112 M. E. Pringsheim on the Radiometer. 



that in gas so highly rarefied the absorption is vanishingly 

 little. In other respects the apparatus showed the same phe- 

 nomena with illuminating-gas as when atmospheric air was 

 employed. 



III. Action of the Vanes. 



By far the greatest influence on the manner of motion of 

 radiometric apparatus is exerted by the quality of the vanes 

 themselves. First of all, it is evident that motion can only be 

 produced by the vanes when on irradiation their two sides act 

 differently ; for one and the same action of both sides would 

 neutralize itself. This different action presupposes a difference 

 in the quality of the two sides, which difference may lie in 

 their consisting of different substances, or having different 

 forms, or being of the same substance but having different 

 temperatures. 



As to the first point, the great complexity of the cooperating 

 properties of the bodies renders it hardly possible to state 

 exactly the influence exerted by each property of the sub- 

 stances used as vanes. 



In the first place, the forces which generally come into play 

 are very slight; hence it will be necessary to employ extremely 

 light substances — such as pith, mica, or very thin metal plates. 



Further, as the motion is produced by irradiation, it is clear 

 that the vis viva of the motion is derived from the rays. But 

 this is only possible by absorption, since even the simplest 

 experiments show that the cause of the motion cannot be a 

 direct transference of the vis viva of the rays to the vanes. 

 Hence the greater the amount of rays absorbed in the vane, 

 and the more the portion of the radiation absorbed by one 

 surface exceeds that absorbed by the other, the stronger 

 will the action be. To this inference correspond also the 

 observations* which show that the action is strongest when 

 thin mica plates, which absorb almost no rays at all, are 

 coated on one side with lampblack, the blackest of all known 

 substances also for heat. Less favourable are metal plates 

 blackened on one side with lampblack, since metal radiometers 

 on being radiated upon by dark sources of heat move with the 

 lampblacked side in front, therefore inversely as in lightf. 

 From this it is to be inferred that the metals are still more 

 impenetrable to heat-rays of very great wave-length than soot 

 — which also agrees with the experience that soot in thin 



* Crookes, Phil. Trans, clxix. (1878) pp. 259, 260. 

 t Crookes, Comptes Mendus, lxxxiii. p. 1289 ; Proc. Royal Soc. xxv 

 p. 812 ; 4 Nature,' xv. p. 226 (1877). 



