494 Intelligence and Miscellaneous Articles. 



if the energy of the motion of translation of the molecules of the 

 first body itself is greater than that of the second. 



Erom these determinations we get at a rigorous definition of 

 the temperature of luminescence itself. 



The temperature of luminescence for any kind of radiation or 

 other intramolecular motion is that temperature to which a non- 

 luminescent body must be brought, so that this motion may be 

 equivalent to that of the luminescent body. It is clear that the 

 second law of the mechanical theory of heat, which is based on 

 Clausius' principle, cannot be applied in the form fd Q/T = in 

 all the cases we have considered, but must be in accordance 

 with the expression fd Q/T< 0. 



I may be permitted to elucidate the first of the above principles 

 by reference to some e :amples. 



"We bring calcspar from the temperature of 0° to phosphorescence ; 

 it is then bright red. Suppose we introduce it into a hollow platinum 

 shell at a higher temperature, say 1°. Let the platinum sphere 

 be supposed to be exhausted, so that there is neither convection 

 current nor conduction. For the ultra-red rays which proceed 

 from the platinum Clausius' law holds in its original form ; but as 

 calcspar is diathermanous for these rays, they do not heat it. We 

 could finally put an envelope at C between platinum and calcspar, 

 which should be adiathermanous for the heat-rays auddiathermanous 

 for the visible rays. The red rays proceeding from the calcspar 

 are absorbed by the platinum, and this is heated thereby ; hence 

 heat has passed from a cold to a warmer body without a simultaneous 

 production of work. 



In many flames we have to do with phenomena of luminescence; 

 this follows in the first place from the experiments of W. v. Siemens *, 

 and further from those of H. Ebert t, who proved that when a 

 flame is dulled by introduction of carbonic acid, and is thereby 

 greatly qooled, the emission in the ultra-violet at about 500° is almost 

 exactly the same as that when the flame of 1000° is deadened by 

 admission of air. If we suppose such a flame burning in a platinum 

 cylinder at 100°, the emission of which is still small in the ultra- 

 violet, and if we interpose a body which absorbs all the visible rays 

 and only transmits the ultra-violet rays, then ultra-violet rays will 

 pass from the flame to the platinum, so that they will be absorbed 

 and the platinum heated. 



If we do not restrict ourselves to the consideration of the trans- 

 mission of energy from the luminescent body, but also take into 

 consideration the source of energy which excites the luminescence, 

 the principle of Clausius is thus valid in its old form, as is at once 

 clear in many other cases. The luminescent body is then as it were 

 only an intermediate member which effects the transition of a hot- 

 body to a cold one ; an intermediate member which, however, is colder 

 in the ordinary sense of the word than the two terminal members. 



"We have excited calcic carbonate by rays of the sun, which cor- 

 respond to a much higher temperature than that which the pla- 

 tinum foil possesses. The transition takes place in such a manner 



* Wied. Ann. xviii. p. 311 (1883). 



t Sitzungsberichte der Phys.-med. Soc. Erlangen, July 1889. 



