Refraction of the Elements and t/teir Chemical Equivalents. 145 



Secondly, that the constants of the bivalent, trivalent, quadrivalent, 

 and apparently quinquivalent groups are practically the same, rang- 

 ing about I'Ol. 



Thirdly, that when a metal combines in a proportion that indicates 

 a lower valency than that ordinarily assigned to it, its constant is 

 tewhat elevated. 



I refrain at present from pointing out minor analogies between 

 closely-allied metals, and from attempting to explain the difference 

 between the univalent and the other groups ; why sodium should fall 

 away from the value proper to the alkaline group, and closely 

 approximate to that of all the other groups; or why beryllium, 

 bivalent tin, and trivalent iron should be represented by such ex- 

 ceptionally high figures. 



It is to be understood that the values given are all deduced from 

 compounds in which the metal plays the part of an electro-positive 

 radicle. Where they combine with oxygen to form the electro- 

 negative radicle, the values are completely altered, just as we find in 

 the case of several non-metallic elements. 



If we calculate these constants for the square root of the atomic 

 weight instead of that of the combining proportion, we shall obtain 

 for the means 



Univalents T30 



Bivalents 1'40 



Trivalents 1'75 



Quadrivalents 2'12 



Quinquivalent 2*19 



This arrangement does not, as in the former case, give a practically 

 identical constant for the bivalent, trivalent, quadrivalent, and quin- 

 quivalent metals. The fact that these numbers increase nearly in 

 the proportion of the square roots of 2, 3, 4, and 5, indicates that the 

 relation involved is not between the specific refraction and the atom, 

 but rather between it and the combining proportion or chemical 

 [uivalent of the metal. This brings the optical property into 

 analogy with Faraday's law of electro-chemical equivalents. 



I propose to give this product the descriptive name, " Refractive 

 mstant of equivalent weights." It may be represented by 



SE* = constant, 



where S is the specific refraction, and E the chemical equivalent of 

 le metal. 



Some physicists may prefer to make use of the square of the above 

 formula, namely, 



S 2 E = constant. 



