NO. 5 SOLAR VARIABILITY ABBOT, FOWLE, ALDRICH 3 



The reader will understand that the beam from the coelostat A B, 

 or that from the coelostat / K, may be brought to the spectro- 

 bolometer, according as the 45° mirror F is in place or withdrawn. 

 The former is the condition for observing the distribution of light 

 over the sun's disk, the latter the condition for observing the solar 

 constant of radiation. 



As explained in the publications already cited, we allow the sun's 

 image to drift across the slit of the spectro-bolometer by the diurnal 

 motion of the earth, thus avoiding all sources of error associated 

 with inequalities of the transmission of the optical apparatus of the 

 tower telescope. For it is obvious that during a single such period 

 of drift the telescope is always directed to the same point of the sky, 

 and treats whatever object passes that point with impartiality, whether 

 it be the sun's limb, center, or the light of the sky itself. We regard 

 this feature as very favorable for exact results, and much preferable 

 to the arrangement used by some investigators, in which the observ- 

 ing apparatus is shifted about from part to part of the solar image, 

 and the results may be affected by inequalities of transmission of the 

 optical apparatus to these different parts of the image. 



On the other hand, we are thereby limited to an east and west 

 course across the sun's disk, and this hardly ever coincides with the 

 solar equator. However, it seems clear that a comparison of the 

 mean of a great number of observations taken during a large part of 

 one year with that of a great number of similar observations covering" 

 the same part of another year, would certainly be fairly comparable 

 irrespective of the various presentations of the sun during these inter- 

 vals. It is by no means so clear, without further investigations 

 which we hope to undertake next year, that short-period changes of 

 the distribution of radiation along a solar diameter may not be asso- 

 ciated with changes of distribution depending on latitude in the sun. 

 Such investigations as have been made on this point heretofore relate, 

 we think, only to total brightness or total radiation. Pickering's * 

 experiments indicated that the contrast of visual brightness along a 

 polar diameter exceeds that along the equatorial, so that for total 

 visual brightness at 95 per cent out on the radius the equator is 

 brighter in the ratio of about 56 to 53. Pickering expressed doubt 

 as to whether this difference is solar or from experimental error. 

 The more numerous investigations of Langley ^ seem to have shown 

 clearly that differences of contrast in total radiation between the equa- 

 torial and polar diameters of the sun are negligibly small, probably 



^ Proc. Amer. Acad. Arts Sci., Vol. 10, p. 42S, 1874. 

 " Comptes Rendus Sept. 6, 187^. 



