of the Atmospheric Absorption. 307 



say that they seem to show that the average amount of blue 

 light (to speak for the moment of blue light only) which is 

 thus scattered from an ordinary pure and cloudless sky at the 

 sea-level already represents a selective absorption of much 

 over 40 per cent, of the original blue in the direct sunlight ; 

 but that also the mean diffusion is, though less than this, still 

 over 40 per cent., and hence that to obtain the actual light of 

 sun or star before absorption Ave must add, at any rate, over 

 40 per cent, to the observed value. To make the meaning of 

 this last statement clear, it may be observed that if there were 

 bright clouds in the sky we should have (as we know by expe- 

 rience) more light from the clouds than from an equal portion 

 of the blue, but that in this case our observing station would 

 gain the added light at the expense of those portions of the 

 earth which were in the clouds' shadow ; and in this case, 

 therefore, we should not be justified in adding the light we 

 receive to the observed sunlight to obtain that before absorp- 

 tion. But with a uniform sky it follows that every point on 

 the horizon enjoys the same sunshine that we do at our own 

 station; and hence it is evident that if the atmosphere were 

 taken entirely away the sun would grow brighter by all the 

 light which the atmosphere now sends us, and, in fact, by 

 much more, since this atmosphere is scattering light not only 

 towards us, but also to outer space. 



Roughly speaking, we may say that we receive on the 

 average at the sea-level as much light from the sky as we do 

 from the sun itself — getting more light from the sun at mid- 

 day than from the sky, but more in the morning and afternoon 

 from the sky than from the sun : There is no reason to 

 believe that any smaller proportion of starlight is diffused 

 earthward, and this is independent of that which is diffused 

 toward outer space. 



All my own investigations, whether through observations 

 at the sea-level or at an altitude of nearly 15,000 feet, lead 

 me to believe that at any rate the mean absorption of lio-ht 

 (and of heat also) by our atmosphere is probably at least 

 double that which is customarily estimated ; and also to con- 

 clude that fine dust-particles, both near the surface and at a 

 great altitude, play a more" important part in this absorption, 

 both general and selective, than has been heretofore supposed. 

 I hope soon to make public the result of an investigation into 

 the composite character of the radiation as an indication of 

 what the sky has selectively borrowed from the direct solar 

 rays. 



