102 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1920. 



per cent (in defect) in the Mount Wilson solar constant values of 



1919. Considerable additional labor was required in the reductions 

 on this account. Furthermore, the sky was unusually hazy and 

 streaky on Mount Wilson in 1919, which also added to the labor and 

 anxiety of determining the best values from the observations. 



Agreement of Mount Wilson and Chilean work. — However, the re- 

 sults when finally worked out proved to agree excllently, except for 

 the systematic error above mentioned, with the results obtained in 

 Chile. Both stations showed simultaneous and nearly equal fluctua- 

 tions of solar radiation through a range of about 5 per cent. After 

 allowing for the aforesaid 2 per cent systematic error of Mount Wil- 

 son, the average deviation of the two stations was but 0.013 calorie, 

 or 0.65 per cent from all the values, about 50 in number, obtained on 

 corresponding days. Omitting five values very discordant, when 

 the Mount Wilson sky was very hazy and streaky, the average devia- 

 tion of the remaining days was about 0.008 calorie, or 0.4 per cent. 



Solar variation confirmed by observations of Saturn. — From corre- 

 spondence with Dr. Guthnick, of the Berlin-Babelsberg Observatory, 

 a most interesting confirmation of the solar variability has appeared. 

 Variations of brightness of the planet Saturn from January to May, 



1920, were shown by Dr. Guthnick's photo-electric observations which 

 could not be accounted for after allowance for all known sources of 

 variability. These outstanding variations were found to be in almost 

 exact correlation with fluctuations of the solar radiation as observed 

 at Calama, Chile. One per cent increase in solar radiation was found 

 to accompany 1 per cent increase of Saturn's brightness. 



These results, however, were only derived in connection with one of 

 two possible interpretations of the nature of solar variation. The 

 sun might vary in such a manner that its changes would be observed 

 simultaneously in all directions and so would occur on identical days 

 on all the planets. This hypothesis does not fit the available observa- 

 tions of the sun and Saturn. On the other hand, the solar radia- 

 tion may be unequal in different directions. Such inequalities are, 

 in fact, indicated by the ragged raylike structure of the solar corona. 

 On this hypothesis a change of solar radiation would occur as ray 

 after ray strikes the earth in the course of the sun's rotation upon 

 its axis. These same unequally intense rays would reach the planet 

 Saturn either before or after they reached the earth, according to 

 the relative heliocentric longitudes of the ea,rth and Saturn. The 

 sun rotates about 14° a day, so that the angular difference in posi- 

 tion of the two planets is to be divided by 14° to indicate the number 

 of days allowance to be made between the dates of corresponding 

 solar and Saturnian measurements. 



Proceeding on this second hypothesis, extraordinarily close corre- 

 spondence between the variations of the sun and Saturn was found. 



