112 



ANNUAL REPORT SMITHSONIAN" INSTITUTION, 1932 



\0 ru 



a.s 

 >.2 



Willistoii, and Yuma. The interesting result comes out that both 

 temperatures and barometric pressures in weather show oj^posite 

 courses depending on whether they accompany and follow rising or 



falling solar radiation. This is 

 shown for Washington weather 

 in Figure 4. Apparently major 

 changes in weather are caused b}^ 

 small fluctuations in the solar 

 radiation. 



If this is so, we ought to expect 

 that the regular periodicities which 

 are proved by Figure 2 to com- 

 prise the principal solar changes 

 since 1918 ought to be reflected 

 in the weather. In Figure 5 one 

 may see that this is indeed so. 

 The principal changes in Wash- 

 ington temperatures since 1918 

 are represented as the sum of 

 six regular periodicities, of which 

 five are those found in the solar 

 radiation. This gives us hope that 

 weather may be susceptible to 

 long-range forecasting. It would, 

 indeed, be a great boon if the 

 characteristics of coming seasons 

 and years could thus be approxi- 

 mately known in advance. But 

 much further study must be made 

 before this hope can be thoroughly 

 tested. 



We see from these exhibits that 

 solar variation is of two types, the 

 long range and the short range, 

 respectively. Two kinds of causes 

 probably are involved. The long 

 periods of 68, 45, 25, 11, and 8 

 months are closely related to the 

 well-known interval of II14 years 

 in which the numbers of sun spots 

 wax and wane. This suggests that 

 these longer-range periodicities are 

 due to increasing and decreasing 

 agitation in the gaseous fluid 



O aj 





