378 RADIANT HEAT. 



At the conclusion the author makes some observations in explana- 

 tion of the phenomena; these ma}- be more briefly and clearly expressed 

 thus: 



The unpolarized rays incident along the axis of the crystals examine*] 

 or parallel to it undergo a certain absorption dependent on the nature 

 of the crystal; and this is no further modified, since in this direction 

 there is no double refraction. 



But if the ray be incident perpendicular to the axis, it is divided 

 into two, oppositely polarized, and these are differently absorbed. 



That which has its plane of polarization parallel to the axis has the 

 same absorption as along the axis. That which has its plane of polari- 

 zation perpendicular- to the axis is absorbed more or less than the 

 former, in different degrees in different crystals, and for the different 

 component rays. 



When the unpolarized rays are incident, then the result is com- 

 pounded of these separate effects. 



When rays previously polarized are employed, the effects are 

 displayed singly. 



The author considers that these distinctions fully account for the 

 observed phenomena. 



Upon this we may observe — 



The investigation cannot with correctness be called one on the 

 transmission of "radiated heat" in general; it is restricted to that 

 peculiar form or case of radiation which is manifested in the solar rays, 

 and proves nothing as to the radiation from hot bodies, or even that 

 conveyed in the ra} T s from artificial lights, unless, as inferred, by 

 analogy. All the differences observed depend simply on the unequal 

 absorption of the rays of light by the crystals and the colored glasses. 



Mellon? s recent Experiments. 



In the " Comptes Rendus," No. 10, p. 429, March G, 1854, Melloni 

 gives some brief remarks in reply to certain objections raised by MM. 

 Provostaye and Desains against the accuracy of experiments with the 

 thermo-multiplier on the passage of heat through screens. 



He points out as the sources of discrepancy the oblique passage of 

 the rays through a thick diathermanous screen, which is greater or 

 less according to the distance, and gives different effects of internal 

 reflection and absorption in different cases. 



To show that the errors objected arise solely from this source, he 

 cribes a careful repetition of his experiments in which it was 

 'led against. 



The series of experiments included the usual set of sources, viz: 



(1.) The flame of an oil-lamp. 



(2.) Incandescent platina kept up by vapor of alcohol. 



(3.) Plate of copper heated by lamp. 



(4.) Vessel of hot water. 



Equalizing the effects on the thermo-multiplier by changing the dis- 

 tance and interposing a rock-salt screen, the diminution of effect 

 appeared the same for the first three sources, but greater for the fourth 



