222 



SAIITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 70 



year (A). The temperature values have been subjected to a com- 

 bined two- and three-years' smoothing. At the to'p is the curve S for 

 the smoothed mean of the sun spot relative numbers according- to 

 Wolfer. The curves show a considerable difference in temperature 

 variations between summer and winter. The variations are great- 

 est in the winter months. December, January, and February, and go 

 then in great measure (particularly in January) oppositely to the 

 variations of the sun spots. At certain times, as for example, be- 

 tween 1841 and 1853, curves for February and March run almost 

 oppositely to the curve for January and directly with the curve for 



Figure 80. Anomalies of the air temperature in Stockholm for each month 

 (I to XII) and for the whole year (A) in combined two- and three-year 

 smoothing. S: relative numbers of the sun spots (scale on the right). 



sun spots, and this occurs also in part for the curves for April, 

 May, June, and July. In the years of the interval 1864 to about 

 1875 the curve for January and also in a slight degree the curve for 

 December goes partly directly with the sun spot curve, while on the 

 other hand the curve for February runs oppositely. In most years, 

 after 1841, the curve for March runs directly with the sun spot 

 curve. After 1885, the curve for April goes directly with the sun 

 spot curve.' Most curves show a tenedency to the already mentioned 

 double division of the eleven-year period. 



^ Krogness, as stated above, has found the same for the temperature in 

 Norway in the later periods, namely, that in January it goes oppositely to the 

 magnetic storminess in Christiania and in March-April and also in part in 

 ■July it goes directly with it. 



