540 Royal Institution. 



It must not however be omitted to notice, however briefly, another 

 recent set of researches of high interest, those of M. Silberman ; in 

 which (among others) the very remarkable fact is established, that 

 on transmitting a narrow ray of heat from a heated wire, through 

 rock-crystal, there is a singular difference according as the ray passes 

 parallel or perpendicular to the axis of the crystal : the effect being 

 indicated by having the further side of the crystal coated with a fine 

 composition of wax, the portion of which in the direction of the ray 

 is melted in a circular form in the first instance and in an elliptical 

 in the second. 



The general fact of the heterogeneity of heating rays, especially 

 from luminous sources, is fully recognised by Melloni as in some 

 sense the conclusion from all his experiments. 



The hypothesis that this heterogeneity consists simply in differences 

 of wave-length would seem a probable one ; though it is still possi- 

 ble, as Professor Forbes suggests, that some other element may also 

 enter into the conditions. 



This view has been extended by M. Ampere so as to refer both 

 luminous and heating effects to the same rays : — a view controverted 

 by Melloni, chiefly on the ground, evinced by several classes of 

 experiments, that the intensity of the heating effect (especially in the 

 solar rays) follows no proportion to that of illumination ; an argu- 

 ment which really amounts to little, unless the theory obliged 

 us to infer that the amount of illumination must follow the same 

 law as that of heat ; which it manifestly does not ; since the 

 nature of the effect in the one case is wholly dependent on the 

 unknown constitution of the optic nerve ; according to which some 

 precise proportion of the impinging vibrations, with a particular 

 wave-length, is that which gives the greatest perfection of vision ; 

 while for heat the effect has no reference to such peculiar conditions, 

 but is dependent in some way on longer wave-lengths, and pro- 

 bably more simply connected with the intensity or amplitude of 

 the vibrations. 



On this theory our view of the case would be thus :— 



A body heated below luminosity begins to give out rays of large 

 wave-length only. As it increases in luminosity it continues to send 

 out these, and at the same time others of diminishing wave-lengths, 

 till at the highest stage of luminosity it gives out rays of all wave- 

 lengths from those of the limit greater than the red end of the 

 spectrum, to those of the violet end, or possibly less. 



Rays of all these species are transmissible and refrangible by rock- 

 salt; and many of them with numerous specific distinctions by other 

 media. 



They are all more or less capable of exciting heat when absorbed 

 or stopped ; though in some the effect is perhaps insensible. Both 

 this property and that of their transmissibility seems to depend in 

 some way on the wave-length, though in no simple ratio to it. 



The absorptive effect due to texture of surfaces has some direct 

 relation to the magnitude of the wave-length, especially near the 

 limit ; while that due to darkness of colour is connected with 



