M. E. Pringsheim on the Radiometer. 119 



predict with certainty the course of each phenomenon. For, 

 in the first place, in most of the experiments such different 

 influences operated that even with exact knowledge of the 

 laws that govern each one of them it would be difficult to cal- 

 culate the result of their cooperation; and, besides, we are 

 still far from being able to state numerically the properties in 

 question of the different parts of the apparatus (such as ab- 

 sorptive capacity for the precise kind of radiation employed, 

 emissive power, internal and external conductivity, specific 

 heat, &c); and much less still do we know of the laws by 

 which the action in the apparatus itself of those various pro- 

 perties is regulated. 



Nevertheless, in order to embrace in the simplest possible 

 propositions the laws of radiometer-motion, we will first re- 

 present to ourselves the experimental conditions on which the 

 occurrence of the motion is dependent. The first condition, 

 which makes the apparatus (radiometer, otheoscope, or tor- 

 sion-apparatus) what we call " radiometric," is that the gas 

 enclosed in the instrument be rarefied beyond the neutral point. 

 Then we get the following general proposition : — 



Forces act upon the movable parts of a radiometric apparatus 

 as soon as the equilibrium of temperature within it is disturbed. 



A more particular statement, how and in what direction 

 these forces act, upon what their magnitude depends, &c, 

 cannot at present be made; but the experiments must be em- 

 pirically combined into larger classes; and thus the following 

 rules can be given: — 



(1) A plane radiometer-vane tends, when irradiated, to 

 recede with its warmer side. 



(2) A curved vane tends, when warmed, to recede with its 

 convex side. 



(3) A vane suspended over against a warmed surface tends 

 to recede from it. 



By these propositions, however, little is gained, unless we 

 can bring them under a common point of view that will make 

 the process intelligible to us and show us on what it really 

 depends. 



We shall arrive at such a point of view if we consider what 

 process in radiometric phenomena is precisely that with which 

 the force comes in. The motion arises through the irradia- 

 tion of the apparatus; the force of the motion must therefore 

 be derived from the force of the radiation. 



If now we investigate the successive changes produced in 

 the apparatus by the irradiation, we very easily come to per- 

 ceive that the motion of the vanes can only be brought about 

 by heat passing from the vane to the gas, or vice versa; and, 



