CYCLONIC STORMS IN TEMPERATURE LATITUDES 497 



are relatively cool and the summers are not apt to show markedly more 

 precipitation than the winters. In view of this, then, we may interpret 

 the curves of figure 2 as meaning that when sun-spots are numerous, rela- 

 tively continental conditions of climate prevail in northern Germany. 

 This carries with it the implication that at such times the -continental 

 areas of high pressure tend to become intensified in winter, so that the 

 air blows outward from them and cyclonic storms are compelled to move 

 along the margins of the continent rather than toward its interior. In 

 summer, on the contrary, the low-pressure areas of the center of Eurasia 

 appear to become intensified, and this causes the winds to blow toward 

 the interior and to bring abundant moisture. The full importance of 

 this will appear later, when we come to discuss changes of climate during 

 tlie past two or three thousand years. 



Cijclonic storms in iemperate latitudes — Eesearches of Professor Kull- 

 mer. — We now come to much the strongest type of evidence as to cyclonic 

 storms and their relation to sun-spots. Prof. Charles J. Kullmer,^^^ of 

 the University of Syracuse, has made a careful investigation of the matter, 

 both in the United States and Europe. His first work appears as a chap- 

 ter in "The Climatic Factor/' but since that was published he has gone 

 into the matter much more thoroughly. From the maps of the United 

 States Weather Bureau he has compiled monthly and yearly data from 

 1874 to 1912. Omitting the earlier years, where the degree of accuracy 

 is not great, he has found the average for the 30 years from 1883 to 1912 

 and has thus prepared the accompanying map, figure 3. His method was 

 to divide the United States and Canada into a series of rectangles, each 

 of which extends 21^° in latitude and 5° in longitude. He then counted 

 the number of storm tracks passing through each rectangle during each 

 month of the year. The total for the year gives the relative storminess. 

 The figures for the various squares are the average values for 30 years. 

 The map shows a pronounced bow-shaped area of great storminess swing- 

 ing down from British Columbia into the northern United States, across 

 the Great Lakes and down the Saint Lawrence Valley. A minor area of 

 relatively high storminess is found in southeastern Colorado. The num- 

 ber of storms in one area as compared with another is not a measure of 



i" In the remainder of this paper the name of Professor Kullraer will recur again and 

 again. W^hen the paper was originally written it was expected that before its publica- 

 tion Professor Kullmer's data would all have been published, but various unforeseen 

 circumstances have prevented this. His results were given to the world, however, at 

 the meeting of the Association of American Geographers at Princeton in January. 1014. 

 Both before and after that meeting Professor Kullmer most kindly permitted me to 

 make the fullest use of all his material. Now, with rare courtesy, he has permitted me 

 to publish a large number of his maps and other data prior to the appearance of his 

 own report. I can not too strongly express my appreciation of his great kindness. 



