MEMOIR XXVIII.] DISTRIBUTION OF HEAT IN THE SPECTRUM. 339 



There are days on which, owing to the excessive absorp- 

 tion taking place among the ultra-red rays, a rock-salt 

 train has no advantage over one of glass. But if it is 

 the visible spectrum alone that we are using, and the 

 prisms are of a material colorless to the eye, we may be 

 certain that they are exerting no selective absorption on 

 any of the radiations of that spectrum, and that the in- 

 dications they are giving are reliable. 



This variable absorptive action of the atmosphere de- 

 pends partly on changes in the amount of water vapor, 

 and partly on the altitude of the sun. At midday and 

 at midsummer it is at a minimum. The disturbance 

 is not merely a therniochrose, for both ends of the spec- 

 trum are attacked. It is a matter of common observa- 

 tion that the horizontal sun has but little photographic 

 power, owing to atmospheric absorption of the ultra-vio- 

 let rays, and under the same circumstances his heating 

 power is diminished, owing to the absorption of the ultra- 

 red rays. But if the day be clear, and the sun's altitude 

 sufficient, the visible spectrum may be considered as un- 

 affected. 



It should be borne in mind that the envelopes of the 

 sun himself exert an absorptive action, which is power- 

 fully felt in the ultra-violet region, as is indicated by 

 the numerous fixed lines crowded together in that region. 

 The force of this remark will be appreciated on examin- 

 ing the plate above referred to in the Philosophical Mag- 

 azine for May, 1843. 



It seems, then, that all the conditions necessary for the 

 solution of this problem will be closely approached if we 

 make use of prisms constituted of any substance which is 

 completely colorless to the eye, and confine our measures to 

 the visible spectrum, collecting all the radiations between 

 the fixed line A and the centre of that spectrum just be- 

 vond D into one focus, and all the radiations between 



