109General and Physical Chemistry,Refractive Indices of Crystallised Alums. By C. SORET(Compt. rend., 99, 867,-869).-By means of the instrument pre-viously described (Compt. rend., 95), the author has determined therefractive indices OE many alums for the lines a, B, C, D, E, b, I?, G,of the solar spectrum. He has also determined the specific gravitiesof the alums by means of the hydrostatic balance. In the followingtable, only the refractive indices for 13 and the specific gravities aregiven :-?D. Sp. gr.Ammonium aluminium alum.. . . 1.45939 1.631Sodium 9 ,Methylamine ,,Pohssium 9 1Rubidium 9,CEsium 9 ,Thallium 9 7Ammonium indium,, galliumPotassium ,,Ammonium chromiumPotassium 9 9Rubidium Y 9Thallium 3,Ammonium ironPotassium ,,Rubidium ,,CEsium ,,Thallium ,,,, ....7 ' * * ..9 9 *...,, ....,, ....,, ....,, ....,, ....,, ....,, ....,, ....,, ....,) ....,, ....,, ....,, ....,, ....,, ....3.438841.454101.456451.456601.458561.497481.466361.465581 *464991-484181.48 13 71.481511.484821-481691.482391.483781.523651.522801.6671.S681.7351.85219112.2572.0111.7451-7191.81 71.9462.2361.7131.8061.9162.0612.385-The molecular volume is not constant for the different terms of thesame series, but it seems to vary in a definite manner for the corre-sponding terms of the aluminium, chromium, and iron series.In passing from one alum to another, the variation in the refractiveindex is sensibly the same in the three series, thus following a lamwhich has been observed in other series of compounds.It is worthyof note, however, that the refractive index of sodium alum is muchless tbau that of potassium alum, whilst in the case of the chloridesthe sodium salt is intermediate between the potassium and ammoniumcompounds.The author's value for the refractive index of thallium alum ismuch higher than that obtained by Fock.Methylamine alum is intermediate between the sodium and potas-sium compounds, and it would seem as if, in the aluminium series, therefractive index varied continuously with the molecular weight of thealkaline radicle. C. H. B.VOL. xr,vm. 110 ABSTRACTS OF CHEMICAL PAPERS.Inversion of the Electromotive Force of a Copper IronJunction at a High Temperature.By F. F. LE ROUX (Compt.rend., 99, 842-84A).-A bar of iron was bent in the form of a horsn-shoe, and attached a t each end to a copper bar. This couple wasplaced in a furnace and heated nearly to the melting point of copper,a current of about 330 amphres being passed through the couple. Anychange in the temperature of the junction was detected by observinga change in the relative luminosity of the junction, the results ofocular observation being confirmed by the action of the radiation on agehtino-bromide plate.It was found that a t about 1000” a current passing from the copperto the iron raises the temperature of the junction, whilst at the ordi-nary temperature a current in the same direction cools the junction.C.H. B.Electrolysis of Silver Fluoride, Chlorate, and Perchlorate.By G. GORE (Chem. News, 50, 150).-A moderately strong solutionof silver fluoride acidified with hydrochloric acid is a very good con-ductor of electricity, and is very readily decomposed by means of silvereltctrodes and a current from a cell containing zinc and platinum indilute sulphuric acid. Crystals of silver are rapidly deposited a t thecathode, Fhilst the anode soon becomes rough, grey i n coiour, andvery friable. In special experiments, no evidence could be obtainedto show that this loss of cohesion was due to the diffusion of liberatedfluorine through the silver.When a solution of silver chlorate is electroljsed by sheet silverelectrodes and a current from two Smee’s cells charged with verydilute sulphuric acid (1 vol.of acid to 50of water), conduction is good,and silver is freely deposited only at first; the deposit being loose,and not very white. The anode also is soon coated with a black film,presumably silver peroxide, which seems to stop the current ; it ispermanently blackened, although but slowly corroded. With oneSmee’s cell, the deposit is formed slowly, and is more coherent,, Thissolution requires a feeble current, a large cathode, and a much largeranode.When silver perchlorate is similidy electrolysed, conduction is verygood, and loose, bulky, silky crystals of silver are soon deposited a t thecathode, whilst the anode quickly becornes black, the current a t thesame time becoming much diminished. With one cell and a moredilute solution, conduction is free, the deposit is smaller, and the anodebecornes less dark.A silver wire anode soon becomes corroded andloosely coated with a black substance ; this falls off and is ultimatelyreplaced by a thick green coating ; no gas is evolved. The solutionrequires a large cathode and a rather small anode. D. A. L.Electro-deposition of Carbon and Silicon. By G. GORE (Ckenz.Xews, 50, 113--114).