The Electrical Conductivity of Solutions of some Fatty Acids in Water and in Alcohols. By K. HARTWIG (Ann. Phys. Chem. [2], 33, 58-80). The author points out that comparatively few investigations have been made into the conductivity of other than aqueous solutions. Of those known to the author, Mateucci (Ann. Chim. Phys., 66, 237) was the first who made experiments on alcoholic solutions, but Wiedemann has thrown doubt upon his conclusion that aqueous and alcoholic solutions of the same substance of the same specific weight conduct equally well. Oberbeck (Ann. Phys. Chem., 155, 595) measured the resistances of aqueous and alcoholic solutions of cadmium bromide and cupric chloride, and found that the conductivity was increased by each of these salts in a manner dependent on the salt and on the solvent. Guglielmo (Atti R. Accad. Torino, 17) determined the conductivity of an alcoholic solution of potassium hydroxide, that of the aqueous solution having been previously determined by F. Kohlrausch. Vincentini (Mem. R. Accad. Torino [2], 36, 22) investigated the conductivity of alcoholic solutions of some chlorides, and found that there is no simple relation between the solubility and the conductivity. Bartoli (l'Orosi, 7, 3) showed that paraffin and naphthalene can be made into conductors by the addition of amylic alcohol and phenol. The same author (R. Acad. Lincei, 1, 550) investigated the conductivity of various mixtures of organic compounds; and Lenz (Mem. Ac. Sci. St. Pétersbourg, 7, 30) made some researches in the conductivity of aqueous and alcoholic solutions, in the course of which he completely demonstrated the falseness of Mateucci's conclusions. The author selected the fatty acids for his experiments, as being very soluble in several media.

The measurements were made by the Wheatstone bridge method, with a telephone as indicator. The substances experimented on were solutions of formic, acetic, and butyric acids in water, and in methyl, ethyl, and amyl alcohols. The measurements were made immediately after the solutions were made, and as none of them occupied more than an hour, it was found that the results were not sensibly affected by the etheritication which always takes place when organic acids and alcohols are mixed.

With the exception of two of the solutions of formic acid, the author finds that the conductivity reaches a maximum at a certain concentration, and that this maximum is reached the sooner the worse the conductivity of the acid.

The table below gives the percentage of acid for which each solution has the maximum conductivity :

Thus the greater the quantity of carbon present in the acid the sooner is the maximum conductivity attained, and the greater the quantity of carbon present in the solvent the later is the maximum attained, the conductivity being diminished by an increase of the quantity of carbon present either in the acid or in the solvent.

The author attributes the anomalous behaviour of formic acid with regard to electrical conductivity, together with similar anomalies which it exhibits with respect to its other physical properties, to the absence of the group methyl from its composition.

G. W. T.

Electric Leakage. By J. J. THOMSON and H. F. NEWALL (Proc. Roy. Soc., 42, 410-429).-The liquid experimented on is contained in a cylindrical, metallic vessel, connected to earth, in which a metal cylinder is suspended by means of a silk thread; this can be connected either with a battery or with a quadrant electrometer. The inner cylinder, after being charged, is connected with the electrometer, and readings taken every five seconds. Curves are plotted showing (1) the decrease of potential with time; (2) the ratios of successive potential values. The liquids examined were benzene, olive oil, carbon bisulphide and paraffin oil, and were filtered many times before use. With the first three, no deviation from Ohm's law could be detected. With paraffin oil, the conductivity is slightly greater with large than with small differences of potential. These results, where the E.M.F.'s were 20-100 volts, differ from those of Quincke (Abstr., 1886, 959), who finds that with E.M.F.'s sufficient to produce a spark the conduction does not even approximate to Ohm's law. With carbon bisulphide, great discrepancies were found in the first results; these were traced to differences in time of charging, and later experiments proved that electric absorption took place. The effect was greatest after redistilling the carbon bisulphide, but at times it was totally absent. The conductivity of all the liquids was increased by rise of temperature, so that in this respect they resemble electrolytes. H. K. T.

