85 General and Physical Chemistry. Refraction and Density. By ISIDOR TRAUBE (Bey., 1896, 29, 2732-2742).-The expression V,,, = mid = SnC + Cov = SnC + 25.9 (see Abstr., 1896, ii, 235) was termed by Kopp the apparent molecular volume, but the author prefers to call it the molecular vibration volume ; it is composed of the true molecular volume SnC, which the author terms the atomic nucleus volume, and the niolecular covolurne Cov. The values of V,,, are obtained from density determi- nation, whilst an expression for the value of SnC is given by the Clausius-Mosotti theory of dielectrica ; if v is the actual volume occu- pied by spherical molecules, I% the dielectricity constant, and p the refractive index for infinite wave-length, the fraction of unit space actually occupied by a molecule is 8nC/Vm, and it should be proportional to (p2 - 1)/(p2 + 2) if the argument be valid.The values of SnC (pz + Z)/Vm (p2 - 1) have been cal- culated for a long series of saturated organic compounds, and are con- stant wikhin very narrow limits, the mean values being 3.53 for Cauchy's constant A, 3.44 for the D line, and 3.460 for H, ; the values of SnC are calculated from the atomic nucleus volumes c! = 9-99, H = 3.1, and 0 = 5.5. Clonsequently, the vibration volumes of the atoms calcu- lated from molecular weight and density are equal to the atomic nucleus volumes or atomic refractions multiplied by constants which vary within narrow limits with the wave-length of the light. The quantity p - 1 is the " loss of time " experienced by a ray in traversing a thickness of the substance, instead of an equally wide vacuous space which is traversed in unit time.The loss of time is also very strictly proportional to that fraction of the space filled by a substance which represents the molecular nucleus volume, so that S92C/Vln ( p - I) is practically constant; this quantity has the mean values 2.086 for Ha, and 2.073 for D, and knowing the density of a given substance, it is possible to calculate with fair approximation the refractive indices for these rays. It is evident that the two above expressions involving ZnC, V, and functions of p are independent of the temperature. On dividing the atomic vibration volumes stated above by 2.086 or 3'460, the atomic refractions for Ha of the corresponding atoms for the Gladstone, or the Lorenz and Lorentz, formula respectively are obtained ; the numbers thus got for the simple formula agree more closely with the numbers in use than do those for the theoretical formula.By applying the formula SnC (p2 + 2)/Vm(p2 - I) to unsatu- rated compounds, it is possible to calculate the increments of mole- cular refraction due to double linkings between the carbon atoms with very fair approximation. Action of Light on Dyed Colours. Report of British Asso- ciation Committee, Section B, Liverpool, 1896 ( C l ~ m . News, 1896, 74, 21 = (k - I ) / ( k + 2) = (p2 - 1)/(p2 + 2), W. J. P. VQL. LXXII. ii, 786 ABSTRACTS OF CHEMICAL PAPERS, 205-207, 218).-The report refers to blue and green colouring matters on wool and silk exposed to light under glass with free access of air and moisture, and examined by comparison with standards at various intervals; a very large number of samples were examined and the general results are tabulated under five classes :-I? Very Fugitive ; 11, Fugitive; 111, Moderately Fast; IV, Fast, and V, Very Fast Colours.Interesting comments and notes are also given. By WILHELM JAEGER and R. WACHS- MUTH (Ann. Phys. Chem., 1896, [ Z ] , 59, 575-591).-The use of cad- mium in place of zinc for the construction of standard cells of the Clark type has been proposed on account of the reduction that is thus attained in the temperature coefficient of the cell. The authors have submitted this point to further investigation, and have also ascertained the ratio of the E.M.F. of the cadmium cell to that of the Clark cell, and the influence of impurities and of time on the constancy of the cadmium cell.The cell is best made of the usual H pattern, with electrolytically amalgamated platinum electrodes sealed through the glass. I n place of pure cadmium, an amalgam of 1 part by weight of aadmium to 6 parts of merciiry is used (m, p. 60'). The mercurous sulphate is mixed with crystals and a concentrated solution of cad- mium sulphate and some mercury to a thick paste, and this paste is placed over the mercury of the positive pole. The negativecpole (1Cd: 6Hg) is covered with a layer of cadmium sulphate crystals, and the rest of the cell is then filled with a