50 



SCIENCE 



[N. S. Vol. XXXI. No. 785 



like any other solution, except that it has 

 the property of conducting electricity ivith 

 concomitant chemical decomposition. There 

 is no way known at present by which any 

 one can foretell whether a given solution 

 will conduct the current or not. The only 

 way to find out is by actual trial with the 

 electric current itself. There is also a mis- 

 apprehension that only electrolytes will 

 cause the coagulation of colloids. Such 

 coagulation can be quite as well accom- 

 plished by non-electrolytes, so that here too 

 there is no essential difference between 

 electrolytes and non-electrolytes. Upon 

 what electrolytic conduction really depends 

 we are still quite ignorant, just as we do 

 not know why a bar of silver conducts and 

 a stick of sulphur insulates. But upon this 

 matter I have already expressed myself 

 more fully on other occasions. 



Again it is necessary to call attention to 

 the fact that there is really no essential 

 difference between colloidal solutions and 

 solutions of crystalline substances. I do 

 not refer to those so-called colloidal solu- 

 tions which from the very mode of their 

 preparation must be regarded as suspen- 

 sions, which view has also been confirmed 

 by the use of the ultramicroscope. We are 

 now able to separate crystalline bodies 

 from each other by dialysis, also crystal- 

 line bodies from those that have never been 

 oitained in the crystalline state by having 

 the latter pass through the septum and the 

 crystalloids remain behind; and indeed, 

 even tivo colloids may be separated from 

 each other by dialysis, as I have demon- 

 strated experimentally in the course of my 

 researches on osmosis. The matter depends 

 entirely upon the nature of the solutions 

 and the chemical nature of the septum, and 

 from a knowledge of these, what will 

 happen may be foretold. 



Water is a great solvent, and because of 

 its abundance and importance to all life 

 on the globe aqueous solutions will ever be 



studied with the greatest interest. But in 

 obtaining a correct conception of the na- 

 ture of solutions, aqueous solutions obvi- 

 ously can have no stronger vote than solu- 

 tions in less abundant and far less readily 

 procurable liquids. Water has a high co- 

 hesion, as is shown by its high surface 

 tension and high latent heat of vaporiza- 

 tion. The hydroxyl group which is char- 

 acteristic of the water molecule certainly 

 exhibits great tendency to cling to other 

 hydroxyl groups. So, for instance, though 

 hydrocarbons are not soluble in water, 

 they become soluble when one of their 

 hydrogen atoms is replaced by hydroxyl, 

 provided that the number of carbon atoms 

 in the compound is small. However, when 

 more than one hydroxyl group is in an 

 organic compound, the latter may have 

 even a relatively high carbon content and 

 yet be soluble in water. A study of or- 

 ganic hydroxyl derivatives shows that 

 compounds consisting of carbon, hydrogen 

 and oxygen, and containing one or more 

 hydroxyl groups for every carbon atom 

 present, are soluble in water, though, to be 

 sure, even considerably less than one 

 hydroxyl group per each atom of carbon 

 in the molecule is frequently sufficient to 

 cause solubility. On the other hand, the 

 multiplication of hydroxyl groups in such 

 compounds tends to diminish their solu- 

 bility in hydrocarbons. From this and 

 similar illustrations that might readily be 

 given it is clear that a study of the solu- 

 bility of a compound in different solvents 

 may well serve as a means to investigate 

 the nature of that compound. 



It need not be feared that by accepting 

 the chemical view of solutions we should 

 lose the advantage of the molecular weight 

 determinations by the boiling-point and 

 freezing-point methods. These methods 

 would serve us as well as ever. But we 

 should not argue that common salt is dis- 



