154 
MESSES. W. E. BOUSFIELD AND C. ELSPETH BOUSFIELD 
Following up the hint given by a consideration of viscosities the present investigation 
showsjthat good results can in fact be attained by treating the changes in both the 
specific heat and the specific volume of the free water as being simply proportional to 
the percentage of the solute present, i.e., 
s w —s F = P x constant, 
W—Wp = P x constant. 
P is proportional to l/(E + H). The constants were also calculated on the 
hypothesis that the changes were proportional to l/II instead of P. The results were 
not very different but not quite so concordant, and the hypothesis that certain 
changes in the free water of a solution which depend on diminution of molecular 
complexity are simply proportional to the percentage concentration of the solute 
appears to be a step in the right direction in the solution of the problem. 
