146 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1920. 



It was not until the time of Sir John Herschel and Pouillet, in the 

 decade 1830-1840, that attempts were made to get accurate measure- 

 ments of the heat of the sun. And it was not until the decade 1900- 

 1910 that the methods and apparatus for this purpose attained such 

 perfection that results accurate to the order of 1 per cent were ob- 

 tained. Inquirers who may live 1,000 years hence can, we hope, 

 refer to the measurements of the Smithsonian Institution begun in 

 1902 in order to settle the question whether the sun's heat has gradu- 

 ally declined in the millennium intervening. 



If we could set up a tube reaching to the outer limit of the atmos- 

 phere and large enough to see through it the whole of the disk of the 

 sun, and exhaust the air from it entirely, then the measurements of 

 the sun's intensity of radiation would be very simple. But situated as 

 we are underneath an ocean of air charged with dust, clouds, water 

 vapor, carbon dioxide, and even, we may say, of the molecules them- 

 selves, such solar researches are very difficult. In order to minimize 

 these difficulties as far as possible, such studies are best conducted in 

 the most dry and cloudless regions at high-altitude stations. 



In the course of the work carried on by the Smithsonian Institution 

 measurements have been made at Washington, sea level; Bassour, 

 Algeria, 3,600 feet; Hump Mountain, North Carolina, 4,800 feet; 

 Mount Harqua Hala, Arizona, 5,600 feet ; Mount Wilson, California, 

 5,800 feet; Calama, Chile, 7,500 feet; Montezuma, Chile, 9,500 feet; 

 Mount Whitney, California, 14,500 feet; and finally from a free 

 balloon which was sent up from Omaha, Nebraska, carrying automatic 

 recording instruments, to an altitude of about 15 miles. The accuracy 

 of these measurements has depended on the able cooperation of my 

 colleagues, Messrs. F. E. Fowle, L. B. Aldrich, A. F. Moore, L. H. 

 Abbot, and others, observers ; Mr. A. Kramer instrument maker ; and 

 Miss F. A. Graves and others, computers. They have been made, some 

 in summer, some in winter, some in clear skies, and others in skies 

 made hazy by the dust from the gigantic eruption of Mount Katmai, 

 Alaska, in 1912, but their results are in substantial agreement and give 

 the mean value of the intensity of solar radiation to a probable ac- 

 curacy of 1 per cent. I say the mean value because the investigations 

 have shown that the sun's output of radiation is not constant, but 

 varies from year to year and even from day to day within the year. 



The solar variations of long interval seem to be associated with the 

 general activity of the sun, so that higher values of the sun's emis- 

 sion are found when sun spots, prominences, faculae, and other solar 

 phenomena are more than usually marked. At such times, strangely 

 enough, the temperature at most weather stations on the earth aver- 

 ages below the normal. This paradox of increased solar heat and de- 

 creased terrestrial temperature may perhaps be explainable on the 

 basis that increased cloudiness occurs at times of sun-spot activity. 