concentrated solution of cadmium sulphate. The E.M.F. of this cell a t 20' is 1,019 volt. The change in E.M.F.with change in temperature between 5Oand 2 6 O is expressed by the equation and is, therefore, for lo, only about i&.G per cent. I n the effect of the influence of impurities and in durability, the cadmium cell was found to be in no way inferior to the Clark cell. D. A. L. Cadmium Standard Cell. Et = E,, - 3.8 x 1 0 - 5 (t - 20) - 0.065 x (t - 20)2, H. C. Heat of Formation of Lithium Hydride. By ANTOINE GUNTZ (Compt. rend., 1896, 123, 694-696. Compare Abstr., 1896, ii, 359). -The heat of formation of lithium hydride was calculated from the heats of dissolution of the hydride and of the metal itself in water. Heat of dissolution of lithium hydridein water a t 1 8 O = + 31.6 Cal. Heat of dissoliition of lithium in water at 1 8 O = + 53.2 Cal.Therefore, Li solid+H gas=LiH solid= + 21-6 Cal., a number which explains the stability of the compound. The fact that the above value for the heat of dissolution of lithium in water is much greater than that obtained by Thomsen, namely, + 49.08 Cal., is attributed by the author to the much greater purity of the metal used by him. The heats of formation of lithium compounds deduced from Thomsen's number are accordingly too low. Lithium hydride melts at 680°, at which temperature its tension of dissociation is about 2'7 mm. At ordinary temperatures, dry chlorine has no action on it ; when heated in a current of that gas to a dull red heat, it burns, giving lithium chloride and hydrogen chloride. At a redGENERAL AND PHYSICAL CHEMISTRY. 87 heat, hydrogen chloride acts on it, forming lithium chloride and hydrogen. It reacts slowly with absolute alcohol, but is without action on dry benzene, toluene, or petroleum, at ordinary temperatures.Lithium hydride difYers markedly in its properties from the hgdrides of sodium and potassium. A. C. C. By PAUL LEMOULT (Conzpt. rend., 1896, 123, 559-562).-The cyanamide used was prepared by the action of mercuric oxide on thiocarbamide, and its molecular weight was determined by cryometric observations on its solution in acetic acid. Combustion in the calorimetric bomb gives 4090 Cal. as the heat of combustion of 1 gram ; the molecular heat of combustion is, therefore, 171.78 Cd. at const. vol,, and 171.5 Cal. at const. pressure, the heats of formation from its elements being - 8.4 Cnl.and - 8.2 Gal. respectively. Its heat of dissolution in water is -3.59 Cal., and hence its heat of formation in solution is - 12.0 Cal., a result which explains the greater stability of the compound when in the solid state. From these data, it follows that the action of mercuric oxide on thiocarbamide develops + 25.2 Cal., whilst the conversion of cyanamide in dilute solution into carbamide would develop +20.2 Cal. The addition of dilute acids to solutions of cyanamide causes no thermal disturbance, but the heat of neutralisation by potassium hy- droxide (one molecular proportion) is +3*79 Cal., and by sodium hy- droxide + 3.6 Gal. I n either case, the addition of a second molecular proportion of alkali causes a slight development of heat, but a third proportion has no effect.The heat of neutralisation by ammonia is + 1.38 Cal. It follows that in aqueous solution cyanamide behaves as an acid, the energy of the acidic function being comparable with that of hydrocyanic acid. It would seem also that there is a difference between the functions of the two replaceable atoms of hydrogen. Thermochemistry of Cyanamide. C. H. B. Thermochemistry of Hexamethylenetetramine and its Ni- troso-derivatives. By MARCEL DEL~PINE (Compt. rend., 1896, 123, 650--653).-The heat of combustion of pure hexamethylenetetramine, determined by burning in the calorimetric bomb, is 1006.53 Cal. (const. press,), and its heat of formation from its elements - 26.73 Cal. Its heat of dissolution in water a t 15' (1 mol. in 1.5 to 2.5 litres of water) is 4.8 Cal., from which the heat of formstion of dissolved hexa- methyleneamine is - 21.93 CaI.In order to verify the accuracy of the latter number, it was employed in calculating the heat of formation of hexamethylenetetramine dinitrate, and the result, 92.3 Cal., compared with that obtained by direct experiment +92*94 Cal. It is incidentally remarked that hexamethylenetetrsmine dinitrate (47.37 per cent. HNO,! gives, on combustion, very little nitric acid. First Nit~/.oso-derivcctzve, (CH2)6(NO)2N4.-Heat of combustion (const. press.) = 872.28 Ual., from which its heat of formation (cryst.) is found to be - 55.78 Cal, Xecmd Nitroso-derivative, (CH,),( NO),N, (com pare Abstr., 1 8 8 9,3 3). 