ON PERIODICITY IN SOLAR VARIATION 

 By C. G. ABBOT 



Greatly interested by the paper of H. H. Clayton, 1 I directed Mr. 

 Eisinger to make the necessary computations to determine by Clay- 

 ton's method whether there occurred periodicities in the short inter- 

 val solar variations in other years than 191 3. I refer to those vari- 

 ations discovered by the Smithsonian Astrophysical Observatory, 

 which often seem to run irregular courses of a week or ten days 

 between maxima. Clayton's method is applied as follows : 



All consecutive days are written down in a column from one end 

 to the other of the observing season of each year. Opposite these 

 days are written in a second column the corresponding values of the 

 " solar constant " of radiation determined on Mount Wilson. As the 

 observations are lacking on some days, vacancies exist in this col- 

 umn. In a third column the same " solar constant " values are written 

 down, but raised one day on the scale of time. In succeeding columns 

 up to 40 in all, the same "solar constant" values are written down, 

 but each column is raised one day's interval as compared with the one 

 before. Thus as we look along from column to column the values 

 are so arranged horizontally that we compare the " solar constant " 

 of each day with those of one, two, and subsequent days to forty 

 days later. Owing to the lack of observations, not every day's value 

 is thus compared with all the values of later days up to forty, but each 

 day enters into some at least of these comparisons. 



The observations being thus arranged, the usual computations are 

 gone through with for obtaining coefficients of correlation between 

 the " solar constant " of given days and those of 1 day, 2 days, and 

 other intervals later. In selecting the groups required in correlation 

 computations, the observations have been separated within ranges 

 of 0.02 calories. To avoid giving undue weight to " wild " values, 

 such as are probably affected by progressive obscuring or clearing of 

 the atmosphere, all values over a certain reasonable maximum or 

 under a certain reasonable minimum are put in with the highest and 

 lowest 0.02 calory groups, and are regarded as falling in these ranges. 

 Such high and low " wild " values seldom number more than 3 or 4 

 in a season. 



1 Smithsonian Misc. Coll., Vol. 68, No. 3, 1917. 



Smithsonian Miscellaneous Collections Vol. 69, No. 6 



