through Dry and Humid Air. 53 



entrance of dry air produced in three successive experiments 

 the deflections 



76°, 74°, 75°. 



" The plate was again removed and the lining reduced to three 

 inches in length ; the deflections obtained in two successive ex- 

 periments were 



66°, 65°. 



"Finally, the lining was reduced to a ring only 1-| inch in 

 width ; the dynamic radiation from this small surface gave, in 

 three successive trials, the deflections 



60°, 56°, 56°-5. 



" The lining was then entirely removed ; and the deflection in- 

 stantly fell to 



7°-5. 



" In the foregoing experiments the lining was first heated by 



the collision of the air, and it then radiated its heat through a 



hick plate of rock-salt against the pile. The effect of the heat 



' ,/as enfeebled by distance, by reflexion from the surface of the 

 salt, and by partial absorption. Still we see the radiation thus 



'weakened competent to drive the needle almost through the 

 quadrant of a circle. Suppose, instead of being thus separated 

 from the lining, the face of the pile itself to form part of the inner 

 surface of the tube, receiving there the direct impact of the parti- 

 cles of air ; of course the deflections then obtained would be far 

 greater than the highest of those above recorded. I do not doubt 

 the possibility of causing the needle of my galvanometer, sub- 



1 jected to such an action, to swing through an arc of 1000 degrees ; 

 and it is my reluctance to derange the magnetism of my needle 

 that prevents me from making the experiment " *. 



Professor Magnus refers to the agreement which subsists 

 between his results and mine in the case of the more powerfully 

 acting gases, in proof of the correctness of his mode of experi- 

 ment. The agreement, however, is not such as to warrant 

 the conclusion drawn from it. The case may be illustrated by 

 reference to a delicate chemists' balance as compared with one 

 of those used in common life. Weighing pounds, both balances 

 would roughly agree, but in weighing milligrammes the coarser 

 balance would infallibly fail. 1 think it vain to expect a correct 

 determination in any case requiring great delicacy with the 

 apparatus which Prof. Magnus employs. 



Prof. Magnus again refers to the hygroscopic character of 



* When the pile was placed entirely within the tube (as Prof. Magnus 

 places it), a single stroke of the pump in exhausting drove the needle 

 through an arc of 115 degrees. 