The Influence of Magnetic Forces on the Nature of the Heat Conductivity of Bismuth. By A. V. ETTINGSHAUSEN (Ann. Phys. Chem. [2], 33, 129–136).—The author points out that Righi (Abstr., 1887, 1009) and Leduc (Compt. rend., 104, 1783; 105, 250) have made some experiments, from which they conclude that the thermal conductivity of bismuth in a magnetic field is diminished to an equal extent with the electrical conductivity, supposing the magnetic lines of force to cut the stream lines for heat or electricity respectively at right angles. Nernst (Ann. Phys. Chem. [2], 31, 760) was unable to detect any change in the thermal conductivity. The author, after some experiments made with great care, finds that the thermal conductivity is diminished, but to a much less degree than the electrical conductivity. The decrease in the thermal conductivity is greater when the bismuth is impure, but still much less than that of the electrical conductivity. G. W. T.

Constancy of the Heat Produced by the Reaction of Silver Nitrate with Solutions of Metallic Chlorides.

By T. W.

RICHARDS (Chem. News, 57, 16-17).-In these experiments, all necessary precautions are observed to obtain comparable conditions: 250 c.c. of the silver nitrate solution (equal to 4 grams silver nitrate) and 250 c.c. of the salt solution (containing a gram or so of the salt in excess of that required to precipitate all the silver) are poured simultaneously into the platinum calorimeter (Berthelot's), and the rise in temperature noted. The average rise in 20 experiments amounts to 16 165 cal. and all the results are practically identical; the experiments included sodium, potassium, ammonium, barium, cupric, zinc, manganous, nickelous, ferrous, aluminic, ferric, and chromic chlorides and hydrochloric acid. Taking the following equation as representing the reaction, the author concludes from these results that the amount of heat is constant no matter what R m or n

may be. (RmCl, + AgNO, + Aq) = AgCl + [†R„(NO2)n = Aq]. (Compare Pickering, this vol. p. 333.)

D. A. L.

Relation between the Heats of Formation of Chlorides and Sulphates in Aqueous Solution. By I. W. FAY (Chem. News, 57, 36-37).-The author conducted a series of experiments with barium chloride and soluble sulphates exactly similar to Richard's experiments with silver nitrate and chlorides (see preceding Abstract). There is not that general regularity observed in the case of the sulphates as with the nitrates, although the rise for allied bases is quite close; the sesquioxide salts give a greater rise of temperature than the protoxide salts do, and with the double salts a still larger quantity of heat is evolved, whilst sulphuric acid gives most of all. The sulphates included in the experiments were sodium, potassium, ammonium, magnesium, zinc, cadmium, copper, nickel, cobalt, ferrous and ferric, and aluminium potassium alum, potassium chrome-alum, ammonium ferrous sulphate, and sulphuric acid. The results are tabulated in the original paper. (Compare Pickering, this vol., p. 333.)

D. A. L.

Alteration in the Volumes, and Density of Liquids by the Absorption of Gases. By K. ANGSTRÖM (Ann. Phys. Chem. [2], 33, 223-233).-The author has already given an account of investigations on expansion of water through the absorption of gases (Abstr., 1882, 687). Since this appeared, experiments of the same nature have been made both for water and ethyl alcohol, by Blümcke (Abstr., 1885, 215; 1887, 435), but they do not agree with those made by the author or by Mackenzie and Nichols (Abstr., 1878, 366) and Nichols and Wheeler (Phil. Mag., 5, 11, 113), probably because Blümcke's method did not allow of several requisite corrections being made; for example, for the compression of the areometer, and for the refraction of light at the surface of the containing vessel.

A summary of the results of the author's previous experiments and of the first series described in the present paper are given in the table below, in which d1, d2, d, are coefficients of expansion due to absorption.

This shows that, except for ether, the ratios /, and /d3 are sensibly constant. Experiments on the successive absorption of two or more gases gave the result that the coefficient of expansion due to absorption is independent of any previous absorption of gas, the total expansion being equal to the sum of the expansions due to the absorption of the different gases.

