324 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1910. 



knowing if the sun was on or off. It is wonderful that out of such a 

 maze the truth was approximately found at last. 



^Vlien the Astrophysical Observatory was founded at Washington 

 the bolometer had been so far subdued that Langley introduced the 

 beautiful device of photographically recording the galvanometer 

 light spot on a moving plate, while the same clock which moved the 

 plate also moved the spectrum over the bolometer tape. Thus an 

 automatic solar spectrum energy curve could be taken without mak- 

 ing a single galvanometer reading. But " drift," though no longer 

 a meter a minute, was still an obstacle. Several devices have since been 

 applied by means of which " drift '' is practically eliminated, so that 

 the galvanometer light spot stays day after day practically unmoved, 

 except as the sun is allowed to shine through the spectroscope. We 

 now usually take an energy curve of the solar spectrum, running 

 from the band of silver transmission near wave length 0.33jLt in the 

 ultra-violet, to wave length 2.5ju, in the infra-red in eight minutes. 

 Such a curve shows more than even Keeler could have found with the 

 old apparatus in a lifetime. One observer may now easily carry on 

 " solar constant " work without help. 



In his " Report of the Mount "W^iitney Expedition," Langley states 

 that the measurement of the " solar constant " encounters two diffi- 

 culties, one of which he describes as " formidable," the other as 

 " perhaps insurmountable." The first is the difficulty of completely 

 absorbing and accurately measuring the intensity of the solar rays 

 as they reach the earth's surface. The second is the difficulty of 

 correctly estimating the loss they suffer in traversing the atmosphere. 

 We shall recur to the latter. After eight years of effort to overcome 

 the former I agree that it was " formidable." 



In 1894 Prof. V. A. Michelson, of Russia, published an account 

 of his pyrheliometer. In this instrument he employed a tube-like 

 chamber, blackened within to absorb the rays, and surrounded by 

 melting ice and water. The amount of solar heating he determined 

 by noting the increase of volume of the ice as it melted, reading for 

 this purpose a graduated capillary tube attached to the outer chamber 

 wall. Michelson's pyrheliometer, which has been, I think, little used, 

 may have given correct results, but excepting for it I believe there 

 has been no accurate standard pyrheliometry until this year, 1910. 

 Unfortunately the importance of Michelson's device was overlooked 

 because his description was published in the Russian language. 



The electrical compensation pyrheliometer of Angstrom was de- 

 scribed in 1898, and has attained wide acceptance. It was adopted 

 by the Solar Union as a standard at the Oxford Conference, but the 

 experience of Kimball and of Callendar is unfavorable to it, for 

 there is a deterioration after some years in practically every in- 