7-288 AHSTRACTS O F CHEMICAT, PAPERS. -Heat of combustion (const. press.) = 745.96 Cal., from which its heat of formation is found to be - 91.76 Cal.The replacement, therefore, of CH, by (NO), diminishes consider- ably the heat of formation of these substances, and renders it strongly endothermic, the second4 substitution producing a greater diminution than the first. The nitroso-derivatives are much less stable than the base itself. A. C. C. Dependence of the Dissociation of some Acids on Tem- perature and the Heat of Dissociation. By HANS EULER (Zeits. physikal Chem., 1896, 21, 257-271).-Determinations of the con- ductivity were made by Kohlrausch's method in the case of benzoic, toluic, salicylic, metahydroxybenzoic, ortho- and meta-nitrobenzoic, and dichloracetic acids, at various temperatures between loo and 50", and at various dilutions.Interpolation formulze mere calculated which, in almost all cases, except with the sodium salts, were of the form X = a + bt - ct2, indicating the occurrence of a maximum (compare Jahn and Schriider, Abstr., 1895, ii, 203). From these values, the dissocia- tion constants are calculated at the different dilutions and temperatures, the values from the different dilutions agreeing satisfactorily with one another, except in the case of or.thonitrobenzoic acid and dichloracetic acid, where, however, the value a t 0" was undoubtedly higher than that a t 2 5 O , results not in accord with those of Wildermann. I n all cases, the dissociation constant is a function of the temperature, a maximum occurring for benzoic acid at about 3 5 O , and for metahydroxy- benzoic acid a t about 28'.For orthotoluic acid, the values decrease continuously as the temperature rises, but increase for salicylic and metanitrobenzoic acids. Calculation of the heat of dissociation of the acids proved it to invariably increase with rise of temperature. I n most cases, it is at 6rst negative, the temperature a t which zero is reached being thatof the maximum dissociation, nrelation indicated by the expres- sion Q = 0.50804 Tl.l/k.dk/dt. from which the values are calculated. L. &I. J. The Determination of Molecular Weights. IV. By ERNST 0. BECKMANN (Zeits. pJhysikal. Chew,., 1896, 21, 239-256).--The paper contains further details of the apparatus and methods employed by the author. An electromagnetjc stirrer is described, for use i n freezing point determinations, whereby the apparatus can be kept, completely closed and the entrance of moist air avoided, an important precaution when phenol or acetic acid are employed as solvents.A form of boiling point apparatus is described, small tetrahedra of platinum foil being recommended for the prevention of bumping ; and the author states that in the Beckmann thermometers a conical junction of the capillary tube to the reservoir is necessary. Experiments are recorded indicating the availability of the apparatus as described, aniline, water, chloro- form, ether, ethglic alcohol, and benzene being employed as solvents (comp. Abstr., 1895, ii, 154, 382; 1896, ii, 236). Exact Cryometry : Application to Sodium Chloride Solu- By FRAN~OIS M. RAOULT (Compt. Yencl., 1896, 123, 475-478). L.M. J. tions,GENERAL AND PHYSICAL CHEMISTRY. 89 -The author has made several series of determinations of the freezing points of solutions of sodium chloride, using the apparatus previously described (Compt. ?*end., 1896, 122), but employing ether instead of car- bon bisulphide. It mas found that the temperature could be kept constant to 0.1" at any point between - 15" and the surrounding temperature for several hours. In the following table, P is the weight of salt in 100 grams of water; C, the apparent reduction when the converging temperature is 3.5" below the freezing point; C,, the real reduction when the converging temperature and the freezing point coincide. P c, co c,-c, co x 0.002 5.1350 3.4435 3.4381 0.0054 0.0068 2.859 1.6880 1.6839 0.0041 0.0034 1.400 0,8286 0,8267 0.0019 0.0017 0.690 0.4132 0.41 11 0.002 1 0*0008 0.341.0.210'7 0.2093 0.0014 0.0004 0-176 0.1113 0.1111 0.0002 0*0002 C, x 0.002 and C,-C, are practically identical, and hence C,- C, = C, x 0.002 or C, = C,(1 + 0*002), and the general expression C, = C,(1 + q) previously arrived a t (q having always a very small value) is experimentally verified for sodium chloride solutions. The real and apparent molecular reductions can be calculated from the figures in the table ; the limitiug value