30 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 65 



true absorption occurring in spectrum lines which are so fine as to 

 have escaped discovery hitherto, although so numerous as to produce 

 a profound effect upon the transmission of the atmosphere. Indeed, 

 Mr. Very says in another place that one may prove that atmospheric 

 losses in the atmosphere are at least three times as great as are indi- 

 cated by Rayleigh's theory of scattering, or by the secant formula of 

 extinction. We have found by balloon experiments, as we shall 

 show, that the radiation at a level of about 25 kilometers, where 

 more than twenty-four twenty-fifths of the atmosphere lies below, 

 is still not greater than 1.9 calories per sq. cm. per minute. Hence 

 the condition of affairs referred to by Mr. Very, if it exists, applies 

 only to the very highest layers of the atmosphere, exerting less 

 than one twenty-fifth part of its pressure. Apparently, however, his 

 strongest evidence of this supposed condition of affairs is his fixed 

 impression that the solar constant must be greater than we have 

 found it. 



As to the effect on solar radiation of particles too gross to diffract 

 the rays, this must refer to dust particles, or agglomerations of dust 

 and other materials about nuclei of one kind or another, perhaps 

 about the hydrols which are thought by some to exist in the atmos- 

 phere. In regard to this we have only to refer to that line of table 

 2 which shows the transmission of the atmosphere for July 26, 1912, 

 when it was filled with volcanic dust. The atmospheric transmission 

 was then greatly reduced, but in a manner to make the sky white, 

 not blue. Hence we may say that the particles composing the dust 

 were large as compared with the wave length of light. But our 

 values of the solar constant obtained both at Bassour, Algeria, and 

 at Mt. Wilson, California, did not differ appreciably from those we 

 had obtained in the clearest of skies. 



It is urged that there are diffuse bands of atmospheric absorption 

 which have escaped detection, but which, if taken account of, would 

 increase the value of the solar constant of radiation. We call atten- 

 tion here to the results published by Mr. Fowle, 1 in which he 

 determined in the ordinary manner, from Washington observations, 

 transmission coefficients in the great infra-red water vapor bands. 

 These transmission coefficients, as he showed, sufficed almost, or 

 quite, to obliterate these bands from the energy curve of the sun 

 outside of the earth's atmosphere, just as they ought to do, if effect- 

 ive, seeing that no water vapor exists in the sun. If, now, there are 

 other bands which are so inconspicuous that they cannot be found 



1 Smithsonian Misc. Coll., Vol. 47. 





