between Metallic Masses having different Temperatures. 461 



64. I have been led to entertain the idea of a new species of 

 mechanical agency in heat, not from a love of introducing novel 

 principles, but after having been driven by experiment from the 

 hypotheses to which I was at first entirely attached. Although 

 the mechanical effect of the repulsive power of heat cannot be 

 said ever to have been demonstrated, experiments are not want- 

 ing which seem to be quite inexplicable without its aid, or some 

 other principle not yet recognised in science. 



65. Several ingenious French experimentalists have furnish- 

 ed us with facts, which, though not completely established as be- 

 longing to any peculiar class of phenomena, and therefore not 

 generally admitted into systematic works, are not the less worthy 

 of notice. Those which bear most directly on our present spe- 

 culation were observed by M. FRESNEL * ; namely, the repulsion 

 of disks of mica, of which one was placed at the extremity of a 

 delicately suspended needle in vacua, and when the disks were in 

 contact, heated by means of a ray of solar light concentrated by 

 a lens. M. SAIGEY f has also described a class of similar pheno- 

 mena observed by himself, with a considerable number of metals, 

 which, after rejecting the influence of aerial currents, of electri- 

 city, magnetism, &c., he ascribes to the repulsive action of radiant 

 heat at sensible distances. There are several other experiments 

 on record which seem to require a similar explanation, but I ap- 

 prehend that the present are the first to establish the existence 

 of some species of mechanical repulsion in the propagation of 

 heat, a principle which can hardly fail to be applicable to the ex- 

 planation of many natural phenomena. 



GREENHILL, EDINBURGH, 

 IQth February 1833. 



* Annales de Chimie et de Physique, xxix. 57. and 107. 



+ See several successive articles in the Bulletin des Sciences Mathematiques, 

 torn. ix. See also POUILLET, Ekmens de Physique. 



2 N2 