is 38-05 for the former and 37.88 for the latter. The curves, with the observed reductions for abscissae and molecular reductions for ordinates, are very similar in the two cases, and cut the axis of the ordinates at practically the same point, which corresponds with the limiting molecular reduction.The experiments with sodium chloride, therefore, confirm the author's pre- vious conclusion that the temperature of the surroundings has no influence on the laws relating to the reductions of the freezing point of different solutions of the same substance. The real molecular reductions given in the table correspond to a superfusion of 0*5", and the absolute values, when the concentration is not altered by freezing, are obtained by multiplying the figures by 0994. The limit molecular reduction for sodium chloride is then found to be 37.82, which is identical with that requiring complete ionisation. The different results obtained by Ponsot (Abstr., 1896, ii, 411, 636) are attributable to insufficient agitation of the liquid, especially in a vertical direction.C. H. B. Exact Cryometry : A Correction. By FRANFOIS M. RAOULT (Compt. ~ e i z d . , 1896, 123, 631--632).-After some reference to an earlier paper (preceding abstract), it is remarked that, in the deter- mination of the true depression of the freezing point, absence of radia- tion is not theoretically necessary, which is fortunate, since that condition is abgolutely incapable of realisation, owing to the develop- ment of heat produced by the agitation of the liquid. A. C. C. Expansion during the Dissolution of Ammonium Salts and of Sodium Thiosulphate. By HUGO SCHIFF and U. MONSACCHI90 ABSTRACTS OF CHEMICAL PAPERS. (Zeits. physikal. Chem., 1896, 21,277-296).-The expansion occurring during the dissolution of ammoniacal salts was first determined, pyknometers of 25 C.C.and 50 C.C. capacity being employed. I n the case of ammonium nitrate, the expansion (throughout referred to 100 parts of the mixed constituents) was found to vary from 4.0 for a 63 per cent. solution to 0.179 for a 4 per cent., with a minimum of 0.119 at 7 per cent. As the expansion may be due to a dissociation into ammonia and nitric acid, the effect of dissolution in dilute nitric acid was determined, but the expansion was found to be even greater than in aqueous solution, a result also obtained by dissolution in potassium nitrate and ammonium chloride solutions. The expansion may be fairly well calculated by assuming that a saturated solution of the ammonium nitrate is first formed, and then mixed with a solution containing the other salt and the remaining water.I n methylic alcohol, however, a contraction of about 0.86 occurs for the saturated solution (14 per cent.). Ammonium chloride gave an expansion of 2.6 at 30 per cent. and 0.46 a t 10 per cent.., that of ammonium bromide being in each case slightly lower. I n the case of ammonium iodide, however, contraction occurs, varying from 1.4 at 60 per cent. to 0.035 at 3 per cent,, and is greater in alcohol than in water. Hydroxylamine hydrochloride gave an anomalous result, contraction first occurring reaching a maximum a t from 10 to 15 per cent. with normalvolume a t 28 per cent., after which expansion occurs, whilst in the case of hydra- zine hydrochloride a perfectly regular contraction obtains.Sodium thiosulphate gave results very similar to that of the hydroxylamine salt, the maximum contraction occurring at 40 per cent. and zero at 78 per cent., after which expansion occurs, the values being not quite concordant with those of Boisbaudran (Abstr., 1895, ii, 486), partly owing to the latter using a higher sp. gr. for the solid salt (1.752) than that obtained by the authors (1.734). Laws of Irreversible Processes. By LADISLAUS NATANSON (Zeits. physikal. Chem., 1896, 21, 193- 21 7).-A mathematical paper, unsuit- able for abstraction, in which the author deduces expressions for the velocity of various irreversible processes, such as diffusion, heat con- ductivity, dissipation of electrical energy. Relationship of the Rate of Diffusion to the Initial Concen- tration of Dilute Solutions.By W. KAWALKI (Arm. phys. Chem., 1896, [ Z ] , 59, 637-65!).