--Carbon, silicon, and boron have each beendeposited during the eleccrolysis of certain fused mixtures.C‘irrbon is dfyosl’ted when a current from 10 Bmee’s elements ispaswd through a fused mixture of 200 grains of sodium hydroxide,170 grains precipitated silica, and 610 grains of mixed anhydroussodium and potassium carbonates, the anode being sheet platinumGENERAL ASD PHYSICAL GHEMISTRY. 111tlhe cathode a wire of the same metal; the deposition of carbon is,however, probably due to a secondary reaction of this character:silicon i s first deposited and &is reacts with the alkaline carbonates,and causes the separation of carbon.The same phenomenon occurswhen a mixture of 475.2 grains of 97.1 per cent. sodium carbonate and217.4 grains sodium borofluoride is similarly treated, and apparentlyalso in the electrolysis of aqueous or alcoholic oxalic acid. The de-posited carbon is not crystalline.Carbon is not deposited either during the electrolysis of sodium andpotassium carbonates, using eight Smee's cells and platinum electrodesa t a red heat, or when boric acid is included in the mixture, or bythe electrolysis of any of the following : potassium cyanide, oxalic acidin solution in hydrochloric or nitric: acid, sodium formate or formicacid, carbonic oxide and anhydride, pyrogallol, liquid hydrocynnicacid saturated with carbonic oxide, fuming sulphuric or syiwpyphosphoric acids saturated with dry carbonic anhydride ; or fromdilute sulphnric acid over which coal-gas was passing during the14 days the electrolysis was continued.Carbon tetrachloride does notyield carbon undcr t h e influence of an electric current. These experi-ments were conducted under different conditions a8 t o strength ofcurrent, length of time, temperature, and composition of electrodes.Xilicon is deposited when a fused mixture of 300 grains of 97.1 percent.potassium carbonate and 442 grains of potassium silicoflnorideis electrolysed, as described above i n the carbon-deposition experi-meats. D. A. L.Relation between the Ordinary Thermometer and the WeightThermometer. By E. BARBIER (Compt. rend., 99, 752-753).-Ademonstration of the theorem that if the ordinary thermometer andthe weight thermometer agree at the two fixed points, they agree atall temperatures. C. H. B.Attraction of Homogeneous Molecules. By C. SCHALL (BPI-.,17, 2555--?577).-In order to interpret experiments on the rela-tion between the rates of evaporation of liquids and their molecularweights and heats of expansion, the author has more particularlystudied the phenomena of cohesion and adhesion of liquids, a subjectof interest to the chemist as dealing witoh the attraction of homo-geneous and heterogeneous molecules.The method of investigationwas based on that of the so-called adhesion plates, which consists, inoutline, in suspending a plate of glass from one pan of a balance, andcounterpoising i t ; the plate being adjusted to a level, a dish contain-ing the liquid to be examined is placed under it, and then raised untilthe surfaces of the liquid and glass are in contact. To the oppositepan of the balance, weights are added until the glass is severed fromtlie liquid ; this excess of weight is then toted. The apparatus used,toget'her with devices for levelling the plate and for the complete sever-ance of the liquid and glass, are described in detail in the paper, Asthe attractive force between two contingent molecules within a liquidis proportional to their mass and inversely proportional to the squareoE the distance between them, and as increase of distance is correla-tive with decrease of specific gravity and also with that of