66 Scientific Intelligence. 



stant; while if the liquid be not monomolecular this constant 

 must be multiplied by x/a where x is the factor of association of 

 the liquid and a that of the vapor. From tabular results of r, d 

 and T given in the paper, the value rd/T is calculated ; and it 

 appears that while the mean value for monomolecular liquids is 

 practically constant at about 0*099, that for associated com- 

 pounds, while consistent with the theory, is considerably greater, 

 that for water being about 3 and acetic acid 2/71. Similar results 

 are obtained from a study of the lowering of the freezing point. 

 While monomolecular compounds in monomolecular solvents 

 agree well with Van't Hoff's theory, associated compounds in 

 monomolecular solvents, monomolecular compounds in associated 

 solvents and associated compounds in associated solvents, differ 

 from it considerably; though in the last case the values equal 

 those required by the corrected theory when x/a equals x x . 

 Hitherto the deviations from Van't Hoff's theory of osmotic 

 pressure obtained by the cryoscopic method have been explained 

 by the hypothesis of electrolytic dissociation. But the above 

 results clearly show that this hypothesis is entirely unnecessary 

 for this purpose. No account has been taken of association in 

 the liquid state, either for the solvent or the dissolved substance, 

 and it is this which has necessitated the introduction by Van't 

 Hoff of the now celebrated coefficient i for aqueous solutions upon 

 which the hypothesis of electrolytic dissociation was founded. 

 The author concludes as follows: " (1) Van't Hoff's law of osmo- 

 tic pressure does not hold in its original form when either the 

 solvent or the dissolved compound are associated. (2) The latent 

 heat of fusion, melting point on the absolute scale, and density 

 at the melting point, of a liquid, are connected by the expression 

 rd/T= const, for monomolecular liquids, or rda/Tx= const, for 

 associated liquids, the value of the constant being 0*099 or roughly 

 0*1. (3) The molecular reduction of the freezing point for mono- 

 molecular compounds in monomolecular solvents is given by 

 Van't Hoff's formula E=0*01976 r F/r or by the derived formula 

 E = 0-2Tt/. If, however, the dissolved substance or the solvent are 

 associated, this formula no longer applies, but ~E = 0-019l6T 2 x/rax 1 

 or E = 0-2 Td/x x . (4) Exceptions to Van't Hoff's formula for the 

 molecular reduction of the freezing point, appear therefore when- 

 ever association of either dissolved substance or solvent takes 

 place, and the hypothesis of electrolytic dissociation is not only 

 unnecessary in explanation of these exceptions but inconsistent 

 with what is now known of the nature of liquids. (5) Electro- 

 lytes are salts in an approximately monomolecular fluid state, in 

 solution in associated solvents." — J. Chem. Soc, lxxi, 925-946, 

 August, 1897. g. f. b. 



2. On the Electrolytic Decomposition of Aqueous Solutions. — 

 A lecture has recently been given by Nernst before the German 

 Chemical Society in which he considers the chemical mechan- 

 ism of the electrolytic decomposition of aqueous solutions. He 

 showed in 1889 that the potential difference between a metal and 



