Dr. H. Hudson on the Theory of Exchanges. 343 



nished canister produces the greater chill; and it is obvious that 

 what might appear & plausible explanation with one ball in the 

 focus must be an utter failure in the case of the other focal ball. 

 In fact, however, when canister, mirror, and thermometer are all 

 at the same temperature as the air, it appears unaccountable (on 

 Prevost's theory) that the " superior radiative energy " of the 

 varnished canister does not heat the focal ball when concentrated 

 upon it by the reflection of the mirror. 



There is a very pretty experiment of Professor TyndalFs with 

 a thermomultiplier, which (although I have never tried it) will, 

 I doubt not, prove the same thing. A warm body at the dis- 

 tance of 4 feet 6 inches from the instrument shows scarcely any 

 action on the pile (most of the rays being scattered in space); 

 but on introducing a polished tin tube (4 feet long) between the 

 pile and the source of heat, the rays are reflected from its sides, 

 and thus, by several internal reflections, reach the pile and pro- 

 duce a decided motion of the needle. Only substitute (in this 

 experiment) a cold canister as the radiator, and the chilling in- 

 fluence of the varnished side will doubtless show its superior 

 efficiency. 



Seeing that the scientific world generally admitted Prevost's 

 theory, I brought forward at the first Dublin Meeting of the 

 British Association (in 1835) what I considered an " experimen- 

 tum cruris." I heated the surface of the mirror (by filling it with 

 hot water) to about 170° Fahr., and (keeping one ball in the 

 focus) shifted the differential thermometer until its liquid stood 

 at zero, both balls being then equally heated by the mirror. 

 The temperature of the air in the room was 55° Fahr. ; the 

 cubic canister (containing water at 67°), being placed just in 

 front of the screen F (see figure), acted as a radiator of heat, 

 and the varnished side was most efficient. On moving the 

 canister nearer to the mirror, the effects diminished and at length 

 ceased, the thermometer remaining at zero whether the varnished 

 or metallic side of the canister faced the mirror ; but on moving 

 the canister still nearer to the mirror, it began to act as a cold 

 body, and the varnished side produced the greater chill. 



I trusted at the time that this experiment (supplementary to 

 those of Leslie already referred to) would have satisfied scientific 

 physicists that Prevost was in error and that there could be no 

 phenomena of radiation so long as a body was of the same tem- 

 perature with the medium in which it is placed ; and therefore 

 I did not enter into the rationale of the experiment, which I 

 believe to be as follows. 



The heated mirror produces a gradation of temperature in 

 front of itself (of course an unstable equilibrium) ; and when the 

 canister is so situated that its surface and the medium in front 



