122 M. E. Pringsheim on the Radiometer. 



atmospheric pressure take place in exactly the contrary way 

 to the motions of the radiometer, in order to maintain these 

 theories it would be necessary to assume that in rarefied air 

 the currents follow quite different laws from those governing 

 them under the ordinary pressure. At the most one might 

 infer these laws from the radiometer-motions, and so would 

 fall into a vicious circle. Moreover it is indubitable that 

 air-currents arise in radiometers ; but these, in the high 

 rarefactions, are very weak, and they are not the cause of the 

 motion, but usually counteract it. 



Of the impossibility of finding the cause of the motion in 

 air-currents one can easily be convinced by a very simple phe- 

 nomenon which I have had occasion to observe. A radio- 

 meter (or, more correctly, an otheoscope) possessed a movable 

 cross with four mica vanes inserted radially and inclined about 

 45° to the plane of the horizon, which were lampblacked on 

 their upper side. Immediately over this cross was a circular 

 disk of mica, easily rotated about a vertical axis, suspended 

 horizontal. Horizontally incident sunlight produced rotation 

 of the cross with the bright side preceding, and an opposite 

 rotation of the disk. 



This same motion continued when a portion of the cone of 

 light was cut off so that only the lampblacked sides of the 

 vanes were irradiated, while the bright sides were in the shade. 

 When, on the contrary, the lampblacked sides of the vanes 

 were shaded and the bright sides irradiated, the motion of the 

 cross was reversed, so that now the blackened sides preceded. 

 But the mica disk preserved its previous direction of rota- 

 tion; and consequently now the vanes and the disk rotated in 

 the same direction. Now it is certainly inconceivable that, 

 in consequence of the change of the illumination, the air- 

 current atone place should be reversed while that immediately 

 above it has the same direction as before. 



From a much more sure foundation, and one that has already 

 by manifold experiments been rendered almost a certainty, do 

 those theories start which rest on the kinetic theory of gases. 

 The thought which is common to all these theories is that the 

 " vacuum " with which w.ehave to do in radiometers is not 

 an empty space at all, but still contains an enormous number 

 of molecules of air. If, then, these molecules meet a surface of 

 higher temperature, apart of the heat contained in the surface 

 passes to the gas-molecules; these become warmer; that is, 

 according to the kinetic theory, they rebound with greater 

 velocity than they approached. According to the law of the 

 equality of action and reaction, the warmer surface must suffer 

 a greater repulsion from these particles retiring with accele- 



