220 The Smithsonian Institution 



Another result of these experiments was the establishment 

 of the fact of selective absorption of the solar rays by the 

 earth's atmosphere. In regard to this Mr. Langley wrote at 

 the time : 



" Our observations at Allegheny had appeared to show 

 that the atmosphere had acted with selective absorption to an 

 unanticipated degree, keeping back an immense proportion 

 of the blue and green, so that what was originally the strong- 

 est had, when it got down to us, become the weakest of all, 

 and what was originally weak had become relatively strong, 

 the action of the atmosphere having been just the converse 

 of that of an ordinary sieve, or like that of a sieve which 

 should keep back small particles analogous to the short wave- 

 lengths (the blue and green), and allow freely to pass the 

 large ones (the dark-heat rays). It seemed from the obser- 

 vations that the atmosphere had not merely kept back a part 

 of the solar radiation, but had totally changed its composition 

 in doing so not by anything it had put in, but by the selec- 

 tive way in which it had taken out, as if by a capricious in- 

 telligence. The residue that had actually come down to us 

 thus changed in proportion was what we know familiarly as 

 'white' light, so that white is not 'the sum of all radiations,' 

 as used to be taught, but resembles the pure original sun- 

 light less than the electric beam which has come to us through 

 reddish-colored glasses resembles the original brightness. 

 With this visible heat was included the large amount of in- 

 visible heat, and, if there was any law observable in this 

 ' capricious ' action of the atmosphere, it was found to be this, 

 that throughout the whole range of the then known heat- 

 spectrum the large wave-lengths passed with greater facility 

 than the shorter ones." 



Most of these observations were carried on in Allegheny. 

 In 1878, however, he made observations of the solar eclipse 

 from the summit of Pike's Peak, at an elevation of fourteen 



