38 



The law of inverse squares has been proved by Melloni to be 

 true for radiant heat passing through air, whence that eminent 

 experimenter inferred that the absorption of such heat by the atmo- 

 sphere, in a distance of 18 or 20 feet, is totally inappreciable. 

 With regard to the action of other gases upon heat, we are not, so 

 far as I am aware, possessed of a single experiment. 



Wishing to add to our knowledge in this important particular, I 

 had a tube constructed, 4 feet long and 3 inches in diameter, and 

 by means of brass terminations and suitable washers, I closed per- 

 fectly the ends of the tube by polished plates of rock-salt. Near to 

 one of its extremities, a T-piece is attached to the tube, one of 

 whose branches can be screwed to the plate of an air-pump, so as to 

 permit the tube to be exhausted ; while the gas to be operated on is 

 admitted through the other branch of the T-piece. Such a tube 

 can be made the channel of calorific rays of every quality, as the 

 rock-salt transmits all such rays with the same facility. 



I first permitted the obscure heat emanating from a source placed 

 at one end of the tube, to pass through the latter, and fall upon a 

 thermo-electric pile placed at its other end. The tube contained 

 ordinary air. When the needle of a galvanometer connected with 

 the pile had come to rest, the tube was exhausted, but no change in 

 the position of the needle was observed. A similar negative result 

 was obtained when hydrogen gas and a vacuum were compared. 



Here I saw, however, that when a copious radiation was employed, 

 and the needle pointed to the high degrees of the galvanometer, to 

 cause it to move through a sensible space, a comparatively large 

 diminution of the current would be necessary ; far larger, indeed, than 

 the absorption of the air, if any, could produce : while if I used a 

 feeble source, and permitted the needle to point to the lower degrees 

 of the galvanometer, the total quantity of heat in action was so 

 small, that the fraction of it absorbed, if any, might well be insensible. 



My object then was to use powerful currents, and still keep the 

 needle in a sensitive position ; this was effected in the following 

 manner : The galvanometer made use of possessed two wires coiled 

 side by side round the needle ; and the two extremities of each wire 

 were connected with a separate thermo-electric pile, in such a manner 

 that the currents excited by heat falling upon the faces of the two piles 

 passed in opposite directions round the galvanometer. A source of 



