The ’38 Hurricane


A favorite story on Long Island concerns a man at Westhampton Beach who received in the mail on September 21, 1938, a barometer purchased a tew days earlier in New York. He found the instrument’s needle pointing down near 28 degrees, at the section of the dial marked “Tornadoes and Hurricanes.” He shook the barometer and banged it with his fist, but the needle refused to move, so he rewrapped it, enclosed a note of complaint, and carried it to the village post office. Soon after he mailed it, his shorefront house was demolished—by hurricane.

It is unlikely that such a barometric warning would be disregarded today on Long Island or along the New England coast, where there have been several hurricanes since 1938. But in that year such storms were almost unknown in the Northeast, and the Weather Bureau didn’t bother much about tracking coastal disturbances north of Virginia, assuming that storms moving up from the Caribbean would blow out to sea at Cape Hatteras, as most of them indeed do.

The hurricane of 1938 was first sighted by a ship in the Atlantic 350 miles northeast of Puerto Rico at 9:30 on the evening of September 16. The turbulence was moving at about twenty miles an hour toward Miami, but on the morning of the nineteenth it began to turn away from Florida, eventually bearing due north toward Cape Hatteras.

A story in the New York Times on the morning of Wednesday the twenty-first described “thousands of relieved residents in South Florida” taking down barricades from before their houses and stores. Quoting the weather bureau in Jacksonville, the story said the hurricane was “turning on a northward arc” and “apparently” was heading out to sea. The Times buried the article on page 27—but editorially praised the “admirably organized meteorological service” of the federal government that enabled “New York and the rest of the world [to] have been so well informed about the [Florida] cyclone.”

The Times ’s prediction (made Tuesday night) for Wednesday’s weather in Massachusetts, Connecticut, and Rhode Island was merely for “rain, probably heavy today and tomorrow, cooler.” The coastal forecast from Eastport, Maine, to New York City was for “fresh southerly winds except fresh north or northeast near Sandy Hook, increasing this afternoon or tonight and overcast with rain.”

Thus the great hurricane of ’38 caught the weather bureaus in New York and Washington completely off balance when, instead of turning eastward oft Hatteras, it roared straight up the coast at an incredible forward speed of nearly sixty miles an hour. The storm cylinder was about 150 miles wide and carried gusts of over 150 miles an hour. Shortly before four in the afternoon of the twenty-first the hurricane hit Long Island. The savage winds and towering waters then boomed into the Connecticut and Rhode Island shore lines with enough force to register on a seismograph in Silka, Alaska.

A hurricane gets its energy from sucking up warm moisture in a chimneylike whirlwind. Usually, after leaving the ocean and passing over land it dies down because the supply of charging wetness peters out. But New England had received heavy rains for several days prior to the twenty-first. The land was warm and saturated; therefore, the hurricane’s force was undiminished as it swept north for five hours after leaving Long Island Sound. The storm tore across the Connecticut River valley, up through Massachusetts, Vermont, and New Hampshire (salt spray from the ocean whitening windows en route), and on into Canada, spreading the worst disaster in the history of New England.

Yet even after noon on the day of the blow, weather personnel were still predicting that it would move off to the northeast. Shortly after one o’clock there were gale warnings on Long Island. Ernest S. Clowes, a newspaper editor at Bridgehanipton, later noted that “in none of these warnings was the word hurricane used… it is doubtful if anyone, the Coast Guard included, considered that anything really dangerous was at hand.” In an advisory report issued at 2 P.M.—with the hurricane less than two hours from Long Island- the Weather Bureau said only that the tropical storm moving north-northeastward seventy-five miles east-southeast of Atlantic City would be “likely” to pass “over Long Island and Connecticut late this afternoon or early tonight attended by. shifting gales.”

Nowadays it is easy enough to show that the Weather Bureau should have been able from its own reports to figure that the hurricane was apt to move straight up from Cape Hatteras. It said on Tuesday night that there was a solid high-pressure area extending from Nova Scotia and Newfoundland southward over the ocean, and that another high existed over the Midwest. High-pressure zones tend to repidse lows—like hurricanes. Between the two highs, as the bureau noted Tuesday night, there was a broad trough of warm, humid low pressure extending southward from New England to Cape Hatteras. That trough was a red carpet; there was simply no place else for the storm to go.