CYCLONIC STORMS IN TEMPERATURE LATITUDES 503 



probably of high importance, but neither its importance nor its verity 

 can rival that of a most striking discovery made by Knllmer. He has 

 ascertained that there is a definite law as to the shifting of storm tracks 

 in response to changes in sun-spots. The American data, as has already 

 been stated, cover three successive periods during which a group of three 

 years of minimum spots may be compared with a similar and succeeding 

 group of maximum spots. The periods are as follows : 



(1) The minimum years of 1877, 1878, and 1879, compared with the 

 maximum years of 1882, 1883, and 1884. 



(2) The minimum years of 1888, 1889, and 1890, compared with the 

 maximum years of 1892, 1893, and 1894. 



(3) The minimum years of 1900, 1901, and 1902, compared with the 

 maximum years of 1905, 1906, and 1907. 



We may] also make a fourth comparison by using the last sun-spot 

 maximum and comparing it with the following instead of the preceding 

 minimum. Thus : 



(4) The maximum of 1905, 1906, and 1907, compared with the mini- 

 mum of 1910, 1911, and 1912. 



For each of these periods — that is, three maxima and four minima — 

 Kullmer has found the total number of storms, and has thus obtained the 

 figures for a series of maps similar to figure 3, except that each represents 

 the average of three years instead of 30. He has then taken the map of 

 a minimum series of years and that of the succeeding (or in the last case 

 the preceding) maximum series, and has compared the two, square by 

 square, to see to what extent the maximum years exceed or fall short of 

 the minimum. The figures thus obtained have been plotted in another 

 series of maps, two of which are given in figures 5 and 6. These repre- 

 sent the two groups of years marked (2) and (3) above — that is, the two 

 which are most perfect and which do not overlap. The other two maps 

 show the same features as these two and might equally well have been 

 used. In the maps the plus sign means that the three maximum years 

 had more storms than the minimum years, while the minus sign means 

 that storminess was greater when the sun-spots were at a minimum. The 

 numbers show the totals for three years, and must be divided by three in 

 order to obtain the average for a single year. Lines have been added at 

 intervals of five to bring out the essential features. Solid lines indicate 

 that when the sun-spots were at a maximum the storminess was greater 

 than when the spots were at a minimum, while dotted lines indicate the 

 reverse. In figure 7 I have combined all four of Kullmer^s maps into a 

 single map, which shows the average difference between all the available 

 years of sun-spot maxima and minima. Here the numbers indicate the 



