208 REPORT— 1886. 



attraction for water, but low rate of diffusion, might show less invapora- 

 tion than one with a weaker attraction for water but a higher rate of 

 diffusion. Thus, potassic chloride diffuses faster than sodic chloride, but 

 the latter has the greater attraction for water vapour. If solutions of 

 these two salts were confined in a space containing water, the sodic chlo- 

 ride solution would at first attract water more rapidly, but the consequent 

 dilution of the surface layer would not be counterbalanced by diffusion so 

 rapidly as in the case of potassic chloride ; so that the rates of invapora- 

 tion might become equal, or that of potassic chloride even greater. Some 

 experiments made by us have given indications of these phenomena. 



If the rates of invaporation, not complicated by diffusion, could be 

 accurately measured, a comparison of such measurements would be 

 valuable, by giving indications of the formation of hydrates in solution. 

 They would also be of value in considering the ' Correlation of Physical 

 Properties of Solution with Concentration,' in the manner indicated by 

 D. Mendeleeff (' Ber. Deut. Ch. Ges.' xix. 370-389). But the subject can 

 be investigated in a different way and with promise of more fruitful 

 results. When two salts are enclosed in the same space with a certain 

 quantity of water, the salts tend to keep the atmosphere dry by con- 

 densing the water- vapour. This goes on until all the water is evaporated 

 except that small portion which remains in the condition of vapour. The 

 question at once presents itself, in what proportion will the two salts 

 divide the water between them ? The proportion will be influenced, 

 probably, by the relative masses of the salts and the water, and by the 

 temperature, as well as by the relative attractions of the salts for water. 

 If the salts are in molecular proportion, they might be expected to divide 

 the water between them much in the same way as equivalents of caustic 

 soda and potash with a simple equivalent of sulphuric acid in solution. 

 That is, if the attraction of salts for the water which dissolves them is of 

 the same nature as that between acids and bases, the partition would be 

 in proportions representing the relative attractions of the salts for water. 

 It was to test the correctness of this reasoning that the following experi- 

 ments were made. 



The salts were carefully dried, and weighed out in small test-tubes 

 (5'5 cm. long and 1 cm. diam.). The quantities used were in the ratio of 

 the molecular weights. Thus, in the first experiment (I.) the masses of 

 the salts were two-hundredths of the gram-molecules, and the quantity of 

 water eight-hundredths of the gram-molecule. As a rule, the water was 

 divided between the two salts, because by so doing we thought to hasten 

 the completion of the experiment. Experience, however, has decided us to 

 abandon this method in favour of enclosing the water along with the salts, 

 the three in separate small tubes, so that the process of invaporation may 

 be watched from the beginning. The salts and water were sealed in a 

 large glass tube (about 10 cm. long and 4 cm. in diameter) before the 

 blowpipe. When the small tube containing the water appeared to be 

 dry, the enclosing tube was opened, the small tubes with their contents 

 weighed, and then resealed. Invaporation was very slow, owing to the 

 small surface exposed by the liquids in the narrow tubes. Shallow 

 vessels would have been better, but the hermetical enclosing of these pre- 

 sented such difficulties that the narrow tubes were used in preference. 

 Heating to 100° C. and cooling gradually was also tried as a means to 

 hasten invaporation. This was found to have the desired effect up to a 

 certain point, beyond which the water began to condense on the enclosing 



