SOLAR VARIATION AND WEATHER—ABBOT 141 
precipitation. I then divided the data into three sections representa- 
tive of dry years, 54 cycles; intermediate, 108 cycles; wet years, 81 
cycles. The results are given in figure 13. 
A VERIFIED PREDICTION ONE YEAR IN ADVANCE 
In March 1943 I informed the Chief of the Weather Bureau that on 
a certain list of dates the average daily precipitation would be higher 
than on the remaining dates of the year. I recently tabulated the re- 
sults: Using curve 3, applicable to years of intermediate precipitation, 
the selected dates of 1943 were expected to show 166 times the average 
aw om ms sraeeenaecea as 
rh T 
net 
6 ams t ms 
4 spat B'9 on rH 
at ib; of EH ath . 
4 dana {+ +t 
ue aa + 
“i eanageeeeaa 4 eee 
REE EN : aH HA 
a =! A PATS 
wap was Hee ease: 
2 a awaeren ro 
; = zs 
0 suusesy tet oF : 
4. + 3 4 - aa 
speuieeee' : 
+H gegsee 
2 Ht + I +-| 
gauaeeenee 
c] oorem 5 = 
ac +. 4 i ct 
2 A + 
tH £ 
ues ty we 
aa Si 
fe) 5 aanT ee Et 
abe uae shape { t 
z snaeee neo a 
2 Seupaeewe! sana 
h nage 
is fo t cot i 
if E Saesavavazedasasascfapuererasessratesieceusnessaae 
oO SEEEEE EEE eae 
/ 3 ey 7 } 4 43 15 /7 19 
Figure 18.—27.0074-day period in Washington precipitation. Curve 1, general 
mean, 248 cycles; curve 2, dry years, 54 cycles; curve 3, intermediate, 108 cycles; 
curve 4, wet years, 81 cycles. 
rainfall of the nonselected dates. The actual ratio, for the 175 selected 
dates compared to 191 nonselected (the work included December 31, 
1942) was 1.58.2. The 27-day cycle has continued so consistently for 
20 years at Washington that one is inclined to think it may be trusted 
to hold for some years to come. 
MONTHLY MEAN SOLAR CONSTANTS 
We will now consider monthly mean values of the solar constant 
of radiation, the variations they disclose, the periodicities found 
us I might add that the 2 days of large rainfall in January 1944 fell on selected dates 
also. 
