NO. 9 



PRECIPITATION AND SOLAR VARIATION — ABBOT 



It will be noticed that the maxima are not exactly spaced at in- 

 tervals of 6.6456 days, but they are approximately. The minima fall 

 as closely to the correct intervals as could be, without dividing the 

 days into hours. The range is from 0.040 to 0.204 inch, more than 

 fivefold. 



Having illustrated the procedure, I give in table i a summary of 

 the places where the maxima were found in tabulations similar to 

 figure I, but covering, for each month of the year, all years from 1924 

 to 1945. This summary is entirely parallel to table 3 of my former 

 paper. It was undertaken to correct for phase shiftings, before taking 

 the general means of the effects. It seemed by no means to be as- 



Table I. — Phases of maxima in the 12 months, and phase corrections 



sumed, without testing, that the lag of the effects of the solar change 

 on precipitation would be identical for different seasons of the year. 

 Table i, however, gives little ground to conclude that there are dif- 

 ferent lags in different seasons. The phase shiftings determined in 

 the present table i do not exceed ± i day. These might easily have 

 resulted from the facts (a) that no account is made of hours in the 

 records; (b) that tabulations, similar to table 3 of my former paper, 

 include frequent cases where a computed phase shifting might have 

 been altered i day by a difference in judgment as to the position of 

 centers of features of graphs. In order to get the best determination 

 of the amplitudes of precipitation effects, I used the phase shiftings 

 indicated in the present table i, when collecting the results of the 12 

 months of the year. 



I give in table 2 the collected results of the precipitation in the 



