid substance, and the rati100 ABSTRACTS OF CHEMICAL PAPERS.a' : b' : c', &c., found.for water, a : b : c, &c., will be found = a' : b' : c', &c.Tf hhen we have only to deal with attractionThe following table illustrates the results obtained :-Collo'id substances.Gplatin .........The solution contained in 100C.C.the following amounts ofsalt w18cn precipitation began.L--v---Ja. 6 .Ammonium Magnesiumsulphate. sulphat e. a : b.12.4 14.8 0.84White of egg.. .... 20.2 19.6 1-039 3 ...... 18.5 19.3 0.95Serum prote'rds .... 17.4 18.5 0.94Albumose. ........ 12.7 13.6 0.9:3,, ......... 14.9 17.6 0.85Amidulin. ........ 20.9 10.5 1-99Glycogen-dextrin . . 44.7 23.7 1-99The differences of the numbers in the last column show thatwater-attracting power is not the only influence at work, but someother relation must exist between the colloid and the salt.Still it ispossible that it may explain some of the precipitations, especially thatof gelatin. Gelatin loses many of its characteristic properties afterthe prolonged heating of its solutions; it, for instance, no longergelatinises on cooling, and its water-holding power is greatly in-creased, yet, the ratio a : b = 0.84 remains constant for the gelatin inall the different stages of this change.With regard to the proteids, in which such wide differences occur,it is thought probable that the explanation lies in the fact that loosecompounds with the salts are formed.The paper concludes with some remarks on the influence of tern-perature in determining precipitation by salts.New Formula for Calculating the Molecular Volumes ofChemical Compounds at the Boiling Points. By J. A.GROSHANS (Rac. l'rav. Chim., 7, 220--225).-The molecular volumeof a sut)stance, CPH4Or, at the boiling point may be represented byt7, = a + 1O(p + q ) - 7.28 B for a fatty compound, or by the sameexpression miiius 15 for an aromatic compound, c1 being the numberof' O.C. equal to the molecular weight of the compound, and B = p + q + Y. Both these formulte obey Kopp's rules, that homologouscompounds differing in their composition by CH, should differ i utheir molecular volumes by 22, and that a fatty compound shouldhave approximately the same volume as an aromatic compound whichdiffers in its formula by C2 - Hq.With hydrocarbons, since 7 = 0, the formula may also be written(us - a)/B = 2-72 for fatty compounds, or (u, - a -+- 15)/B = 2-72for aromatic, both of which are fonnd t o agree well with experiment.For halogen-deiaivatives, an addition of 15 must be made for each atomof halogen oontained. H. C.W. D. HISORGANIC CHEMISTRY. 101Molecular Lowering of the Freezing Point of Benzene byPhenols. By E. PAT ERN^ (Ber., 21, 3178-3180).--The author madea, number of experiments to ascertain whether the fact that certainsubstances containing the hydroxyl-group produce an abnormal lower-ing of the freezing point of benzene WRS true of all substances, andwhether this abnormal behaviour was sufficient proof of the pi*esenceof the hydroxyl-group (compare Raoult, Abstr., 1884, 959). The re-sults showed that although phenol behaves in an abnormal manner,the following compounds : ethyl phenol, acetylphenol, two isomericnitrophenols, tribromophenol, picric acid, paracresol, methyl salicylate,thymol, nitrothymol, nitrosothymol, a-naphthol, /-3-naphthol, andbenzylphenol, all produce the normal lowering of the freezing point ofbenzene and of acetic acid, either in dilute or moderately concentratedsolutions, the variations caused Ey change in concentration of coursebeing taken into consideration.The molecular weight of water determined by Raoult's method inacetic acid solution was found to be 18 (compare Rxonlt, lor. cit.), butthe author points out that tlhis result is not by any means conclusive,as, even if the molecules were originally more complex, they wouldbe simpli6ed by the act of solution.Hentschel's experiments with acetic acid (Zeit. phys. Chenz., 2, 308)seem to point to an opposite conclusion, but in this case the freezingpoint of the benzene solution was considerably below that of aceticacid. The fact that the molecular weight of water is found to be 36when the freezing point of the solution lies below 0" shows that thetemperature a t which the mixture freezes is a most important factorin the case. F. S. K