i 3 6 SCIENCE PROGRESS 



to the practical exclusion of the hydrates S0 3 . 150H2O and 

 S0 3 . I52H 2 0. The tendency at the present time is rather to 

 regard the sulphur trioxide or sulphuric acid molecules (or ions) 

 in a dilute solution as nuclei round which water molecules are 

 condensed. The size of the " watery atmosphere " (F. Kohlrausch, 

 Proc. Roy. Soc. 1903, 71, 338) may vary with the concentration, 

 pressure, and temperature of the solution (Bousfield, Phil. 

 Trans. 1906, 206, 101), and similar differences may exist between 

 individual molecules in the same solution ; although therefore 

 at any one moment a molecule of acid might be associated with 

 a definite number of water molecules, the number would vary 

 with each collision, and the conception of a series of hydrates 

 present in definite proportions under definite conditions could 

 have only a statistical meaning. 



The introduction into the hydrate theory of the idea of 

 reversibility had the result of destroying the theoretical justifi- 

 cation for Mendeleef s breaks — a justification that was of excep- 

 tional importance in view of the doubt attaching to their actual 

 existence. The abrupt changes of direction which he expected 

 to find in his experimental curves are of frequent occurrence 

 when saturated solutions are studied, although this involves 

 working at the temperature of crystallisation instead of a 

 constant temperature of 20 or 50°. Thus when a substance 

 such as magnesium sulphate crystallises from its solution with 

 different proportions of water the break in the solubility- 

 temperature curve which marks the appearance of a new hydrate 

 is accompanied by a corresponding break in the density-tempera- 

 ture curve, although the isothermal density-concentration curves 

 may be perfectly regular. But in the case of solutions which 

 are not saturated similar breaks can only occur when compounds 

 are formed which do not dissociate when fused, a condition that 

 is altogether exceptional in the case of aqueous solutions, and 

 would be even rarer in the case of less active solvents. 



But although dissociation is fatal to the appearance of abrupt 

 breaks in the properties of a series of solutions, it must be 

 recognised that very rapid alterations of direction and curvature 

 may occur when hydrates of great stability are formed. This 

 appears to have been the view entertained by Pickering, who 

 admitted that the hydrates of sulphuric acid might dissociate in 

 solution of their own accord without resorting to the extreme 



