CYCLONIC STORMS IN TEMPERATURE LATITUDES 511 



a division into six groups. When this is done, the table appears as 



follows : 



Table 5 



Per cent 



7 years, with sun-spot numbers above 60 (average, 70.4) . . .storminess 114.1 



8 years, with sun-spot numbers from 35 to 60 (average, 46.8) .storminess 101.6 



7 years, with sun-spot numbers from 10 to 35 (average, 21.1) .storminess 99.3 



8 years, with sun-spot numbers below 10 (average 5.7) storminess 85.4 



The similarity between these figures and those for tropical hurricanes 

 in Table 2, on page 494, is striking. It seems impossible to avoid the 

 conclusion that the same phenomena are repeated on both the northern 

 and southern borders of the region where cyclonic storms prevail. The 

 maps which we have discussed above make it equally clear that in the 

 intervening region there is a corresponding decrease of storminess. 



Before leaving this subject let us take the area of maximum change 

 of storminess — ^which is the area included within the heavy lines in fig- 

 ure 3, and which is also the area of maximum storm frequency — and 

 compare it with the sun-spots year by year. In this comparison I shall 

 include the early years from 1874 to 1882, as well as the later and more 

 reliable years when the data cover the whole country. These earlier years 

 are probably more or less closely comparable with one another, although 

 the first years are less reliable than the later ones, and there is a slight 

 break between 1882 and 1883. Moreover, as already implied, the years 

 1883 to 1885 may not be so reliable as those that follow. The results of 

 the comparison are given in figure 9. The upper heavy line represents the 

 variations in the number of storms in the American area of maximum 

 storminess. From 1876 to 189 la dotted line has been added, indicating 

 the combined storminess of the American area and of a similar area in 

 Europe, but the European area is less well defined than the American and 

 the data are of less value. Confining our attention to the American 

 curve, we see that for the seven years from 1874 to 1880, inclusive, the 

 curve of storms agrees closely with that of sun-spots, which appears below 

 it. Even the little maximum of 1877 appears in both curves. In 1881, 

 however, the amount of storminess drops suddenly, although the number 

 of sun-spots keeps on increasing. Then for four years the two curves 

 again agree, for both approach a maximum in 1884 and begin to fall in 

 1885. The next year the storms again behave differently from the sun- 

 spots, but this does not continue long, for in 1887 and 1888 both curves 

 again decline. The following year the storms increase in number, while 

 the sun-spots remain stationary. If the two sets of phenomena are really 

 due to the same cause, the delay of the sun-spot curve may possibly be 

 due to the fact that the atmosphere of the earth is much smaller and per- 



