103 



m precisely the same degree, of imparting motion to still ether. 

 Films of gas on surfaces of polished metal are shown to act like 

 coats of varnish. 



The author has appended a theoretic chapter to his memoir, in 

 which he investigates the physical connexion of radiation, ahsorption, 

 and conduction. In the foregoing experiments we have dealt with 

 free atoms and molecules, and thus fixed upon them individually 

 the responsibility of the effects observed. The effects are thus de- 

 tached from considerations of cohesion and aggregation which suggest 

 themselves in the case of liquids and solids. 



He points out that the reciprocity of absorption and radiation is a 

 necessary mechanical consequence of the theory of an ether. 



But why is one molecule competent to stop or generate a calorific 

 flux so much more powerfully than another? The experiments 

 point as follows : The elementary gases which have been examined 

 all exhibit extremely feeble powers both of absorption and radiation 

 in comparison with the compound ones. In the former case we 

 have oscillating atoms, in the latter oscillating systems of atoms. 

 Uniting the atomic theory with the conception of an ether, it follows 

 that the compound molecule which furnishes points tfappui to the 

 ether must be capable of accepting and generating motion in a far 

 greater degree than the single atom, which we may figure to our 

 minds as an oscillating sphere. Thus oxygen and hydrogen, which 

 taken separately or mixed mechanically, produce a scarcely sensible 

 effect, when united chemically to form oscillating systems, as in 

 aqueous vapour, produce a powerful effect. Thus also nitrogen 

 and hydrogen, which, when separate or mixed, produce but little 

 action, when combined to form ammonia, produce a great action. 

 So also nitrogen and oxygen, which, as air, are feeble absorbers 

 and radiators, when united to oscillating systems, as in nitrous 

 oxide, are very powerful in both capacities. Comparing small 

 volumes at equal tensions, the action of nitrous oxide is 250 

 times that of air ; a fact, which perhaps furnishes a stronger pre- 

 sumption than any previously existing, that air is a mixture, and 

 not a compound. Carbonic oxide is about 100 times as powerful as 

 its constituent oxygen ; carbonic acid 1 50 times as powerful ; while 

 olefiant gas, as already remarked, is 1000 times as powerful as its 

 constituent hydrogen. In the case of the hydrocarbon vapours, 



i2 



