SOLAR VARIATION AND WEATHER—ABBOT 139 
vals of nearly 20 days, may give hope that a method of forecasting for 
many days in advance may be evolved therefrom. I have, indeed, made 
a preliminary test of this possibility. 
It will be apparent that after computing basic curves of the tempera- 
ture effects of solar variation for a given station, it may be assumed that 
when a sequence of rise of the solar constant is descried in the daily 
observations, one may write down in a column for some 2 weeks there- 
after the departures in temperature expected to follow this sequence of 
rising solar activity. As other sequences occur, some rising, some 
falling, other parallel columns of expected temperature departures are 
written down on the proper dates, appropriate to each. 
i] 243] 4|5]6]7]8 [9 opin [re] sig pee 242: ee 1|2|3|4 siclaialaiie We Jia lagi este 7 [refi] ~ 
Observed. ete 
: EEREREE 
; Ss SU EUANOTUONGEUSEOUUESUREUENE 
LUTTE nT 4 THER 
CEN AAEM PTT TT 
SS mETALCeAn AMT TIFELSTEAMA TE HIPAIASAELANGL 
CUVEE PAGE eT 
a aU VANONAVGNOGN/ CAMARLOW//ONEAOAIAEREAPAOTASUNOIE | 
y Sf TAANAMEMEHALATRTRDRER AE ALOEA HUB ELOAAREREREREE 
TTT AEE: 
EEE NE 
4shOhse fet, +429 SAAIHOTOWANOWOHOVGVOVGNOO0 (“OTOOTGKOREE ; 
| sf HT EEE CePeteeeeeerteeil | 
ane per ature Departures Washin gton. 
o[qjObserved - -20--30 
eH! 
Pceescn 12.—Forecast and verification of. Washington temperature departures. 
By the summation of all of these columns day by day, one finds an 
expression of the total influence of solar variation. This summation 
may go on continuously, always for as much as 10 days, in advance of 
the calendar. Figure 12 gives such a summation for September and 
October 1935, prepared from solar-constant basic curves for Wash- 
ington and Ebro dates. 
Unfortunately the solar-constant daily values of first-class quality 
are too scattered as yet, with only our two first-class stations observing. 
However, I have found several months in the long record of calcium- 
flocculi measurements kept at Ebro when the breaks were so rare 
