THE SHIFTING OF CLIMATIC ZONES. 



201 



we should get a double maximum where previously only a single maximum was found. 

 In the present case the ideal storm track — -that is, the mean of all tracks — may be con- 

 sidered as running in a great arc somewhat southeasterly from a point in western Canada, 

 approximately in latitude 55° N. and longitude 115° W., to the northern part of Lake 

 Superior, eastward across southern Canada, and then with a sUght northerly trend toward 

 Newfoundland. The area of greatest storm frequency is limited to a small region, approxi- 

 mately 45° to 50° north of the equator. Southward the number of storms decreases gradu- 



^•^Aplproximate .. .... ..^ _.,._.,_, _ . -.. 



l(r~A*pcaxmiate center of storm freq uency 1899-190! 

 A^proxirTiate-|fn6an &tarnjjtrack ISSSr '^"8 



Fig. 64.— Storm Frequency. Year Maps for 1878-1887, after Dunwoody, and 1899-1908, after Kullmer, showing Shift of Storm Track. 



ally; northward the decrease is very rapid. A southerly sliift of the storm track of this 

 region would cause a decrease in storm frequency in the squares just north of the center 

 of the curve, which is seen to be the case in the map for the years 1899-1908 as compared 

 with the earlier one. At the same time the southward movement of the storms would 

 produce increased storm frequency south of the former mean storm track, which is actually 

 the case, as appears from the increase of 10.3 in the New York square. Under such circum- 

 stances a double maximum might perhaps be produced in this fashion. Suppose that in 

 the earlier of our two periods the curve of the mean storm track were tangent to the parallel 

 of latitude 47.5° N. which, of course, is the mean latitude of the tier of squares 45° to 50° N. 

 Being tangent to this parallel at the most southerly swing of the track and also at the point 



