2 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. IO7 



ings for the precipitation tables identical with those I used in collecting 

 temperature effects. 



Hence in figure i of the present paper the reader will find the date 

 figures, which I have written small, not all in the column of zeroth 

 day. They are found as follows: In the years 1931 and 1933 in 

 column +2; in the years 1935 and 1936 in column +i ; in the year 

 1937 in column — i ; in the year 1932 in column —2; in the year 1934 

 in column —3. These shiftings are those idicated by tables 3 and 4 of 

 my former paper. Without these phase shiftings the present results 

 would be confused. 



-4 -3 -2 -I 



2. 3 



9 to It 12 (3 14- 



CicO 



336 ^18 Ay4 4(S- S79 43? 476 416 ifeO /09 4Jl 383 6«6 'ir34f4 /W 417 ;i?3 ^a 4«8 

 'a«« cwso.loOAiy/ o.<i?i" CMlo.iwe.erjxe.l'K, <).(3Z*1«9 alt? 0.1^7 «.*»" ej47a(fli A(4io.f*« 



a.«6o 



Fig. I. — Computation of average effect, 1931 to 1937, of the 6.6456-day solar 

 period on Washington precipitation for the month of March. 



In figure i of the present paper, the top line indicates days from 

 — 5 to 4-14, with respect to zeroth day indicated basically in table i 

 of my former paper. The next following 29 lines of figure i give for 

 20 days each the total precipitation, midnight to midnight, in hun- 

 dredths of an inch, as published by the Weather Bureau for March 

 1 93 1 to 1937. These data fix the form of the four or five returns of 

 the 6.6456-day period which occurred in March of each year — thus in 

 1931 four, in 1932 five, et cetera. The next to last line of figure i 

 gives the sums for each column, and the last line the average daily 

 precipitation in inches for each of the 20 columns. At the right is a 

 scale of inches, and in the body of the table a graph gives the results 

 as written in the lowest line. 



