388 Professor Tyndall [June 7, 



currents ; the strength of the current being an accurate measure of 

 the quantity of the heat. As long as both faces of the pile were at 

 the same temperature, no current was produced ; but the slightest 

 difference in the temperature of the two faces at once declared itself 

 by the production of a current, which, when carried through the gal- 

 vanometer, indicated by the deflection of the needle both its strength 

 and its direction. 



The two faces of the pile were in the first instance brought to the 

 same temperature ; the equilibrium being shown by the needle of the 

 galvanometer standing at zero. The rays emitted by the current of 

 hot air already referred to were permitted to fall upon one of the 

 faces of the pile ; and an extremely slight movement of the needle 

 showed that the radiation from the hot air, though sensible, was 

 extremely feeble. Connected with the ring-burner was a holder con- 

 taining oxygen gas ; and by turning a cock, a stream of this gas was 

 permitted to issue from the burner, strike the copper ball, and ascend 

 in a heated column in front of the pile. The result was, that oxygen 

 showed itself, as a radiator of heat, to be quite as feeble as atmo- 

 spheric air. 



A second holder containing olefiant gas was also connected by 

 its own system of tubes with the ring-burner. Oxygen had already 

 flowed over the ball and cooled it in some degree. Hence, as a 

 radiator in comparison with oxygen, the olefiant gas laboured under a 

 disadvantage. It was purposely arranged tliat this should be the 

 case ; so that if, notwithstanding its being less hot, the olefiant gas 

 showed itself a better radiator, its claim to superiority in this respect 

 would be decisively proved. On permitting the gas to issue upwards, 

 it cast an amount of heat against the adjacent lace of the pile sufficient 

 to impel the needle of the galvanometer almost to its stops at 90°. 

 This experiment proved the vast difference between two equally 

 transparent gases with regard to their power of emitlmg radiant heat. 



The converse experiment was now performed. The thermo-electric 

 pile was removed and placed between two cubes filled with water kept in 

 a state of constant ebullition ; and it was so arranged that the quantities 

 of heat falling from the cubes on the opposite faces of the pile were 

 exactly equal, thus neutralizing each other. The needle of the 

 galvanometer being at zero, a sheet of oxygen gas was caused to issue 

 from a slit between one of the cubes and the adjacent face of the pile. 

 If this sheet of gas possessed any sensible power of intercepting the 

 thermal rays from the cube, one face of the pile being deprived of the 

 heat thus intercepted, a difference of temperature between its two 

 faces would instantly set in, and the result would be declared by the 

 galvanometer. The quantity absorbed by the oxygen under those 

 circumstances was too feeble to affect the galvanometer ; the gas, in 

 fact, proved sensibly transparent to the rays of heat. It had but a 

 feeble power of radiation : it had an equally feeble power of absorp- 

 tion. 



The pile remaining in its position, a sheet of olefiant gas was 