--Nernst has shown, as a consequence of the dissociation theory, that the rate of diffusion of dissolved substances should alter very little with the concentration when the solutions reach a certain degree of concentration. I n former experiments (Abstr., 1894, ii, 345), the author found that with very small initial concentrations the values calculated for the diffusion coefficient k do not correspond with one another, andit was pointed out that the be- haviour was probably due to convection currents. Careful experiments with dilute solutions of sodium acetate and carbamide have since served to confirm this view and support Nernst’s conclusions. The diff nsion coefficient, k, for dilute alcoholic solutions was also deter- mined, and the ratio kl/k found for sodium acetate=2.19, and for carbamide = 1-79.H. C. L. N. J. L. M. J.GENERAL AND PHYSICAL CHEMISTRY. 91 Determination of Isosmotic Concentrations. By SVEN G. HEDIN (Zeds. physikul. Chew,., 1896, 21, 272-276).-Chiefly a contro- versial paper and a criticism of Koppe's results (Abstr., 1895, ii, 208). The author considers that errors occur in Koppe's work owing to (1) the use of salts which directly affect the blood corpuscles, (2) the use of standard solutions which are not actually isosmotic, whilst, further, he considers it necessary to defibrinate the blood employed. L. M. J. Explosive Properties of Acetylene. By MARCELLIN P. E. BERTHELOT and PAUL VIEILLE (Compt.vend., 1896, 123, 523-530).- When acetylene under ordinary pressure is subjected to the action of an electric spark, a red-hot wire, or a discharge of fulminate, the decom- position of the gas does not extend beyond the immediate neighbour- hood of the source of decomposition, but under pressure the results are different, and when the pressure exceeds two atmospheres the gas shows the ordinary properties of explosive mixtures. Under these conditions, if decomposition is produced at any point by one of the methods indicated, it very rapidly spreads tbrough the whole mass of the gas, which is thereby resolved into hydrogen and bulky pulverulent carbon. Under an initial pressure of 21 kilos. per square cm., the pressure developed by the decomposition is 10 times as great, and the change is complete in 0.018 of a second.The ratio of the final to the initial pressure decreases, and the time required for complete decom- position increases, the lower the pressure. Even with an initial pres- sure of 21 kilos. per square cm., the rate of propagation of the change is much below the velocity of the explosive wave in the oxyhydrogen mixture. The calculated temperature of decomposition is 2750", and the calculated pressure 11 times as great as the initial pressure. The observed pressure agrees fairly well with the calculated. Liquefied acetylene decomposes in the same way as the gas ; with 18 grams of the liquid in a bomb of 48.96 C.C. capacity, the final pressure was 5,564 kilos per square cm., and under these conditions the explo- sive force is nearly equal to that of guncotton.The decomposition of the liquid is, however, relatively slow when excited by simple ignition. When the bomb contains both liquid and gas, there is a change in the curve of pressure which indicates two distinct phases of the explosion, one most probably corresponding with the decomposition of the gaseous part, and the other, which lasts longer and raises the pressure much higher, to the decomposition of the liquid portion. Mere shock, such as is caused by a fall from a considerable height, seems incapable of causing the explosion of either compressed or liquefied acetylene. If the vessel breaks there is still no explosion in the case of the compressed gas ; but if the vessel contains liquid acety- lene, the fracture is followed after a short interval by an explosion.The latter, moreover, differs essentially from the explosive decomposi- tion of the gas, and is not accompanied by the separation of free carbon ; it results from the admixture of air with the acetylene, and the ignition of this mixture by sparks that result from the friction of the breaking metal. If, however, liquid acetylene is decomposed by the discharge of a92 ABSTRACTS OF CHEMICAL PAPERS. small quantity of fulminate contained in the same vessel, violent deto- nation takes place, and the fragments of the vessel have the appearance of those produced by a true explosion. All the fragments are covered with the carbon liberated from the gas. I n the action of small quantities of water on excess of calcium car- bide in a closed vessel, there may be sufficient local elevation of tem- perature to initiate the decomposition of the whole of the compressed gas.This local elevation of temperature may also produce polymerides, which are themselves endothermic. Other causes of dangerous local heating are too rapid compression, or the local pressure that arises when the gas is allowed to escape very suddenly from a vessel in which it is highly compressed. The precautions needed to prevent accidents arising from these causes are obvious. C. H. B. Influence of Pressure on the Inversion Constants of some Acids. By 0. STERN ( A m . I'hys. Chenz., 1896, [ 2],59, 652-663).