Mr. M. M. Pattison Muir on Chemical Classijication. 269 



structure of these compounds. There is little doubt that the 

 periodic law, when it is more fullj developedj will be of 

 equal service. 



The arrantrement of all the known elements in a series 

 beginning with that element which has the smallest atomic 

 weight, and regularly progressing through those having 

 larger and larger atomic weights, and the connexion which is 

 thus shown to exist between the positions of the elements 

 and their general properties, lead us to believe that there 

 are several gaps in this series which are yet to be filled by the 

 discovery of new elements. Mendelejeff has even ventured 

 to speculate regarding the properties of some of those ele- 

 ments ; and it appears as if his predictions were to be realized 

 in one instance at least. In this Magazine, for April, 1877, 

 I have gathered together some of the facts which bear out 

 MendelejefF's idea that that metal which is the latest addition 

 to the list of elementary bodies, really occupies the position 

 assigned by him to a hypothetical element called eka- 

 aluminium. 



33. The "' periodic law " evidently points to a quantitative 

 connexion between the atomic weights of the elements and 

 the chemical properties of these bodies and of their compounds. 

 It is difficult to measure what we call " chemical properties ;" 

 and hence it is difficult to state the connexion which appa- 

 rently exists in any other than a vague and indefinite 

 manner. But the " periodic law" also enables us to connect 

 together atomic weight and physical characters ; and in each 

 of these we have a quantity susceptible of very accurate 

 measurement. 



If we trace back the history of chemical theory, we 

 find that chemists have ever been attempting to connect 

 together in a definite manner a variable and a variant. 

 In the early days of modern chemistry the variable was 

 the elements which combine and the proportions in which 

 they combine; the variant was empirical composition. A 

 little later it was attempted to connect the empirical com- 

 position, as variable, with general chemical behaviour, as 

 variant. I have endeavoured to show how^ chemists now 

 attempt to connect atomic structure as the variant with the 

 variable, composition — and also how they attempt to connect 

 the latter, which they almost regard as sjaionymous with 

 atomic structure, with physical properties as a variant. 

 MendelejefF's law presents us with a definite variable, viz. 

 atomic weight; and it attempts to represent the physical and 

 chemical properties, both of elements and compounds, as 



