NO. 10 



SOLAR-CONSTANT VALUES ABBOT 



The reason for omitting periodicities of 21 and 25 months, and in- 

 serting one of 22^ months, is that in the present analysis the periodic- 

 ities of 21 and 25 months failed to persist without change of phase 



jWfl*y '923 36 29 3 7 ^ « 



H I H^ Hf SI 



tff H7 HW H<i ^«/Vf' 



V2- 3*9 V/ V3 Hy ^9Yi 



tiS W6 «/i' ^2L Ul i4-Ci 



Ho H-L Q3 HZ i/l H(, 



HI HI. t+S" H^/y^ *f2 



V7 ^3 HA HV H3 t/8 w: 



5"/ 5-; 



i" 39 



10 n 



li. 



w I <f 3 v>" 



V? •^t ^^ 



36 3^ 3«7 



V-L W7 W3 



MZ 4^ Q7 HI 



«#o 3^? V2- 



47 t<7>f4 



Meoxs V32 41 1 W«<Oy<Vjr *^5-/ V43 VV? VW V^^ yyi- VV7 





"(7 4<? W6 M*/ V8 W7 



ro HI hS W2. </3 4/ 



V7 V8 </8r t'V HI. ^0 



SQ SI yo V6 HI H(» 



Si. s>"o So H-7 H<o S5 



^/ 4? 49 



W9 47 WSf 



V3 ^« H% 



V4 wr WV W6 



V3 W6 V6 



If6 V9 H9 



u« H7 m 



48 W9 S7 Wf 



45- 43 4V 



4 9 A'C^ 4^" 



Ntar*^ mo H 



V^V V64 V^i V&o y7/ v^3 



44 



4*? 48 W<? W« 44 4^. 

 q4 44 44 3*7 42. 

 42 47 HT. HI 



|Vfaum7 4^ VJ" ¥•>• 



t/4 ¥^ V<» V4 WV 



H3 46 46 M7 4f 



m9 -i"^ 48 5"^ 3~/ S3 



SI H9 45" 4 7 So SO H^ 



43 HX HS hS' H 



is 47 so V6 



VX 4 2. 4z V 



A/e*»i5 45-7 4irr y^ 



Fig. I. — Computation of iii-month periodic variation of the solar constant, 

 with graphs of three general means, 1923 to 1930, 1930 to 1937, and 1937 to 1945. 



from 1923 to 1945, but that of 22^ months did seem to persist 

 throughout. 



At this point I may say that since some periodicities are nearly 

 integral multiples of others, and since, indeed, the longer periodicities 

 may be obscured in the data, unless the shorter ones are first removed 



