2 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 65 



regretted that it has not generally proved practicable in connection 

 with other duties for the observers to secure measurements with the 

 air mass as great as 3. 



At Prof. Pickering's desire the observations are reduced and 

 published by the Smithsonian Institution. They were made at 

 Arequipa mainly by Dr. Leon Campbell, and in part by H. Perrine. 

 Computations are mainly by L. B. Aldrich. The position of Are- 

 quipa is: Long. 4' 1 46™ 11.73 s W., Lat. 16 22' 28" S. Alt., 2,451 

 meters. 



Nothing would be gained by making a series of pyrheliometer 

 measurements at a station no higher than Arequipa if such a series 

 did not throw light on the variability of the sun or on the varia- 

 bility of the transparency of the earth's atmosphere. Two kinds of 

 solar variability are thought to exist. One is associated with that 

 general solar activity which is indicated by faculse, sun spots, and 

 other visible solar features. This type of variability may be ex- 

 pected to march in rough correlation with the eleven year sun spot 

 cycle. Another type of solar variability appears to be of short 

 irregular intervals in its fluctuations, which are to be measured by 

 days or months rather than by years. 



As for the variations of atmospheric transparency, we need not 

 consider those caused by ordinary cloudiness. Pyrheliometer meas- 

 urements are made only when the sky around the sun is cloudless. 

 Water vapor and dust are the two variable elements which princi- 

 pally affect the atmospheric transmission of solar radiation. Water 

 vapor is effective in two ways : it absorbs radiation of certain wave- 

 lengths, particularly in the infra-red spectrum ; and it associates 

 itself with dust to produce haze which scatters the solar radiation 

 of all wave-lengths, thus increasing sky light at the expense of 

 direct sun light. 



At so high a station as Arequipa, dust, except as associated with 

 water vapor to form haze, is generally not very effective to diminish 

 solar radiation. But after forest fires or great volcanic eruptions 

 it may be of very great influence. 



The hindrance of solar rays by the atmosphere is of course de- 

 pendent on the length of path of the solar beam therein. For zenith 

 distances (Z) less than 70 the length of atmospheric path is closely 

 proportional to secant Z. Suppose one could observe the solar 

 radiation outside the atmosphere, and also at the earth's surface at 

 zenith distances whose secants were 1, 2, and 3. Let the four values 

 of the intensity of radiation be c Q , c x , c 2 , c 3 , respectively. Let the 



fractions-^-, -^-,-^-, be denoted by a x , a 2 , a 3 , respectively. These 



