680 [June 18, 



the competence of the apparatus to measure an absorption of this 

 magnitude, it is only necessary to state that the galvanometric de- 

 flection corresponding to this absorption was 1 1 degrees. Were it 

 worth while, it might be shown that the absorption by a plate of the 

 gas not more than K^OO^ of an inch in thickness is capable of 

 measurement. A layer of olefiant gas 2 inches in thickness inter- 

 cepted nearly 30 per cent, of the entire radiation. Such a layer, 

 encompassing the earth as a shell, permitting the passage of the 

 solar rays, and preventing the escape of the terrestrial ones, would 

 probably raise the surface of the earth to a stifling temperature. A 

 layer of the gas three-tenths of an inch in thickness intercepts 1 1'5 

 per cent, of the radiation. Such a layer, if diffused through 10 feet 

 of air, would be far more attenuated than the aqueous vapour actually 

 existing in the air ; still it would effect an absorption greater than 

 that which the author has ascribed to the atmospheric vapour 

 within 10 feet of the earth's surface. In the presence of such facts, 

 arguments drawn from the smallness of quantity of the atmospheric 

 vapours are entirely devoid of weight. 



For larger thicknesses of gas and vapour, a tube was employed, 

 which was divided into parts capable of being used separately or 

 together. The mode of proceeding was this : A brass cylinder 

 49*4 inches long had its two ends stopped with rock-salt ; a source 

 of heat was attached to it exactly as in the author's experiments 

 described in former memoirs. The pile and the compensating cube 

 also occupied their old positions ; but instead of determining the 

 absorption effected in a column of gas or vapour equal in length to 

 the whole tube, the tube was now divided into two independent com- 

 partments by a third plate of rock-salt. Let us call the compart- 

 ment furthest from the pile the first chamber, and that nearest to 

 the pile the second chamber. The experiments were commenced with 

 the first chamber very short, and the second chamber long ; and the 

 plate of salt was subsequently shifted so as to lengthen the first 

 chamber and shorten the second one, the sum of the lengths of both 

 chambers being preserved constant at 49*4 inches. 



The absorption effected in the first chamber acting alone was first 

 determined ; then the absorption effected by the second chamber acting 

 alone ; and finally, the absorption effected when both chambers 

 were occupied by the gas or vapour was determined. This arrange- 



