NO. 3 EFFECT OF SOLAR RADIATION CLAYTON 1 5 



curve shows a slight minimum about October 5, not indicated by the 

 solar curve. The St. John's curve, instead of showing the greatest 

 minus variation about October 11, as indicated by the solar curve 

 shows the greatest minus variation about November 3. The Stykkis- 

 holm curve follows the solar variation until about October 6, after 

 which it becomes inverted as indicated by the light broken curve. 

 The Sacramento curve follows the solar radiations in an inverse 

 sense until October 10, after which it varies directly with the solar 

 radiation as indicated by the light broken curve. The curve for 

 Winnipeg is inverted throughout, but shows a tendency to a direct 

 correlation about October 24. 



These diverse effects appear to be associated in some way with 

 shifts in the centers of action in the atmosphere, as for example, the 

 shift of the anticyclonic center in the Atlantic and Pacific oceans, 

 and that of the low pressure centers near Iceland and the Aleutian 

 Islands. 



In order to study the changes accompanying each individual 

 maximum of solar radiation A, B, C, and D, figure 5, the deviations 

 of temperature accompanying each solar maximum were plotted on 

 charts. The dates of the solar maxima were taken as September 19, 

 October 2, October 19 and November 8. Supposing that in the 

 average for the whole world the greatest departures of temperatures 

 would follow these solar maxima about three days the dates were 

 taken as September 22, October 5, October 22 and November 11. 



The deviation of the five-day means from monthly means on these 

 dates for the various stages of the world were plotted on maps. The 

 five-day means were for the day of observation and the two days 

 following and the two days preceding the date. The monthly means 

 were for the 15 days preceding and 15 days following the given dates. 



In figure 5 the zero lines represent the 30-day mean and the plotted 

 lines the deviations from this mean. It will be seen from this chart 

 that on October 22, the departure at San Isidro, Manila, was -I-0.5, 

 while at Stykkisholm it was —3.7. 



After plotting this data on maps, lines were drawn separating the 

 plus from the minus areas and also lines were drawn around the 

 stations showing maximum departures. 



An inspection of the charts 3, 4, 5, and 6 reveals the fact that the 

 centers of maximum departure are not fixed but show a tendency to 

 shift from point to point on the earth's surface. On chart 7 the 

 centers are plotted as well as could be done from the meager data 

 at hand, and they indicate an oscillation of the maxima around the 