cohesion, i 112 ABSTRACTS OF CHEMICAL PAPERS.it follows that a decrease of the latter caused by warming the liquidis proportional to tihe square of the former. But the superficial expan-sion, which is equal to the $j power of the cubical, is inversely propor-tional to the specific gravity.As the superficial expansion increases,the number as also the mass of molecules under the plate and theircorrelative cohesion diminishes, and therefore the latter diminishes indirect proportion to the Hence if sand s1 be the specific gravities for any two degrees of temperature, Gand G' the excess of weight necessary for the disruption of the plate,thenpower of the specific gravity.Experimental results are tabulated which demonstrate the validityof the formula, and of the law deduced therefrom, that the force bywhich homogeneous molecules are attracted varies in direct proportionto the square of the specific gravity, and also to the mass of the mole-cules.From the experimental results can be deduced the diminutionof cohesion for each degree of temperature, and thus the critical pointat which the cohesion is nil.Rut the results obtained with some of the liquids examined, espe-cially water, benzene, and its derivatives, are not in strict accordancewith the law enunciated above, so that it would appear that theforce of cohesion is dependent to some extent on the chemical consti-tuttion of the liquid.I n the case of t w o liquids, it is further de-monstrated that the relation between the respective cohesions and alsotheir specific gravities at boiling points within restricted limits ofpressure are approximately identical.These experiments are also of importance in regard to the pheno-menon of capillary attraction, a force which depends on the differ-ence between the force of cohesion of the molecules of the liquidwith one another, and of adhesion to the molecules of the glass.If the force of attraction as represented by the capillary height = h,that of the adhesion of the liquid to the glass = a, and of the cohe-sion of the liquid = c, then-1~ = a - c.The form of the meniscus is concave if a > c.but convex if a < c.B u t from the above formula, = (i)2(:)', then if the capillaryheights are h and h' at two different temperatures, then h = a-c,a-c I1 and h' = a'-c', it follows that h' = ~- , or 72' =Experimental results are also adduced in support of these formule,although water and liquid sulphur offer instances of marked exception ;it is thus probable that the molecular constitution of these liquids isthe cause of the discrepancy.Relation between Molecular Weight and Velocity of Evapo-ration of Liquids.By C. SCHALL (Bey., 17, 2199--2212).-ThisV. H. VGENERAL AND PHYSICAL CHEMISTRY. 113paper contains a description of the apparatus nnd method of workingemploged by the author in his experiments with benzene, carbonbisulphide, and water (Abstr., 1884, 551).Experiments with substances of nearly coincident molecular weightand boiling point :-Calculated.Phenol ................ wb = 94Aniline ................ rn = 93Toluene ................ m = 92Valeraldehyde .......... m = 86AIonochl(lrobenzene ...... m = 112.5Acetic anhydride m = 102Benzoic chloride ........ nz = 140.5Ethyl benzoate 112 = 150{{ ..................Found.95.57, 95.11, 95.1191.48, 91.91, 91.9193.8584.3108.2106144-:3, 144.5145.7, 145.2Experiments with substances of nearly equal molecular weights,but of different boiling points :-Calculated. Found.Ethyl acetate............. m = 88 86.32, 87.76Amyl alcohol(fermentation) m = 88 89.71, 87.76Benzddehyde .......... m = 106 104.7, 110Acetic anhydride ........ m = 102 103.3, 98.27{{{{{{{{ Amyl alcohol (fermentation) nz = 88Experiments with substances of unequal molecular weights, but ofnearly coincident boiling points :-Calculated. Found.Acetic chlori~le .......... m = Z3.5 78.06Acetone ................ WL = 58 58.33Alcohol ................ m = 46Benzene nz = 78 87.12, 87-12Toliiene ................ m = 92 94.28Phosphorus oxychloride . . m = 153.5 149-841.2, 41.2 ................Substances with different boiling points and different molecularweights :-Calculated.Found.Benzene ................ m = 78 75..5, 75.5Toluene ................ m = 92 95.04, 95.04Methyl alcohol’ .......... rn = 32 30.66Propyl alcohol nz = 60 62.63Ethyl alcohol.. .......... m = 46 45.48Isobutyl dcohol. N L = 74 74 84Isobutyl alcohol ......... m = 74 65.6599,19...................Recent determinations of heats of vaporisation show thst these areproportional to the time of vaporisation. I n the followiug table, t isthe boiling point a t which the heat of vaporisation L was determined,DL the product of the latter niultiplied by the theoreiical density, 112the calculated, and nz’ the found molecular weight :11 4 ABSTRACTS OF CHEMICAL PAPERS.---Water.. . .. . . . .. . .Wood spirit . . . . . , .Ethjl alcohol.. . . . .Amy1 alcohol . . . . , .litliyl acetate.. . . . .Metlijl butyrate.. . .Oil of lemon . . . . . .0 1 1 of turpentine.. .Hutjric acid . . . . . .Ethjl valerate . . . . .t.100"66 -5"7.81317493165156164113 *5L.---538 -0 cal.261 -7 ,)206 *4 ),120 -0 ,,105 .o ,,86.0 ,,69.5 ),68-5 ) )114 0 ,,68.4 ,,DL.--331 3'290.1328 ' 8368 *7320 *O31 13 .8327 * 4322.7347 -5308 .Om.---3246888810213613688130m'.---36.646 -489 .o91 - 2111.2138 *3139.884 -0140 -0The author has further compared the velocity of evaporation ofacetic acid with that of toluene, amyl alcohol, and isobutyl alcohol,and the results obtained show that the molecular weight of aceticacid a t its boiling point is 89.8 This may also be calculated fromF'avre and Sil bermann's determinations of its heat of vaporisation,arid likewise for formic acid, the molecular weight 69.A.K. M.On Crystallisation. Observations and Conclusions. By G.Stability of Compounds. By W. ALEXBEFF (Jour. XUSS. Chem.SOC., 16, 641-642) .-The author communicates his researches as tothe conditions determining the stability of a compound ill the presenceof an excess of one or the other of its constituents. The results agreewith what he found with regard to the stability of hydrates of alco-hols in their aqueous and alcoholic solutions. The difl'erence in thestability of hydrates determines the difference of the vapour-tensions,at one and the same temperature, for two solutions, which areformed by water, and a liquid capable of yielding a hydrate.More-over, in an aqueous solution, this tension is always smaller when thejormula of the hydrate is A + nH,O, n being generally greater thanunity. With solutions formed by water and ether, a difference intemperature of 8" corresponds with equal tensions.Phenomena af Condensation. By D. MENDEL~EFF (Jour. Russ.Clwm. rSoc., 16, 643--644).--The author remarks that the phenomenaof condensation, as shown in the case of the formation of solutions or ondiluting some liquids, is analogous to what takes place when sphericalbodies of different diameters, sach as samples of differentl seeds (peaseand millet), are niixeci together.When spherical bodies are mixed, ASmay be shown b-y experiment or by geometrical analysib, the weightof a measure containing a large numberof such small spheres of bothkinds is greater than the mean calculated from the weight of bothkinds taken alone. I n the same manner, the sp. gr. of a solution isgreater than it should be, when calculated from the sp. gr. and theynantity of the constituent liquids. The analogy in the change ofvolumes which takes place in both the above cases shows that when aI~R~GELMAKN (Ber., 17, 235942372).B. BQEKERAL AND PHYSICAL CHE3fISTRT. 115small bulk of light spheres of small diameter is added to heavierspheres of large diameter, the sp. gr. of a cubic: measure of the lastmay become greater, exactly as the firsf addition of water to normalsulphuric acid raises its sp.gr. The above geometric question is,unfortunately, up to the present inaccessible for full geometricanalysis, and the experimental investigation is rendered very di ficultby the impossibility of obtaining the neeessary balls of regular sizeand equal diameters.Experiments with mixtures of millet, and gunpowder, however, haveconvinced the author t h a t the above phenomenon exists, but it is onlya statical representation of a dynamical phenomenon which takes placein the case of dissolution as a simple act of chemical association ofheterogeneous particles. B. B.Connection between Pseudo-solution and True Solution.By ITT. W. J. NICOL (Chem. News, 50, 124).