CYCLONIC STORMS IN TEMPERATURE LATITUDES 507 



average number of storm tracks for a single year and not for groups of 

 years^ as in the other maps. 



The most striking fact about these maps is that all of them show the 

 same general features^ and that although the location of the features 

 varies somewhat from map to map^ the variation is so slight that all the 

 features of the individual maps reappear in a somewhat generalized form 

 in the combination map of figure 7. The most noticeable feature is an arc- 

 shaped area which lies along the southern border of Canada and is char- 

 acterized by increased storrniness during periods of many sun-spots. It 

 begins in latitude 55° north on the west, drops down to about latitude 

 48° in the Great Lakes region, and then bends northward once more 

 down the Saint Lawrence Valley. A second prominent feature is a pe- 

 culiar projection which on each of the maps extends southward from the 

 center of the area of excess. It almost divides into two parts the next 

 prominent feature, which is a pronounced band of decreased storminess 

 extending from Oregon through the central United States to Nova Scotia. 

 South of this we have a second great area of increased storminess during 

 times of many sun-spots. Although slightly broken in some of the maps 

 of individual groups of years, it forms a continuous band in the combi- 

 nation map of figure 7. It is particularly well developed over the arid 

 regions from California to western Texas and over the Atlantic Ocean. 

 In one map, however, it becomes intensified in southern Texas. Finally, 

 in the far south we find that in times of many sun-spots there is a slight 

 suggestion of another little belt where storms decrease during times of 

 many spots. This, however, is not well enough developed to be given 

 much importance. In all these cases it should be understood that the 

 words "increased^^ and "decreased" are used in a purely relative sense. 

 In the northern belt of increased storminess the number of storms is 

 always great, both at times of few sun-spots and of many. In the south- 

 ern belt, on the contrary, the storminess is never very great, although at 

 times of many spots it is distinctly greater than at times of few. I shall 

 deal with this matter more fully later. 



As a further test of the relation of storms and sun-spots Kullmer has 

 constructed still another map, figure 8. In this case, instead of taking 

 merely years of maximum and minimum sun-spot frequency, he has com- 

 pared two complete sun-spot cycles with one another. Unfortunately, 

 only two cycles are as yet available, but they are enough to provide an 

 interesting test. One cycle consists of the 12 years from 1899 to 1900, 

 inclusive, and the other of the 11 years from 1900 to 1911. In the first 

 cycle the average sun-spot number was 38.8 and the average storminess 

 107.4 per cent. In the second the average sun-spot number was 33.3 and 



