50 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. ']'] 



shown later, there is an increased solar radiation when the faculse 

 are in that position. Simultaneously, or nearly simultaneously with 

 the occurrences of these maxima of solar radiation, there are maxima 

 of pressure and minima of temperature at Winnipeg, while with the 

 minimum of solar radiation, when the spot is near the central meridian 

 of the sun, there is a minimum of pressure and a maximum of 

 temperature at Winnipeg. From this central continental region in 

 which Winnipeg is located the pressure waves progress southward 

 and eastward, reaching stations like New York about three to four 

 days later. 



The reasons for the maxima C and D are not so evident, but 

 as mentioned before, they are probably related to periodic or semi- 

 periodic changes within the solar mass, especially as they are found 

 in almost every class of solar changes. 



Departures from normal temperature were next obtained and 

 studied by seasons for several stations in the United States and 

 Canada. Dividing the year into four parts, the first three months, 

 January to March, were called Winter, the next three months, April 

 to June, were called Spring, the three months, July to September, 

 Summer, and the three months, October to December, Autumn. 



The mean temperature departures by seasons for the dates pre- 

 ceding and following the central passage of sun spots are given in 

 table 15 for Winnipeg,. Chicago, New York, and Father Point. When 

 thus subdivided, the number of large sun spots was not sufificient to 

 form satisfactory means, so that all observed sun spots were used. 

 The number of cases were : Winter, 20 ; Spring, 30 ; Summer, 25 ; 

 Autumn, 25. The results for the four divisions of the year at 

 Winnipeg from seven days before to ten days after the central 

 passage of the spots is plotted on the left hand side of figure 39. The 

 results for the four seasons show a striking resemblance, and thus 

 furnish proof of the existence of short-period changes in the sun 

 dependent on the position of sun spots. Apparently, however, the 

 effect comes somewhat later in summer than in winter. This delay 

 is more evident when the mean of several stations is taken. Allowing 

 one day for the drift of atmospheric changes from Winnipeg to 

 Chicago, and another day to Father Point, the mean results for the 

 three stations were obtained from the data in table 15, and are plotted 

 on the right-hand side of figure 39 for the interval from five days 

 before to eleven days after the central passage of the sun spots. A 

 distinct seasonal lag is here indicated. The maxima and minima 

 occur about one day earlier in winter than in spring and autumn. 



