10 



SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 128 



Having shown by these two figures examples of the smoothing of 

 the solar-radiation periodicities by removing overriding shorter peri- 

 odicities integrally related, I add that these two examples are by no 

 means exceptional. Twenty others as convincing might be similarly 

 presented. My ignorance of hydrodynamics prevents me from sug- 

 gesting a theory to explain why this large family of integrally related 

 periods occurs in solar radiation. Surely, however, it is a problem 

 that deserves the attention of experts in stellar theory, for the sun 

 is a convenient laboratory for stellar investigations. 



55 61 67 



-The 68i-month period in solar variation as cleared of subordinate interfering 

 integrally related periods. 



Some critics, incredulous of the causation of weather changes by 

 such small percentages in solar radiation, may be inclined to turn the 

 phenomena about. They may suggest that some terrestrial atmospheric 

 condition sets up the family of periods first in weather elements. 

 Then they may suggest that these atmospheric changes produce, by 

 our inability to entirely eliminate such influences, the observed tiny 

 percentage periodic fluctuations in solar-constant measures. Such an 

 explanation seems untenable to me. For the solar-constant periodic- 

 ities go on with perfect regularity of phases over long periods of 

 years, as table 2 and figure i show. This is so, despite the measure- 

 ments being made in two hemispheres. The terrestrial periodicities, 

 on the other hand, are altered in phases (a) with the season of the 



