through Dry and Humid Air, 47 



appeared to me best calculated to eliminate all sources of error. 

 My mode of operating was as follows : — The brass tube (A B, 

 fig. 2 [see ante, p. 34]), open at both ends, formed the conduit 

 for a portion of the thermal radiation from the source (S). 

 These heat-rays, after passing through the tube, traversed 

 several inches of intervening air-space before entering the cone 

 of the thermo-electric pile, where they produced their effect, in 

 opposition to that arising from another constant source of heat 

 affecting the opposite face of the pile (the compensating cube). 

 The differential action was indicated as usual by a delicate galva- 

 nometer. These arrangements being once for all made, I was 

 able by means of an air-pump to introduce at pleasure into the 

 tube either the ordinary air of the laboratory, dry air, or air ren- 

 dered moist by passage over extensive surfaces of wet glass. At 

 the commencement of the experiments the tube was of course full 

 of the common air of the laboratory ; the needle of the galva- 

 nometer marked 42°, and remained steady for a quarter of an 

 hour within a degree of that point. I now interposed in the 

 path of the rays entering the brass tube a sheet of tin-plate ; the 

 needle at once bounded from 42° up to 90°. It was thus evident 

 that any obstruction to the passage of the rays of heat through 

 the tube, or, in other words, any cooling of that face of the pile 

 which was turned towards the tube, would be indicated by an in- 

 creased deflection of the needle on the same side of zero, which I 

 will call the — side, whilst a heating of the same face of the 

 pile would be attended by a diminished deflection, or even by 

 a passage of the needle to the opposite or + side of zero. The 

 following are the results which I observed : — 



Permanent 

 Tube filled with deflection. 



Common air —42 



Air dried by contact with monohydrated sulphuric"! , „ 



acid and introduced gently into the open tube. J 

 Common air which had spontaneously displaced! __.„ 



the foregoing dry air J 



Common air gently drawn in by air-pump . . — 43 3 

 Common air gently blown in from caoutchouc bag. —45 

 Same air gently blown in from caoutchouc bag,l fi 



but dried by passage over sulphuric acid . J 



Air from same bag but not dried —46*5 



Air of laboratory . —42 



Air dried and introduced as before . . . . . -f 14 

 Air dried as before with sulphuric acid, but after-"! 



wards passed over fragments of glass moistened > — 46 



with water J 



Common air gently drawn in by air-pump . . — 42*5 



