B. P. Langley on the Selective Absorption of Solar Energy. 157 



plane always perpendicular to the line joining its centre and 

 the slit. The mirror and the bolometer, with their attach- 

 ments, were fastened to this movable circle. 



An allowance has been made for the Fig. 1. 



absorption of speculum-metal and silver; 

 but the absorption of the iron strips of 

 the bolometer has only been indirectly 

 allowed for. This has been done by 

 comparison with the action of a bolo- 

 meter with lampblacked surface. It 

 will be seen hereafter that the whole 

 experiment has been repeated with a 

 lampblacked bolometer without in any 

 way affecting the results. The wave- 

 lengths are derived from the measured 

 angles by the use of the formula 

 nsX = sin i + sin r, where n is the 

 order of the spectrum, s the space 

 between the lines of the grating, \= 

 the wave-length of the radiation, i the 

 angle of incidence (in the present 

 instance 0°), and r the angle of dif- 

 fraction. 



In the early observations it ap- 

 peared from the examination of the dif- 

 fraction-spectrum up to X = 1^*0 that 

 the energy in the invisible part as far 

 as this was much less than in the visible. 

 Nothing definite is even at this time 

 known to physicists as to the extent 

 of the normal solar spectrum; but the 

 prismatic spectrum is still very com*- 

 monly supposed to be limited by theo- 

 retical considerations to an extent little greater than this; and 

 one of those most conversant with the subject has treated this 

 wave-length (i. e. l^'O) as marking the limit of everything 

 known to exist*. 





* Draper, "On the Phosphorograph of a Solar Spectrum, and on the 

 Lines in the Infra-red Region," Philosophical Magazine, vol. xi. p. 167, 

 March 1881. He asks, "Do we not encounter the objection that this 

 wave-length, 1O750 - 10 metre (the limit of Captain Abney's map), is 

 altogether beyond the theoretical limit of the prismatic spectrum ? " 

 Previous measurements of heat had, it will be remembered, been made by 

 comparing its total amounts in the visible and invisible prismatic spec- 

 trum, which gives us no knowledge as to wave-lengths in any case ; and 



