140 



SCIENCE PROGRESS 



That this is so may further be seen by referring to the 

 densities of this series of solutions (Knietsch, Ber. 1901, 33, 

 4102). The formation of the monohydrate is marked by a 

 remarkable break in the density curve for 15 C. which falls to a 

 sharply defined minimum at 81*14 per cent. SO s (theory = 81-63 

 per cent.) ; this is shown on a large scale in the figure, from 

 which it will be seen that the change of direction is almost 

 as abrupt as that in the conductivity curve. Although, therefore 

 it may be true from the standpoint of strict theory that " no 

 homogeneous liquid mixture can show a discontinuous change 



Fig. I. — Diagram showing the properties of concentrated solutions of sulphur trioxide in 

 water. The densities are from data by Pickering and by Knietsch, the conductivities 

 are after F. and W. Kohlrausch, and the capillarities are those given by Knietsch ; the 

 viscosity data are too incomplete to be represented by a smooth curve and have been 

 omitted from the diagram. 



The concentrations shown are from 60 to 100 per cent. S0 3 ; the physical properties are 

 plotted on arbitrary scales. 



of density with composition whilst its temperature and pressure 

 remain constant " (H. Hartley, loc. cit. p. 542), there are evidently 

 instances in which a very high degree of experimental refine- 

 ment would be required to demonstrate the continuity of the 

 curve, whilst the " break " is so obvious as to render it extremely 

 easy to locate its position, and almost impossible to overlook the 

 existence of the hydrate by which is produced. 



