Section III, 1922 [75] Trans. R.S.C. 



The Crookes Radiometer 



By John Satterly, F.R.S.C. 



* (Read May Meeting, 1922) 



In a series of six papers published in the Philosophical Trans- 

 actions for the years 1873-1879,^ Sir William Crookes described 

 the long series of experiments that he made on Repulsion arising 

 from Radiation. Partway through the work (in 1874) he invented 

 the Radiometer or Light Mill, a little apparatus consisting of a light 

 movable vane system delicately pivoted and mounted in a vacuum 

 tube. In the usual form of instrument there are four mica vanes 

 mounted on four arms at right angles to each other. The vanes 

 are blackened at one side and mounted vertically and radially at 

 the ends of the arms so that when the mill is set spinning the black- 

 ened surfaces are all advancing or all retreating. When exposed to 

 a source of radiation, e.g. the sun, or a lamp, or a hot ball, the mill 

 turns in such a way that the blackened surfaces are apparently 

 repelled by the radiation more than the untreated surfaces. In 

 1875 Crookes measured the rates of rotation for a particular instru- 

 ment placed at different distances from a constant source of radiation 

 (an oil lamp) and enunciated the law that the speed of rotation was 

 proportional to the intensity of the incident beam. He varied the 

 experiment by arranging a number of candles in a circle of 2ft. 

 diameter round a radiometer and starting with all the candles burn- 

 ing he blew them out one by one and measured the corresponding 

 speeds of rotation. He mentions that he did not make the experi- 

 ments with any great degree of preciseness but they taught him just 

 what he wanted to know, namely, that the instrument could be used 

 as a Radiometer. He suggested that the Light-mill might be used 

 as a photometer, also by photographers as an exposure meter. 



If we assume for a moment that Crookes' lamp acted as a point 

 source, the intensity of the radiation at a distance D would be inversely 

 proportional to D'^. Crookes plotted Times of Rotation T in seconds 

 as abscissae and Distances D as ordinates. The points lie on a curve 

 which fits fairly well to a parabola D~ = kT whence Crookes concluded 



iVol. 163, p. 295; Vol. 164, p. 501; Vol. 165, p. 519; Vol. 166, p. 325, Vol. 169, 

 p. 243; Vol. 170, p. 87, Vol. 172, p. 387. Also Phil. Mag. August, 1874. Proc. 

 Roy. Soc. XXV., p. 136 and p. 304. 



