244 



ANNUAL REPORT SMITHSONIAN INSTITUTION, 19 3 9 



500 KILOMETERS 



Figure 1. — rressure map for September 21, 1938, at 7 : 10 a. m., and track of hurricane. 

 (After Tannehill.) Pierce (op. cit., fig. 12) places the center farther east, on the 73d 

 meridian, along which it moved straight northward tUl 7 p. m., when it was in west- 

 central Vermont. 



lowest pressure, even making it go faster than the general current. 

 Moreover, the cool, dry air could not furnish more than half as much 

 latent heat, the lifeblood of a tropical storm, as the warm, moist air 

 had been providing. On the other hand, the 20° F. contrast in tem- 

 perature between the east and west sides of the storm created potential 

 energy that made up in part for the loss of latent heat, so that the 

 rate of decrease in intensity was not rapid,^^ Also, the storm moved 

 northward so fast that the reduced energy had little time to make 

 itself felt. 



The gradual weakening of the storm over New England could have 

 been accounted for altogether by the increased friction of the wind 

 with trees, buildings, and hills of an average roughness of 9 m. (about 

 30 ft.) height, according to computations by Raymond Wexler.® 



The circular pressure field of the cyclone merged, naturally, into 

 the general pressure trough extending northward through New Eng- 

 land. The pressure gradient on both sides steepened rapidly, par- 

 ticularly on the east side, where the stronger high was. The wind 



** See Pierce, op. cit., flg. 40. 



» Wexler, Raymond, Friction and the filling up of the hurricane of 1938 over New 

 England. Bull. Amer. Meteorol. Soc, vol. 20, pp. 277-281, September 1939. 



