76 DIVISIBILITY OF MATTEK ATOMIC THEORY. 



In treating thus far of the divisibility of matter, 

 and of the atomic theory as resultant from it, I have 

 said but little of those imponderable agents which we 

 can in no way dissociate from matter, and cannot 

 easily define otherwise than as functions of it, yet 

 which severally involve atomic actions and conditions 

 peculiar to each. Light and heat come perhaps in 

 closest relation, as connected with these conditions ; 

 and the discovery of a constant proportion between 

 the specific heat of bodies and the atomic weight of 

 their particles is one of the exponents of relations 

 which may well be called very profound, seeing how 

 far they go beyond the region of the senses or the bare 

 conjectures of thought. How important, again, in its 

 connexion with the atomic theory is the curious fact 

 (established as far as experiment has yet gone) that 

 the power or index of refraction of light in compound 

 bodies is the sum of the indexes of refraction of their 

 component parts. We may further fairly presume (and 

 this is confirmed by Tyndall's later researches) that 

 the undulations of light the ' luminous waves of 

 ether ' have definite relation to the atomic con- 

 ditions of the bodies they permeate. If simple light 

 can unite with sudden explosion the atoms of hy- 

 drogen and chlorine, we may well anticipate its action 

 in other cases in dissociating the molecules of com- 

 pound bodies, and with some relation, it may be, to 

 those differences in luminous waves on which the 

 differences of refrangibility depend. Of electricity, 

 under its various forms, as one of these great agents in 

 the world of nature, it may be said that atoms in 



