NO. 3 SOLAR VARIATION AND WEATHER — ABBOT 45 



hours at 315. Monthly means of Mrs. Hill's values were subtracted 

 from 315 yielding the results given in table 6. (See fig. 39, p. 48.) 



From these figures the average probable error of the monthly mean 

 departures is ±15 over the years 1944-1957. Combining the de- 

 partures with observed mean monthly values of h'Fa, the observa- 

 tions are cleared of the yearly march and become adapted to show 

 if they indicate relationship to the family of harmonics in solar 

 variation. The result of this comparison is graphed in figure 39 

 and appears in table 7. 



Within the limit set by the probable error, the graph shows sine 

 forms for 7 of the 9 solar periods included. The other 2 are too 

 short and irregular to judge. 



Washington daily precipitation. — The sun's rotation period is 

 approximately 27 days and is almost an exact submultiple of 2 years. 

 The daily rainfall at Washington was tabulated from 1924 to 1941, 

 18 years, in 9 tables of 27 columns each. The 9 means of the 27 

 columns were then tabulated. This disclosed a small departure 

 from exactly 27 days. The exact period of the solar rotation for 

 the purpose in view was found to be 27.0074 days. P. 3765, figure 1, 

 here reproduced as figure 40, shows how the rainfall of Washington 

 was distributed over the period of the sun's rotation during the 

 years 1924-1941. (See p. 49.) 



The curve for years of average rainfall. No. 3, indicates that 

 days 1, 2, 3, 4, 5, 12, 17, 22, 24, 25, 26, and 27 of the sun's rota- 

 tion cycle (based on January 1, 1924) are apt to be more rainy 

 than other days. From 1941 to 1954 I published yearly a pamphlet 

 predicting 175 days of the next following year when precipitation 

 was more likely to occur than on the other 190 (or 191) days. The 

 experiment was successful beyond expectation. Indeed in 1948 no 

 less than 14 brides wrote or phoned inquiring if their wedding day, 

 to come in a month or six weeks, would be fair. I predicted 13 fair 

 days and one probably rainy. As it turned out 14 were fair, but one 

 was overcast without rain. The most outstanding success concerned 

 the hostess in a restaurant. I told her that her day would probably 

 be fair, but there would probably be rain the day before and the 

 day after. That was exactly what happened ! 



Table 6. — Monthly mean departures of h'Fz from 315. 



Jan. Feb. Mar. Apr. May June 



— 69±9 — 27±14 — 27±24 -|-29±30 -t-88±26 -|-89±2S 



July Aug. Sept. Oct. Nov. Dec. 



+93±19 +73±35 0±11 -52±13 -71±10 -73±9 



