abbot] studies of solar constant of radiation 75 



water vapor in the atmosphere is also both very effective and very 

 variable.^ 



The procedure employed here to determine the general absorption 

 of the air consists chiefly in making holographs — that is, automatic 

 energy spectra — of the solar radiation as often as possible throughout 

 days of uniform and excellent sky without alteration of the sensitive- 

 ness of the apparatus. Such energy spectra are altered in appear- 

 ance from one to another by the varying absorption of the different 

 thicknesses of air, so that at a little after noon the height of the curve 



^ K. Angstrom has, however, attributed much importance to the absorption 

 of carbonic acid gas, implying by his computation that not less than 6i percent 

 of the solar radiation which reaches the outer layers of the earth's atmos- 

 phere is cut off by the absorption of this gas in a vertical transmission 

 through the air. (See Annalen der C hemic und Physik, vol. 39, pp. 309-311, 

 1890.) He locates the absorption of this gas principally in the bands at 2.6/^ 

 and 4.3,"; so that, as he says, its effect is not allowed for in the procedure 

 for obtaning the value of the solar constant of radiation adopted by Mr. 

 Langley in his research on Mount Whitney, and which is essentially that 

 employed here. Angstrom, while using the same method in part, adds a 

 second term amounting to more than half the whole in his computation, 

 solely referring to the absorption of carbonic acid gas, and thus he attains 

 his oft-quoted result for the solar constant of radiation of 4.0 calories per 

 square centimeter per minute. For several reasons I am inclined to think 

 Angstrom has greatly overestimated the importance of this carbonic acid 

 absorption term : First, as he shows, the selective absorption of carbonic 

 acid gas is, so far as I am aware, almost wholly for wave-lengths greater 

 than 2.5 /" and principally in two bands between wave-lengths 2.5/<and 2.85;" 

 and between 4.20/' and 4.50," respectively, where the total amount of the solar 

 radiation is apparently less than one percent of the whole, as determined 

 not only from the appearance of the observed holographic solar spectrum 

 energy curve itself, but from a consideration of the probable temperature of 

 the sun and the distribution of energy in the spectra of bodies at high tem- 

 perature. As a very evidently too great estimate of the energy in these wave- 

 length regions, it may be seen that if the radiation outside the atmosphere 

 (see plate xxn) was of the same intensity throughout these bands as at 2.I|M. 

 the area they would include would be only about one-fiftieth the total area 

 under the curves of plate xxn. It is of course very improbable that the height 

 of the curve at 4.3/^ is nearly as great as at 2.1 . Thus it would appear 

 that the selective absorption of this gas for direct solar radiation is almost 

 negligible. Second, if carbonic acid exercised a general absorption through 

 the more intense parts of the solar spectrum, it is not apparent why such a 

 general absorption is not included and allowed for in the coefficients of 

 absorption here determined. Third, values of the solar constant computed 

 here for the same day, but from observations made through very different 

 thicknesses of air, are found to agree excellently, which appears to confirm the 

 accuracy of the method of determining the atmospheric absorption which is 

 here employed. 



