108 M. E. Pringsheim on the Radiometer. 



or a secondary one produced by the heat radiated from the 

 heated glass side being more strongly absorbed by the front 

 than by the back surface of the mica. The latter, however, 

 is in the highest degree improbable ; for certainly a great 

 part of ihe heat conveyed into the glass by absorption is 

 communicated to the colder air in the interior of the apparatus; 

 and of that withdrawn from the glass by radiation, only an 

 extremely small part can pass over to the mica plate, since it 

 is radiated in the space uniformly in all directions. Hence 

 the proportionally strong action is scarcely to be attributed to 

 this minute quantity of heat, but, at any rate for the most 

 part, proceeds from the heat communicated to the gas by 

 conduction. 



The mica, too, would certainly absorb a portion of those 

 rays which come to it direct from the source of heat, which 

 must have the same action as those emitted from the glass ; 

 so that with the irradiation first employed, from the front, 

 the two actions would be added together, while with the 

 lateral irradiation only the action of the glass is present. 

 Therefore, the mica plate being at equal distance from the 

 same source of heat, the deflection ought to be greater in the 

 first case than in the second. But it is not so, as the follow- 

 ing observations show : — 



Distance of the bright gas-flame from the outermost per- 

 pendicular edge of the mica plate, 305 millim, 



Resting-position of the Spot of Light. 



I. With irradiation from the front : 



Without illumination. With illumination. Deflection. 



500 465 35 



II. With lateral irradiation : 



With illumination of With illumination of D fle f n 



the whole. one half. 



500 447 53 



The deflection in the second case is therefore not only not 

 weaker, but considerably stronger than in the first — which 

 probably results from the circumstance that with irradiation 

 from the front the side of the glass facing the back side of the 

 mica plate is heated by absorption, and thus counteracts the 

 action of the front half of the glass, while with the irradiation 

 lateral only the favourably acting half of the sphere is heated. 



Besides these probable reasons, a strict proof by experiment 

 presents itself that the repellent action of the heated glass 

 side is a primary and not a secondary action. If, namely, it 

 proceeded from the absorption by the vane of the rays emana- 

 ting from the glass, its amount would depend on the absorp- 

 tive capacity of the vane, and consequently on its substance. 



