32 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. IO7 



the mean of 310 recurrences in the months March to November. The 

 remaining over 800 recurrences had to be omitted because of lacunae 

 in the record. 



Tabulations were made of solar-constant data under two assump- 

 tions : First, that the phase of the periodicity remained unchanged in 

 the solar emission ; second, that the solar phases shifted in the same 

 way that the phase of terrestrial temperature effects shifts. The first 

 hypothesis proved correct. There is no shifting of the phase in the 

 sun, and the dates of recurrence of the period in solar radiation may 

 be predicted indefinitely. 



Tabulations were made of solar-constant data in two groups. One 

 group included only months when the temperature effect was of great 

 amplitude, the other group the opposite. No appreciable difference 

 was found in the two averages of solar change. 



The average amplitude of the solar change is so surprisingly small, 

 compared to the 0.7 percent estimated in 1936, that two tabulations 

 were made for occasions when changes of the solar constant had been 

 individually recorded with tolerable completeness. One group con- 

 tains only large, the other only small, observed solar changes. There 

 was a decided difference in the two averages of results. The high 

 group average was about 0.5, the low group average about 0.2 percent. 

 It is believed that the solar changes attending the periodicity range in 

 reality from zero to i percent or more. The unselected group of 310 

 solar-constant values, giving 0.13 percent range in amplitude, doubt- 

 less contains a majority of solar changes too small to observe with 

 certainty separately in the record. 



The average amplitude of solar change is so small, compared to the 

 average amplitude of temperature change associated with the 6.6456- 

 day periodicity, that one is inclined to attribute the terrestrial tem- 

 perature effect to some indirect action, rather than to direct heat 

 effects of change in the insolation. Such an indirect action may be 

 produced by large percentage changes in extreme ultraviolet solar 

 radiation, producing large percentage fluctuations of atmospheric 

 ozone, and large changes of absorption by ozone in outgoing earth 

 rays, at about 10 microns wave length. I have not been able to test 

 this hypothesis. 



Attempts were made to discover correlations between the ampli- 

 tudes of terresterial heat effects of the cycle and several known cosmic 

 fluctuations, but without success. Tabulations are given of changes 

 in phase and amplitude of temperature effects in Washington, St. 

 Louis, and Helena, in the hope that meteorologists may thereby dis- 

 cover correlations with atmospheric circulation, which may enable 