The results obtained enable the change in the specific weight of a liquid due to the absorption of gas to be calculated, which is of practical importance for specific weight determinations, but the effect in such cases is so small that it need only be taken into account when the most extreme accuracy is desired. The author considers the constancy of the ratios / and 2/3 to be the most important of his results from a theoretical point of view, as it shows that the dilatation depends on the gas, and that the relations between the coefficients of dilatation due to absorption are independent of the nature of the liquid; from this he concludes that Ostwald's hypothesis (Stöchiometrie, p. 356), founded on the values of the coefficient of dilatation due to absorption obtained by the author and by Sarrau, namely, that "the volume of the absorbed gas is almost exactly reduced to the volume of its molecules," is à priori extremely improbable. G. W. T.

Differential Tonometer. By G. J. W. BREMER (Rec. Trav. Chim., 6, 122-136). An apparatus for measuring the difference between the vapour-tension of a saline solution and the vapour-tension of water by allowing the vapours to act on the opposite ends of a column of olive oil. It consists essentially of small flasks connected with vertical tubes about 150 mm. high communicating with one another by means of a horizontal tube at the bottom. The particular apparatus described consists of four such flasks. The flasks should have exactly the same cubical contents, which may be adjusted by introducing pieces of glass rod, and should contain equal volumes of the solutions. After introduction of the liquids the flasks are cooled to 0°, rendered partially vacuous, and oil allowed to enter from the bottom horizontal tube until it rises to about half the height of the vertical tube. The exact pressure of the air in the flasks is observed, and the flasks and their contents are heated to different

temperatures and the differences between the levels of the oil in the tubes are observed.

Determinations of the vapour-tensions of aqueous solutions of calcium chloride of different strengths show that if the salt is regarded as existing in the liquid in the anhydrous condition, the reduction of the vapour-tension increases more rapidly than the amount of calcium chloride present, but if the solution is supposed to contain the hydrate CaCl2 + 6H2O, the reduction of the vapour-tension is proportional to the amount of salt present, as Wullner has previously stated. C. H. B.

Dynamical Method of Determining Vapour-pressure. By G. TAMMANN (Ann. Phys. Chem. [2], 33, 322-337).-The author on recalculating some tables given by Regnault (Ann. Chim. Phys. [3], 15, 158) for the vapour-pressure of pure water determined by the dynamical and statical methods respectively, found that the close agreement of the results was only apparent, and that there were really some considerabe discrepancies which could not be accounted for. The author now describes a number of experiments carried out principally with a view to testing the dynamical method, following' Regnault with some modifications in details. He concludes that the method is not capable of giving good results. In the case of unsaturated solutions, when the current of air is passed over the surface of the solution, the upper layers become more concentrated than the lower, through the evaporation being entirely superficial, and if the current of air is led through the solution errors are introduced, as it is impossible to prevent variation of pressure and the scattering of drops of the solution upon the interior of the glass vessels. The author attempted to use it with saturated solutions, using filter-paper soaked in the solution, then partly dried, and introduced into a U-tube through which the current of dry air was allowed to pass, but then he was met by the difficulty that on taking the tube out of the bath used to maintain the required temperature, drops of water condensed on the inner surface of the tube, and dissolving some of the salt adhering to it gave rise to unsaturated solutions. The author next tried the method with hydrated salts in crystals. Here, however, the results were found to depend on the velocity of the stream of air, which appeared to be due to the water-vapour given off dissolving a portion of the crystals and forming hydrates containing less water, which would give a lower vapour-pressure.

From some of the results obtained during the last series of experiments, combined with Pape's result (Ann. Phys. Chem., 124, 329) that rates of evaporation from the different surfaces of irregular crystals are not the same, the author draws the conclusion that the vapour-pressure at the surface of a crystal is not the maximum value of vapour-pressure due to the crystal. The vapour-pressure at the surface depends on the nature of the surface, and is considerably less when the surface is uninjured than when it is broken or otherwise changed. G. W. T.