M. E. Pringsbeim on the Radiometer. 105 



occurs between tbe glass case and tbe vanes, it migbt easily 

 be believed that tbe temperature of tbe former lies between 

 tbe temperatures of tbe brigbt ami tbe dark surface, and 

 tberefore exerts an opposite influence on botb. But tbe 

 glass case can be kept at tbe temperature originally possessed 

 by tbe entire apparatus by inserting in tbe patb of tbe rays 

 a layer of plane glass plates of sucb tbickness tbat all tbe rays 

 generally absorbable by glass are kept back in tbem; and in this 

 way it is easy to ascertain tbat even tben a sufficiently power- 

 ful source of ligbt produces tbe same effect, altbougb in a 

 less degree, as without tbe interposition of tbe glass plates. 

 Further, that all tbe rays absorbable by glass are actually 

 absorbed by the glass plates before reaching the apparatus, 

 can with facility be known from this — that the insertion of 

 additional glass plates between the source of light and the 

 instrument has no effect upon the motion. This proves, 

 therefore, that the above assumption does not correspond to 

 the real cause of the opposite action upon the brigbt and the 

 dark side of the vanes, and consequently furnishes a con- 

 firmation of the view enunciated by Cooke *, that the rays 

 which have passed through the mica, being absorbed by the 

 soot, produce sufficient heating to cause the motion. 



Now, although it is not possible, as above remarked, to 

 decide by experiment whether the glass case is absolutely 

 necessary for tbe production of the motion, yet we can get 

 some elucidation of its actual operation. For as all the rays 

 incident upon the vanes must first pass through the glass, it 

 is evident that it must absorb a portion of the radiation and 

 thereby have its temperature raised. Probably it is now 

 possible to ascertain what influence this rise of temperature 

 of the glass sides alone would have if the vanes themselves 

 retained their initial temperature during the entire experi- 

 ment, or at least showed no difference of temperature on their 

 two sides. 



To effect this, in the above described torsion-apparatus a 

 mica plate as thin as possible was employed as a vane, from 

 which it was to be presumed that it would be capable of ab- 

 sorbing very few rays indeed, and that, with its extraordinary 

 thinness, both sides would always have the same temperature. 

 When now our apparatus, suspended in the open air without 

 its tin case, was irradiated, the side of the glass case which 

 was turned to the source of light must have absorbed more 

 rays than that which was turned from it, while it could be 

 assumed that the mica plate possessed the same temperature 



* Silliman's Journal, xiv. p. 237 (1877). 

 Phil. Mag. S. 5. Vol. 15. No. 92. Feb. 1883. I 



