2 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. Ill 



Table 2 gives the dates for 1950 when the average daily precipita- 

 tion in Washington is expected to exceed the average daily precipita- 

 tion in this city on all other days. In the first column are given in 

 Roman numerals the day numbers of the 27-day cycle when higher 



Table 2. — Predicted dates for the year igso ivhen average daily precipitation 



should exceed average daily precipitation of all other dates of the 



year igso in Washington, D. C. 



"Preferred" 



cycle places Jan. Feb. Mar. Apr. May June 



I II 7 6 2, 29 26 22 



II 12 8 7 3,30 27 23 



III 13 9 8 4 1,28 24 



IV 14 10 9 5 2, 29 25 



V 15 II 10 6 3,30 26 



XII 22 18 17 13 10 6 



XIII 23 19 18 14 II 7 



XV 25 21 20 16 13 9 



XVII 27 23 22 18 15 II 



XVIII 1,28 24 23 19 16 12 



XXII 5 1,28 27 23 20 16 



XXVI 9 5 4, 31 27 24 20 



XXVII 10 6 5 1,28 25 21 



"Preferred" 



cycle places July Aug. Sept. Oct. Nov. Dec. 



I 19 15 II 8 4 I, 28 



II 20 16 12 9 5 2,29 



III 21 17 13 10 6 3,30 



IV 22 18 14 II 7 4,31 



V 23 19 15 12 8 5 



XII 3,30 26 22 19 15 12 



XIII 4, 31 27 23 20 16 13 



XV 6 2, 29 25 22 18 15 



XVII 8 4, 31 26 23 19 16 



XVIII 9 5 1,28 25 21 18 



XXII 13 9 5 2,29 25 22 



XXVI 17 13 9 6 2, 29 26 



XXVII 18 14 10 7 3,30 27 



precipitation is expected. These values arise from the statistical study, 

 1924 to 1941, above mentioned. The other columns give the actual 

 days in the 12 months of 1950 when these Roman cycle dates will 

 occur. In other words the remaining columns give the "preferred" 

 dates for 1950. While it is expected that for the entire year 1950 

 the "preferred" dates will yield higher average precipitation than all 

 others, and even that this will be so for most of the individual months 

 of 1950, the probability that any individual "preferred" day will yield 