- The change in the conductivity of electrolytes with an increase in the external pressure has been accounted for on the supposition that the rise in pressure produced an increase in the electrolytic dissociation.Rontgen found, however, that an increase in the pressure does not accelerate, but retards, the rate of inversion of cane sugar by hydro- chloric acid. The author has extended.Riintgen's observations, and finds that the rate of inversion of solutions containing 23 grams of cane sugar per 100 c.c., and varying amounts of hydrochloric, sulphuric, or oxalic acid, is always reduced when the external pressure is increased from 1 to 500 atmospheres. The influence of the pressure on the rate of inversion is smaller, the smaller the amount of acid added. If the inversion is brought about by phosphoric or acetic acid, the reverse is true, as pressure here increases the rate of inversion, and this increase is the greater, the greater the amount of acid added.The rate of inversion decreases somewhat as the concentration of the sugar solutions is decreased, but the change in the rate of inversion with the pressure remains about the same. A rise in temperature of about 10" produces a very marked increase in the rate of inversion, but here, again, the effect of an increase of pressure remains about the same. H. C . Crystal Symmetry. By VIKTOR VON LANG (Zeds. physikal. Chena., 1896, 21, 218-224).-The author draws attention to the simple method which he used many years ago in his Lehduch dey Krystallographie (Wien, 1866) for deducing the thirty-two possible types of crystal symmetry from fundamental crystallographic laws ; the argument is restated and its simplicity urged in favour of its more general use.A system of nomenclature for the thirty-two crystal systems is proposed, which has the advantage over many others that the names indicate immediately the prominent characteristics of the types of symmetry they are intended to describe. W. J. P. Atomic Weights of the Elements. By DELAUNEY (Conyt. vertd., 1896, 123, 600-603).--The author arranges the elements in four groups, according as their atomic weights, expressed by the nearestGENERAL AND PHYSICAL CHEMISTRY. 93 whole numbers, are O+a multiple of 4, 1 +a multiple of 4, 2 +a multiple of 4, or 3 + a multiple of 4. The first group is the largest, and the elements in it can be arranged in three columns, the successive members in each of which differ from the element at the top (C (12), Ce (92), or x (172) by 4, 8, 4, 4, 8, 4, 4, 4, 4, 24.There are gaps due to undiscovered elements or to inaccurate determinations of known elements. The next largest class contains elements whose atomic weights are 3 +a multiple of 4, and here there are two columns, the successive members of which differ from the first in the same way as in the first group, but only three known elements are contained in the second column. On the other hand, several elements that properly belong to this group do not fit into the columnar arrangement. The groap 2 + a multiple of 4 contains eight elements which, starting from the first, helium, increase by 12, 56, 20, 16, 20, 56, 12, whilst the three elements in the groups 1 +a multiple of 4 increase after the first (Be) by 56 and 20. A certain number of the less known elements do not fit into any of these groups. C. H. B. Hypothesis of the Atomic Motion of the Elements and their Genesis. Bly FLAVIAN FLAWITZKY (Zeits. anorg. Chena., 1896, 12, 182--187).-The author advances the hypothesis that the atoms of an element move in curves which lie in planes parallel to one another. The atoms of different elements move in planes which are inclined at certain definite angles to one another. The orientation of the motion determines the character of the element, and can be regarded as due to the influence of some selective dualistic force, such as electricity, in the formation of the element. H. C. A New Form of Turbine for Use in Laboratories. By GEOBGE F. JAUBERT (Bull. Soc. Chinz., 1896, [3], 15, 9--1O).-The author describes a new form of turbine for use in laboratories supplied with water under high pressure. The turbine is designed for carrying out operations on a moderate scale. No details are given with regard to construction. M. W. T.