I08 CARNEGIE INSTITUTION 



fifth, bolometric observations on the spectral energy distribution of 

 the standard radiator or " black body." 



The Absorption of the Sun's Atmosphere. — While there is not as 

 good an opportunity to study the Sun's absorption as the Earth's, 

 much may be learned by forming spectrum energy curves at various 

 parts of the disc, for it is apparent that as the limb is approached 

 greater and greater thicknesses of the outer solar layers must be 

 traversed by the radiations from the interior. Thus a comparison 

 of the ordinates of the. spectrum energy cur\^es of a solar beam start- 

 ing at 98 per cent of a radius from the center with that of the 

 central beam itself discloses a powerful absorption of the shorter 

 wave-lengths in this lengthened path through the solar envelope 

 amounting even in the yellow to more than 50 per cent of the whole 

 radiation at the center. It has been supposed by some that there is 

 considerable variation in the temperature and consequent absorption 

 capacity of the solar envelope, and it is to this that Halm, in a recent 

 paper, attributes the eleven-year sun-spot period. The investiga- 

 tion of this question through a term of years by the aid of spectrum 

 energy curves from various parts of the Sun's disc would be of great 

 interest. A comparison of such observations at the high and low 

 station would prove whether or not such results, uninfluenced by 

 the absorption of the Earth's atmosphere itself, may in future be 

 obtained at low stations. 



It is unnecessary, after what has been said, to describe at length 

 the particular application of the general method to a study of sun- 

 spots, faculse, and other solar features, as the suitability of it for this 

 study is apparent. 



3, Apparatus and Accessories Required. 



The Coelostat. — The first essential for each station is the provision 

 of a fixed horizontal beam of sunlight. While there is opportunity 

 for difference of opinion as to the best provision for this purpose, 

 I am disposed to recommend, on the score of simplicity and satis- 

 factoriness of operation, the coelostat, so called, in preference to 

 any siderostat. I would propose for each station a long polar axis 

 driven at the rate of one revolution in 48 hours, by powerful clock- 

 work, and capable of carrying in the plane of its center several 

 plane mirrors. One of these mirrors, of not less than one meter 

 aperture, is to be used for bolometric purposes. The others may be 

 used to furnish fixed beams for other researches at the same time. 



