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



metals are situated on the ascending portions of the curve, at 

 no great distance from the points at which the minima are 

 reached. An element whose atomic volume is greater than 

 that of the element with the next smaller atomic weight is 

 generally comparatively easily fusible, while an element 

 having a smaller atomic volume than that of the element next 

 below it in order of atomic weight, is fusible only v/ith diffi- 

 culty. The same relation holds between atomic volume and 

 volatility of the elements. The ductile metals, whose position 

 on the curve is at, or very near to, a maximum or minimum 

 point, crystallize, as a rule, in the regular system ; those ele- 

 ments which are more brittle and more volatile, and are 

 situated on ascending portions of the curve, crystallize for 

 the most part in systems other than the regular. 



The more volatile elements, which are placed on ascending 

 portions of the curve, generally possess a greater coefficient 

 of expansion by heat than those more fixed elements found at 

 the minimum points of the curve. What we know concerning 

 the refraction-indices of elementary bodies points to a periodic 

 connexion between these numbers and the atomic weights of 

 the same elements ; but our knowledge is as yet too limited 

 to allow of any thing more than a general statement of such a 

 connexion. The specific heats of the elements appear to bear 

 some relation, not only to the atomic weights, but also to the 

 atomic volumes of the elements. It would appear as if a cer- 

 tain magnitude must be attained by the atomic volumes of 

 those elements Avhich have small atomic weights before the 

 law of Dulong and Petit is obeyed. 



- That the conductivity for heat and for electricity of metals 

 varies periodically with their atomic weights is also apparent 

 from a study of these three quantities. 



The facts which I have now stated evidently point to a close 

 connexion between atomic weight and general physical proper- 

 ties of the elements. The atomic v/eight may be regarded as a 

 variable, and the physical constants of the elements as variants. 

 Physical phenomena, says Meyer, must probably be regarded as 

 functions, not only of space, of time, &c., but also as functions 

 of the atomic weights of the elements exhibiting these pheno- 

 mena, or of the elements constituting the compound bodies 

 which exhibit these phenomena. Such a consideration ofters 

 a wide field for future physical determinations ; it also de- 

 mands more exact atomic-weight determinations than have yet 

 been made for many of the elements. 



The elements may be roughly arranged in an electric series, 

 wherein some are more positive, some more negative than 



