Mri Mi M. Pattison Muir on Chemical Classijicati&n, 83 



the eleinents ; but the arrangement of the elements into groups 

 according to their valency does not interfere with the arrange- 

 ment of the same elements into other groups in accordance 

 with their specitic gravity, their electric and thermal conduc- 

 tivity, their malleability, their atomic weights, &c. &c. In fact, 

 the nearer we rise to a perfect system of classification (which, 

 however, we can never attain to) the more useful do the sub- 

 sidiary schemes of classification become. We begin even now 

 to see some connexion between the grouping of the elements 

 in accordance with their atomic weights, and the grouping in 

 accordance Avith general physical properties, or with valency. 



3. In an actual system of classification as practised in any 

 branch of science we find a more or less complete fulfilment 

 of the requisites of a perfect system. The qualities by which 

 each class is marked off from all other classes of objects some- 

 times admit of very accurate definition, sometimes they are 

 more vague. Each and every property regarded as character- 

 istic is not always existent in every member of the class. 



In cases where it becomes very difficult to define the essen- 

 tial characteristics of a class, recourse is sometimes had to 

 classification by types. An ideal type is erected, to which 

 the members of a given class more or less approximate. 

 This method of classification has been used to a consider- 

 able extent in the biological sciences. " The type of each 

 genus should be that species in which the characters of its 

 group are best exhibited and most evenly balanced "*. 



A similar method has been adopted in chemistry. Per- 

 plexed with the difficulty of giving such a definition of the 

 term '^ metal" as should mark off one group of elementary 

 bodies from that other group the members of which are com- 

 monly knovrn as metalloids, chemists have endeavoured to 

 imagine a substance possessing certain well-marked proper- 

 ties, some of which are always exhibited by those elements 

 which by common consent are designated as metals. Ac- 

 cording as any element in its general properties approaches 

 to or recedes from this ideal type, so is it to be classed among 

 the metals or among the non-metals. Such a scheme of clas- 

 sification is evidently most incomplete, and is altogether w^ant- 

 ing in those characteristics which are the marks of a perfect 

 System. In how far must the individual element approach 

 the type in order that it shall be admitted to the class of 

 metals? Must an element aspiring to be called a metal possess 

 three fourths of those properties which are characteristic of 

 the type, or will the half be considered sufficient ? A scheme 



* Waterhouse, (Quoted by Woodward in ' Rudimentary Treatise of Re- 

 cent and Fossil Shells,' p. (31. 



G2 



