302 Radiation of Rough and Polished Surfaces, 



influence of the unpolishing is nothing, this is owing to their 

 emissive power reaching the limit of maximum, beyond which 

 an augmentation can scarcely occur, because the emissive sur- 

 face no longer supplies a hindrance to the escape of the heat ; 

 whilst in metals, far removed from this limit, the alteration of 

 the state of the surface must necessarily exercise all its influence, 

 and render it sensible by a marked variation in the quantity of 

 caloric emitted ? 



Although this reasoning is based upon a pure hypothesis, 

 viz. that lamp-black opposes no resistance to the radiation 

 from the surface ; and although the emissive powers of the 

 three substa\ices employed are indeed, on the one hand, so far 

 removed from the number 100, as to permit us to appreciate 

 the variations produced, and yet, on the other, are so energetic 

 that the smallest proportion of change occurring in their values, 

 would make them overcome the whole distance which separates 

 them from this number ; notwithstanding, let us for a moment 

 abandon all these non-metallic substances, and endeavour to 

 determine the question from the bodies themselves with which 

 we are now more especially concerned. 



Copper, zinc, tin, and white iron, which, so far as I know, 

 are the only metals which have hitherto been employed in the 

 experiment described at the commencement, on being exposed 

 to the action of the air, are quickly covered with a slight film of 

 invisible oxide, the presence of which is deduced, in a very satis- 

 . factory manner, from certain electrical phenomena. Now, it is 

 known that the emissive power is much stronger from oxides 

 than from the metals themselves. It may happen then, that 

 the streaked surface, presenting a greater number of points of 

 contact to the air, is more oxidated than the polished surface, 

 and thus augments its radiating power by the simple effect of 

 oxidation, without the more or less regular arrangement of the 

 superficial parts having any thing directly to do with the result. 



For ascertaining if this explanation could be maintained, all 

 that was required was to make the experiment with gold and 

 platina, and this I speedily did ; when I found that the streak- 

 ed plates of platina and gold always yielded a much more abun- 

 dant calorific emission than the polished surfaces of eitlier metal. 



Oxidation then, as well as the influence of polish in non-me- 



