RADIANT HEAT. 337 



in the different eases. This very singular result is established by 

 numerous detailed experiments, of which a tabular statement is given; 

 and the author states it in the following terms: "The calorific rays 

 issuing from the diajrfianous screens are, there/ore, of different qualities, 

 and possess, if we may use the wrm, the diathermancy peculiar to each of 

 the substances through which they have passed. 7 ' 



He next investigates the effects of different colors in glass on the 

 absorption of heat. He infers, in general, that the coloring matter 

 diminishes the power of transmission, and examines the question, 

 Does it stop only rays of a definite refrangibility analogous to what 

 happens in the absorption of light ? 



With this view (following a similar mode of operation to that 

 adopted in the last instance) he used, successively, glasses of different 

 colors, for each of which the distance of the source was varied till a 

 standard effect (about 40° deviation of the needle) was produced on 

 the galvanometer. In this position, in each case, a plate of sulphate 

 of lime was then interposed, and diminished the deviation to about 

 18° for all the colored glasses, except green, in which case it was to 

 about 8°. When alum was substituted the deviations were reduced, 

 in the first case, to 8°, in the second to 1°.6. Hence he concludes 

 that all the colored glasses, except green, produce no "elective 

 action" on heat; green glass, on the contrary, transmits rays more 

 easily stopped than the others. 



Connecting this with his other inference, that rays are stopped in 

 proportion to their refrangibility, he instituted another series of ex- 

 periments to put this to the test. The sources of heat compared 

 were an argand lamp and incandescent platinum, the rays of heat 

 from the former being the more refrangible. The quantities of heat 

 from the lamp and the metal transmitted by the green glass were 

 nearly equal; by all the others, nearly in the ratio of 2 to 1. Hence 

 he infers that green glass is more diathermanous for rays of less 

 refrangibility. 



Again, the rays transmitted by 'citric acid and some other sub- 

 stances are those only of the greatest refrangibility. They should, 

 therefore, be the least transmissible by green glass. This was found 

 to be the case. Of 100 rays passed through citric acid, all the other 

 glasses transmitted various preparations, from 89 to 28, while green 

 glass transmitted only from 6 to 2. 



Without the citric acid the rays from incandescent platinum were 

 more copiously transmitted by the green glass than by the others. 



The whole of the rays of low refrangibility emitted by the platinum, 

 and for which alone the green glass is transparent, had been stopped 

 by the interposition of the plate of citric acid, which had, as it were, 

 sifted it free from these rays. 



Hence, the author concludes that "green glass is the only kind which 

 possesses a coloration/o?* heat, [if ice may use the expression,) the others 

 acting upon it only as more or less transparent glass of uniform tint 

 does upon light." 



In a subsequent part of the memoir, M. Melloni gives a tabular 

 view of the effects, observed in the same manner, of the constant 

 22 



