634 SCIENCE PROGRESS 



separately, but not to a mixture of the two substances. The 

 experiments were supposed to furnish conclusive evidence of 

 the existence of some molecules (hydrates) in solution, which 

 were larger than those of either substance, and were conse- 

 quently unable to pass so readily through the meshes of the 

 porous diaphragm. Recent investigation, however, by Barlow, 

 and by Findlay and Short, has shown that the experimental 

 basis of this conclusion is unsound. 



Evidence is accumulating in favour of the view that a semi- 

 permeable membrane dissolves or imbibes only the substance 

 to which it is permeable ; such a membrane, therefore, is 

 efficient in proportion as it selectively absorbs one of the two 

 substances in contact with it. But the fact must be emphasised 

 that the result of applying thermodynamics to solutions is 

 independent of the way in which the semi-permeable membrane 

 acts. 