-Arguing from themolecular theory of solution, according to which the dissolution of asalt in water is the result of the attraction of the water molecules fora single salt molecule exceeding that of the attraction of the saltniolecules for one another, the author demonstrates that the differencebetween .pseudo-solution and true solution lies only in the degree ofsubdivision of tlie solid.For by this theory dissolution dependsgreatly upon cohesion, and where cohesion is small dissolution is easy,and vice we?-& ; taking barium sulphate as an example, the cohesionis great, the solubility almost nil ; if, however, the cohesion is dimin-ished by any means, then tlie finely-divided salt will remain suspendedin water for a long time, that is, in a state of pseudo-solution, whichshows that the water molecules alone were not able to overcome thecohesion, but this being to a certain extent overcome, pseudo-soh~tionis the result.Supposing now the insoluble salt could be resolvedinto its molecules, t h a t is, further subdivided, then it is easy tfo con-ceive that it would be possible to dissolve it to a great extent in water,and produce a trite soZuticJw, from which the solid would separatebut slowly, owing to the solid molecules seldom comiug in contact insuiiicient numbers for their niutual attraction to overcome that of thewater for them. As examples of such cases, the author refers to thefact that many almost insoluble compounds are precipitated withextreme slowness from cold dilute solutions.D. A. L,Rise of Solutions in Capillary Tubes. By M. GOLDSTEIN andA. DAMSKI (Juuy. Buss. Chew. Soc., 16, 642-643).-According toValson, the rise of a large number of solutions of salts in capillarytubes is inversely proportional to their specific gravities. Thiserroneous conclusion is explained by the fact that Valson workedwith solutions showing no great diffei-ences in sp. gr. and, therefore,in the rise. Very different results are obtained on using solutions ofgreat concentration (2 or 3 gram-mols. of salt to 1 litre of water) andnarrower tubes ; here the specific gravities and rise in the tubes differconspicuously from those of pure water, and the regularity, shown byValson, does not existi, e.g. :116 ABSTRACTS OF CEEMlCAL PAPERS.Height of rise.h.118.2117.49 ) # 9 7 ,, ..117-3,, . . 116.4 :: f :: ,, . . 11.5.3,, 2 9 , ,, . 114.17 2 3 9 , ,, . . 112.4Pure water. . . . . . . . . . .KCl $ mol. wt. to 1 litre-Sp. gr.d.1.0001.0091.0161.0251-0481.1001-155h . d.118.2118.4119.1119.31120.8125.5199.8The value h . d equals that for water only for very dilute solutions.The determinations of the above values for pot,assium chloride,bromide, and iodide have shown that the rise of solutions of potassiumbromide of different degrees of conceiitration is the mean of the riseof corresponding soh tions of potassium iodide and chloride, themolecular weight of KBr being t,he mean between those of KC1 andKI :-Height of rise.KC1.KI. KBr.w-calc. fromSolutions. a. 6. cc+b = c*2 2 mol. in 1 litre.. 117.4 - -9 , .. 117.3 114.5 115.57 9 . . 116.4 113.0 114.797 , . 115.3 108.5 111-92 7, .. 114.1 100.2 107.13 ?, .. 112.4 93-5 lu2-917 -foullaC.115-911 5.2114.6111.310 7.1102.8B. B.Capillary Phenomena in Relation to Constitution andMolecular Weight. By J. TRAUBE (Bw., 17, 2294-2316).-Allthe experiments described in this paper were made with aqueous solu-tions of organic substances : the advantages of the use of such solu-tions over organic liquid compounds, being the much greater heightat which the former stand in capillary tubes, and the much greaterdifferences in capillary height shown in the case of closely relatedsubstances. Voluminous tables are given, a t once showing the dif-ference in capillary height caused by difference in the concentration ofsolutions of the same substance, and comparing the capillary heightsof unlike substances in solutioris of the same degree of concen-tration. The following are amongst the more important conclusionsdrawn from these experiments :-1.The capillary height of the solu-tion of an orgariic body decreases with increasing concentration.Equnl differences of height are not, however, caused by equal incre-ments of concentration, the rate of difference first attaining a maxi-mum, and then diminishing. 