92 Mr. M. M. Pattison Muir on Chemical Classification. 



idea might be extended to the elements themselves. The hy- 

 drogen in water was capable of replacement in halves, not only 

 by compound radicles but also by elementary substances. 

 Two derivatives of water, in which the hydrogen was partially 

 or entirely replaced by potassium, could be prepared. Other 

 elements, again, were capable of replacing at once the whole 

 of the hydrogen. But one derivative of water was known in 

 which the hydrogen was replaced by calcium ; but in this deri- 

 vative the whole of the hydrogen was replaced by the metal. 

 Hence arose the idea of varying capacities of saturation of the 

 elements — that is, of varying valencies. 



It would be out of place in such a paper as the present to 

 trace the course of the development of the idea of valency. By 

 the discovery of the glycols, of the alcoholic nature of glyce- 

 rin, of the relations existing between polybasic acids, polyacid 

 bases, and polyatomic alcohols, the idea gradually advanced to 

 maturity. Today valency is one of the fundamental ideas 

 upon which the scheme of chemical classification rests. 



11. I have thus endeavoured to show that the methods of 

 classification adopted during the earlier stages of the develop- 

 ment of chemical science may be classed as methods based 

 upon (1) the reactions, and (2) the composition of the sub- 

 stances classified. I have also attempted briefly to indicate 

 how the fusion of these two methods into a wider system, 

 which should pay regard alike to function and to composition 

 in attempting to frame an exhaustive classification of chemical 

 substances, was accomplished. 



In the first paragraph of the present paper I indicated the 

 leading features of a perfect system of classification — viz. the 

 iise of qualities capable of accurate definition as means for dis- 

 tinguishing between class and class, the need of employing 

 qualities some of which at least should be capable of ready re- 

 cognition, and the necessity for the existence of each and every 

 distinguishing property in each individual of a class. 



Tried by these tests, each system of classification which we 

 have briefly noticed must be pronounced wanting. The pre- 

 decessors of Lavoisier were generally totally unable to define 

 the characteristic qualities of a class. Lavoisier himself gave 

 a sure and satisfactory definition of elementary and compound 

 bodies ; but he made much of the composition of salts ; and to 

 the question what is a salt? the chemist can hardly yet give a 

 satisfactory answer. Berzelius viewed all compounds as dual 

 bodies ; but how are we to define the characteristics of that 

 vast array of radicles by the combination of which all com- 

 pounds were supposed to be produced? What definite know- 

 ledge did the school of dualism afford as to the correlation of 



