4 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I48 



I have now analyzed Wexler's smoothed values to see if 27 exact 

 submultiples of 27Z months, which I have long believed in, are verified. 

 The number of repetitions decreases from 17 , for 4^ months, to 5 



to 

 63 28 



Section 2. ±i± to ^ii« Section 3. 



27 18 



Monthly solar constants, smoothed, 1923 to 1952. Formula: C'= 



1 



100 



273 



15 



1 



100 



Solar constant = 2. 



■^—5. Sub-periods removed. 



4 



(.A+2B+AC+2D+E). 



Ordinates scale 



Fig. 2. — Solar periodic variation, 



273 



to^" —r~ = 68^ months. With only 3 repetitions for 91 months, I 



273 

 omitted -ir-. All periods longer that 1 5^ months had one or more 



over-riders, exactly their submultiples in length, whose amplitudes I 



