RADIATION OR HEAT BY GASEOUS AND LIQUID MATTER. 
203 
V, through which cold water continually circulates, entering at the bottom and escaping 
at the top. The heat is thus prevented from passing by conduction from the source C 
to the first plate of rock-salt S. This plate forms the end of the hollow cylinder A B, 
dividing it from the front chamber F, with which the cylinder A B is connected by 
suitable screws and washers. Within the cylinder AB moves a second one, II, as an 
air-tight piston, and the bottom of the second cylinder is stopped by the plate of rock- 
salt S'. This plate projects a little beyond the end of its cylinder, and thus can be 
brought into flat contact with the plate S. Fixed firmly to A B is a graduated strip of 
brass, while fixed to the piston is a second strip, the two strips forming a vernier, v v. 
By means of the pinion B, which works in a rack, the two plates of salt may be separated, 
their exact distance apart being given by the vernier. P is the thermo-electric pile with 
its two conical reflectors ; C' is the compensating cube, employed to neutralize the radia- 
tion from the source C. H is an adjusting screen, by the motion of which the neu- 
tralization may be rendered perfect, and the needle brought to zero under the influence 
of the two opposing radiations. The graduation of the vernier was so arranged as to 
permit of the employment of plates of gas varying from 0 - 01 to 2 - 8 inches in thickness. 
They were afterwards continued with the pieces of the experimental tube, already re- 
ferred to, and in this way layers of gas were examined which varied in thickness in the 
ratio of 1 : 4900. 
In my former experiments the chamber F was always kept exhausted, so that the rays 
of heat passed immediately from the source through a vacuum ; but in the present 
instance I feared the strain upon the plate S, and I also feared the possible intrusion of a 
small quantity of the gas under examination into the front chamber F, if the latter were 
kept exhausted. Having established the fact that a length of 8 inches of dry air exerts 
no sensible action on the rays of heat, I had no scruple in filling the chamber F with dry 
air. Its absorption was nil, and it merely had the effect of lowering in an infinitesimal 
degree the temperature of the source. The two stopcocks c and c' stand exactly opposite 
the junction of the two plates of salt S, S' when they are in contact, and when they are 
drawn apart these cocks are in communication with the space between the plates. 
After many trials, the following mode of experiment was adopted : — The gas-holder 
containing the gas to be examined was connected by an india-rubber tube with the 
cock c', the other cock c being at the same time left open. The piston was then moved 
by the screw R until the requisite distance between the plates was obtained. This 
space being filled with dry air, the radiations on the two faces of the pile were equalized, 
and the needle brought to zero. The gas-holder was now opened, and by gentle pressure 
the gas from the holder was forced first through a drying apparatus, and then into the 
space between the plates of salt. The air was quickly displaced, and a plate of the gas 
substituted for it. If the layer of gas possessed any sensible absorbing power, the equi- 
librium of the two sources of heat would be destroyed ; the source C' would triumph, and 
from the deflection due to its preponderance the exact amount of heat intercepted by 
the gas could be calculated. 
