554 On certain Movements of Radiometers. [Dec. 20, 



that the order is different for different components of the total radia- 

 tion, and that the ratio of the intensity of those components has been 

 changed. 



It is perhaps hardly necessary to observe that the radiometers men- 

 tioned in this paper are of the usual form — that is to say, that their arms 

 are symmetrical, so far as figure is concerned, with respect to a vertical 

 plane passing through the point of support. Accordingly the rotation 

 which is attained, for instance, with a radiometer with concave disks of 

 aluminium, alike as to material on both faces (of which kind, again, I owe 

 a beautif ul specimen to Mr. Crookes's kindness), has not been referred to. 

 This rotation, depending on the more favourable presentation, to the bulb, 

 of the outer (and therefore nearer and more efficient) portions of the fly 

 on the convex than on the concave side, has nothing to do with the one 

 isolated subject to which the present paper relates, namely, the elucida- 

 tion of the peculiar behaviour in certain cases of certain kinds of radio- 

 meters, by a consideration of the heterogeneous character of the total 

 heat-radiation. 



November 20, 1877. 



2nd P.S. — This morning I received from Mr. Crookes an account of the 

 behaviour of a kind of radiometer which he was so good as to construct at 

 my suggestion. The consideration of an experiment mentioned in a paper 

 of his presented to the E-oyal Society, which will shortly be read, and which 

 he has kindly permitted me to refer to, suggested to me the desirability 

 of investigating the effect of mere roughness of surface, all other circum- 

 stances being alike, and the disk of the radiometer being metallic, so that 

 the two faces may be regarded as practically at the same temperature. 

 Mr. Crookes's experiment, above referred to, led me to suspect that mere 

 roughness might increase the efficiency of a surface ; and I suggested to 

 him some experiments with heated glass shades, or with a hot poker pre- 

 sented to the radiometer, the bulb being covered with a cool tumbler to 

 defend it from being heated by the rays easily absorbed by glass. The 

 result in every case answered my expectation; and it may be stated 

 shortly that the law of the motion is that when the fly is hotter than the 

 bulb the rough surface is repelled, or, say, the motion is positive ; when 

 cooler, negative. 



I subjoin Mr. Crookes's memorandum of the results of experiment : — 



" Aluminium Eadiometer (1326), one side of the vanes being ruled 

 closely with a sharp knife. 



" 1. Exposed to standard candle 3 inches off. Continuous positive 

 rotation (ruled side repelled) at rate of 3| revolutions a minute. 



"2. Exposed to non-luminous flame of a Bunsen burner 3 inches off. 

 Continuous positive rotation at the rate of 7^ turns a minute. 