2. I n a homologous series, the capillaryheivht diminishes with increasing molecular weight. The differencereaches its maximum sooner in more concentrated than in moredilute solutions.8. Isomeric Substances, although of related consti-tutions, have not necessarily equal capillary heights. With regardto the capillary relatlions of different organic series, the author giveGENERAL AND PHYSICAL CHEXISTRT. 117the following as the result of his observations:-An increase incapillary height is observed in passing, 1, from the fatty alcohols tothe correspoiiding aldehydes or acids ; 2, from the fatty acids to thehydroxy-acids; 3, from the monohydric to the di- and tri-hydricalcohols ; 4, from the normal and iso-alcohols to the tertiary alcohols ;5, from the ethereal salts of formic acid t o isomeric ethereal salts ofthe higher fatty acids; 6, from compounds of the propyl series tothose of the ally1 series.Probably an increase in capillary heightalso occurs in passing from aldehydes to isomeric ketones, and fromfatty acids to their monosubstituted halogen-derivatives, although onfurther substitution a decrease occurs ; f iirther observations arerequired on these two points. Aldehydes show a lower ca,pillaryheight than the corresponding fatty acids in concentrated solutions,but in dilute solutions the reverse is observed. Normal alcoholsshow a lower capillary height than iso-alcohols in concentrated solu-tions. A. J. G.Mutual Relations of the Physical ,Properties of the Ele-ments. By H. FRITZ (Ber., 17, 2160-2165).-This paper contains atable of most of the heavy metals, with their melting points, specificgravities, atomic weights, and specific heats ; from these, it may beshown by calculation that the product of the atomic heat b,y the relativeheat i s equal to the cube root of the product of the melting point mul-tiplied by the spec$% heat, As .Ds = V%; A being the atomicweight, s the specific heat, D the sp. gr., and t the melting point.The elements are arranged in groups in which different values aresubstituted for t. In the case of lithium, sodium, and potassium, thevalue ’ - i- 50 is substituted for t ; in the case of magnesium and2.50aluminium the value tG, and in that of strontium and barium7.4t + f;o30 *If the metals be arranged I, according to the amount of heatliberated by their union with oxygen aud chlorine, and IT, accordingto their conductivity for heat, the one series will be found to bethe revarse of the other.A. I(. M.A General Statement of the Laws of Chemical Equilibrium.By H. LE CHATELIER (Compt. rend., 99, 786--789).-The authorextends and modifies Van tl’Hoff’s general statement of the chemicalequilibrium of a system by including in it t,he “ condensation ” of thesystem, that is, pressure, concentration, number of molecules in unitvolume, &c., and by giving it a form similar to that of the lawsrelating to changes of equilibrium which effect mechanical work.Reversible chemical changes are thus brought into the class ofreciprocal phenomena.When a system in stable chemical equilibrium is acted on by anexternal cause which tends to alter the temperature or condensationeither of the whole system or of some of its parts, the system can onlIlls ABSTRACTS OF CHEMICAL PAPERS.undergo such inLerna1 modifications as would, if they had taken placespontaneously, have produced a change of temperature or con-densation of the contrary sign to that resulting from the action of theexternal cause.These modifications are generally progressive and incomplete.They are, however, sudden and complete if they can take place with-out changing the individual condensation of the different homo-geneous parts of the system, whilst a t the same time they alter thecondensation of the system as a whole,They are nil when their occurrence cannot, produce changesanalogous to those due to the external cause.Although modifications may be possible, they do not necessarilytake place. I n cases where no change occurs arid the system remainsunaltered, the original stable eqnilibr*ium becomes unstable, and thesystem can then only undergo such modifications as tend to bring itback t o stable equilibrium. Many well-known reactions, includingthe phenomena of fusion, evaporation, solution, &c., are cited asexamples. C. H. B