NO. 2 SMITHSONIAN EXPLORATIONS, I924 29 



On such grounds we have not to expect any simple or easy solution 

 of the problem of the influence of variations of the sun on our weather. 

 Apparently there is only one way to proceed. It is to observe dili- 

 gently and accurately for years the sun's changes of radiation, and 

 then, with the basis of fact so laid, examine the behavior of atmos- 

 pheric pressure, temperature and rainfall at a great number of stations 

 over all the world, at all times of the year, in order to work out at 

 last the exact dependence of weather on solar changes. 



This is a very large program. The Astrophysical Observatory 

 began with it in 1902, when it commenced to observe the solar radia- 

 tion in Washington. But the atmosphere in Washington was loaded 

 with clouds, smoke, dust and humidity. It was necessary to remove 

 to a purer sky. Since then stations have been occupied as follows : 

 Mt. Wilson, Calif., during summer and autumn, 1905 to 1920; Mt. 

 Whitney, Calif, (the highest mountain of the United States outside 

 Alaska) on several days in 1908, 1909, and 1910; Bassour, Algeria, 

 in summer and autumn 1911 and 1912; Hump Mountain, N. C. 

 almost a full year beginning May, 1917 ; Calama, Chile, July, 1918, to 

 July, 1920; Montezuma, Chile, August, 1920, to the present time; 

 Mt. Harqua Hala, Ariz., October, 1920, to the present time. 



During all this time better and better methods have been developed, 

 obstacles to accuracy overcome, and new sources of error recognized 

 and avoided. The body of experience which has thus come to the 

 members of the stafl:' is unique, and a very great asset. There is, 

 indeed, the highest necessity for it. The results of Mr. H. H. Claj-ton, 

 who has attempted with much success to unravel the effects of solar 

 changes on the weather, indicate that variations of the sun as small 

 as 0.5 per cent, or 1/200 of the whole solar output, produce well 

 recognizable efl:ects. Changes greater than 3 per cent do not often 

 occur, though once in a great while they go even higher than 5 per 

 cent. Observing, as we do, at the bottom of a sea of atmosphere, 

 loaded more or less with variable elements, like dust, water vapor, 

 etc., it is very difficult to reach a high enough standard of accuracy to 

 reveal surely changes as small as these. 



During the past four years, with the newest methods of observation 

 and computation, our two observ-ing stations in Arizona and Chile 

 each determined the sun's heat on upwards of 70 per cent of all days. 

 Taking all fairly good days when both stations observed, the average 

 deviation of these independent measures, made over 4,000 miles apart 

 and reduced independently for atmospheric losses, is about 0.5 per 

 cent. During October, 1923, however, both stations reported almost 

 every day, and the average daily difiference was less than 0.2 per cent. 



3 



