SOLAR CONSTANT OF RADIATION ABBOT. 323 



After Langley, everybody admitted that " solar constant " work 

 required observation of homogenous rays, but nobody practiced it 

 until 1902, when such observations were begun in Washington at the 

 Smithsonian Astrophysical Observatory. In the meantime im- 

 portant advances had come. The brilliant work in Germany from 

 1890 to 1900 had fixed the laws and constants of radiation for the 

 perfect radiator or " absolutely black body " of Kirchhoff. Hence 

 we knew approximately that the sun was of the order of 6,000° in 

 absolute temperature (centigrade), and that as its spectrum energy 

 curve determined by Langley was generally similar to that of a " black 

 body," there could be no very appreciable fraction of its radiation 

 beyond 3^ in the infra-red, or beyond 0.3ju, in the ultra-violet. The 

 positions of the infra-red atmospheric absorption bands had been de- 

 termined, and they had. been assigned to water vapor and carbon 

 dioxide. The bands of the latter compound had been found to lie 

 beyond the spectral region just named, and hence to be of little ac- 

 count to diminish solar radiation, so that Angstrom, in 1901, with- 

 drew his solar-constant value 4 calories, which he had based on a 

 supposed enormous carbon-dioxide absorption. 



Great improvement had been made in the bolometer. For " solar 

 constant " work this instrument comprises essentially two little threads 

 or tapes of platinum, each about 1 centimeter long, 0.01 centimeter wide, 

 and 0.001 centimeter thick. They are blackened to absorb rays, but 

 one is hidden from the spectrum while the other is exposed to it, so 

 that the latter is warmed by the rays with respect to the former. 

 Two equal resistance coils are joined to the two bolometer tapes, so 

 that the whole forms a " Wheatstone's bridge." The rise of temper- 

 ature of one of the tapes increases its electrical resistance, and causes 

 a very minute electrical current to flow and deflect a highly sensitive 

 galvanometer. In ordinary bolometric practice a rise of temper- 

 ature to" oVf^ centigrade is readily observed. 



You will not wonder that when this instrument was a new one it 

 was almost unmanageable. Langley has told me often that in the hot 

 lent at Lone Pine the galvanometer light spot used to rush oif the 

 scale, 1 meter long, in a single minute. Hence it took several men to 

 make an observation. One sat, sweltering (this was the immortal 

 Keeler), reading the scale as fast as he could, while another recorded 

 his numbers and also set the spectroscope. One let the sun on and off 

 the spectroscope and kept the irregularly running siderostat reflecting 

 the beam approximately right. A fourth observed with the Violle 

 pyrheliometer. It took thousands and thousands of observations to 

 determine a "solar constant" under such circumstances. The de- 

 flections could be worked out only by plotting the almost innumer- 

 able galvanometer readings which the observer had made without 



