Eddy.l 342 [June 15^ 



Now the rotating screen c of the syren may be regarded as such a mill, 

 the surfaces of Avhose vanes may be so inclined as to return radiations 

 coming from B partly to apertures in front of those from which they 

 emanated and partly to those behind, so as to exert no force either to 

 accelerate or retard c. 



Should, however, energy be expended in moving c against the reflected 

 ray, this energy must exist immediately after the reflection in the reflected 

 ray and be transmitted by it to B. Hence we are led to the following 

 remarkable result : — on the hypothesis that radiations cause pressure at 

 surfaces at which they suffer total reflection, a part of the energy of the 

 radiation may be expended in moving the reflector against a resistance 

 while the remainder is all reflected to the bodj" from which it emanated. 

 It is to be noticed that this process of the reflecting mill or mill as it may 

 be called for brevity, is, if possible, in more pronounced and unequivocal 

 contradiction to the second law than that of the syren. 



For the latter calls in question the accepted law of mutual exchanges 

 and the second law as depending upon it, but the former applies to a single 

 body alone as B, and a moving reflector. For example, let B have no 

 radiations except those through the apertures b, then if that part of its 

 radiations which are not expended in turning c are returned to it, it is 

 possible for the mill c to be turned by radiations from B until the energy 

 of B is all expended in performing work, thus withdrawing all heat from 

 B while no heat has been transferred to any other body in the manner 

 required by the second law, and this regardless of the temperature of 

 surrounding objects. It therefore seems to me that the supposition of a 

 pressure at I'eflecting surfaces is more directly opposed to the second law 

 than that of no pressures. 



In regard to the second point mentioned, it seems quite possible to con- 

 struct a syren such that the reflections in it shall all take place from 

 stationary surfaces, or from those whose velocity differs from zero by less 

 than any assignable quantity. For let the mean velocity ic of the screens 

 be the same as before, but not coiitinuous. Instead, let it consist of sud- 

 den steps forward, each of which is half the width of an aperture. The 

 possibility of a mechanical arrangement which could effect this motion, 

 without expenditure of energ}^ with the aid of perfect springs, fly-wheels, 

 detents, etc., to any required degree of approximation will, I think, be ad- 

 mitted, certainly by any one who can admit that Maxwell's "sorting 

 demon" expends no energy in opening and closing apertures. 



It will be seen that the reflections all take place from screens at rest (or 

 nearly so) in this modified syren, and that the same transmissions occur 

 through its apertures as have heretofore been supposed to take place. 



I am not inclined, however, to insist on the special kind of apparatus 

 which I have proposed for rendering sensible the phenomenon which I 

 believe to exist during the time in which the radiations are in process of 

 becoming established, as contemplated in the ordinary law of thermal ex- 



