10 



SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I46 



h9 



1.8 



|/.7 



.3 





^f July 



— OcU9rt ft^ — 



^Aii. 





jAbhoi.f t-WiUwt. 

 ' Max. 





Ocm. %. 



20 



Ala. 



6\0 



MO 



Bdrometer. 



Fig. 7. — Maximum pyrheliometry, sea level to 15 miles altitude. 



three or even five determinations of the solar constant per day and 

 could utilize days vv^ith cumulus clouds intermittently — days quite unfit 

 for Langley's method. The short method is, indeed, empirical and 

 must be set up separately for each observing station by observing a 

 year or more simultaneously with Langley's method to standardize it. 

 We continually improved the "short method" till 1926, but after that 

 we used it exclusively except for occasional checks by Langley's 

 method. 



3. ACCURACY OF "SHORT METHOD" SOLAR CONSTANTS 



In volume 6, page 163, Annals, A.P.O., are compared the solar 

 constants observed on 616 identical days at Mount St. Katherine in 

 Egypt and Mount Montezuma in Chile. Winter at one station cor- 

 responds with summer in the other. The difference between daily 

 results ranged from to 0.028 calorie. The weighted mean differ- 



