HURRICANES INTO NEW ENGLAND — BROOKS 243 



ORIGIN OF THE STORM 



Tlie hurricane that devastated Long Island and New England on 

 September 21, 1938, had its origin far east of Puerto Rico. On the 

 morning of the 16th it was already of full hurricane strength; in 

 fact, as early as the 13th a cyclonic circulation was in evidence in 

 37° W., 19° N/ This whirl was almost certainly that which formed 

 on the night of the 7th over French West Africa, lat. 10°-20° N., 

 long. 10°-20° W., between the northeast trade, the southwest monsoon, 

 and the equatorial easterlies. The disturbance initiating the whirl 

 formed in the interior, however, and moved westward to the coast. 

 It was first noted as a weak low at Bilma oasis, in the south-central 

 Sahara, on September 4.^ 



The direction (W. by N.) and speed (15 to 20 miles an hour) of 

 progression of the hurricane as a whole were practically the same as 

 the gradient wind in the general pressure field at the southern and 

 western margins of the North Atlantic high. This high, centered 

 south of Newfoundland, was attended by a broad stream of tropical 

 air, which curved northward east of Florida. The hurricane moved 

 along in this stream more or less as would have a floating balloon. 

 Its entry into the stream may be considered as essentially accidental. 

 But once in the stream, the effect of the hurricane was marked. The 

 storm became a great whirl that sucked air into itself from a belt 

 about 300 miles wide and discharged it upward, thereby putting it 

 through the cyclonic "wringer." Within the Tropics this process did 

 not alter the warm, moist character of the tropical air stream — the 

 supply of moist tropical air was ample for long distances on each side. 



Once the storm left the Tropics, however, it no longer had an un- 

 limited supply of tropical air on its left. Polar air reached the 

 southeastern States as the hurricane recurved. In fact, the arrival 

 of this air probably helped the recurving; for its lower temperature 

 increased the density of the atmosphere and thereby favored a slope 

 in the upper-air pressure surfaces from east to west, a direction of 

 slope that calls for a northward movement of the upper air. 



The drawing of cool, dense air into the storm from the northwest 

 and west favored a more and more rapid rise in pressure south of it 

 as the incoming air got cooler and cooler with higher and higher 

 latitudes — pressure at Hartford, for example, fell 0.4 inch one hour 

 and rose 0.75 inch the next. This raising of the pressure on the south 

 should have accelerated the northward movement of the center of 



'Tannehill, I. R., Hurricane of September 16 to 22. Monthly Weather Rev., vol. 66 

 pp. 286-288, 2 maps, September 19^8. 



* Hubert, H., Sur I'origlne afrlcaine d'un cyclone tropical dgvastatenr dans la region de 

 New York. Comp. Rend. Acad. Scl., Paris, vol. 208, pp. 456-457, 1939; Origlne Afrlcaine 

 d'un cyclone tropical Atlantique. Ann. Phys. Globe France d'Outre-Mer, vol. 6, pp. 97- 

 115, 17 figs., Paris, 1939. Reviewed In Geogr. Rev., New York, 1940. 



197855 — 40 17 



