by Lee Sandlin
The book was published not by the Signal Corps but by a commercial publisher, and the corps took no public stand on the book one way or the other. But the next year, Greely published a book of his own, titled American Weather, designed to puff the work of the Signal Corps and its continuing importance to American life. He went out of his way to praise Finley’s tornado work. “No other person,” he wrote, “can be accorded greater credit for collecting and arranging data respecting these storms.”
But this was just another empty gesture. As it happened, Finley was just then engaged in the most impassioned fight about tornadoes he’d ever had, and Greely was about to side with his opponent—somebody Finley worked with every day in the corps’s headquarters.
The late general Hazen had been a man passionately devoted to his family—not just his immediate family, but a whole tangled tree of first cousins, second cousins, great-aunts, and great-uncles by marriage. This is how it happened that he had brought into the Washington City office a distant cousin of his named Henry A. Hazen.
Henry Hazen had been born in India, the son of Congregationalist missionaries, but educated in America and sent by his relatives to Dartmouth College. After graduation, he had been appointed assistant professor of meteorology at Yale, under Elias Loomis. He had come to the attention of his august cousin the general early in the 1880s, just as the Study Room was first being set up. The general had sent his chief civilian forecaster, Cleveland Abbe, on a scouting expedition for meteorological talent. Abbe had thought of approaching the famous and well-regarded Loomis, but he then heard through the academic grapevine that Loomis’s work was really being done by his assistant Hazen. Abbe interviewed Hazen and quickly concluded he would be the better hire. The general was only too happy to accept Abbe’s recommendation and put his young relative on the Signal Corps payroll.
Professor Hazen proved to have a vigorous and original mind. At the Signal Corps he did work on the design of thermometers; he studied the microclimates of urban environments like Chicago; he researched ways of using railroad lines to deduce altitude above sea level. He often gave volunteer lectures on meteorology at local colleges. He took on the editorship of the corps’s Monthly Weather Review. He published papers on sunspots, on the possibility of lunar influences on the weather, on methods of calculating flood stages of rivers, on the calibration of anemometers, and on the reason for the spectacular sunsets seen all over the world after the eruption of Krakatoa. He was also a passionate advocate for meteorological exploration of the upper air and several times made balloon ascensions to take readings—this was when ballooning was still considered a fairground stunt.
Among the duties that Professor Hazen ultimately took on was assisting John Finley in tornado research. But this proved to be a disaster. The two men were a poor match from the start. Finley was bluff, clumsy, dogged, and overbearing; Professor Hazen was brilliant, cultured, aristocratic, and haughty. Finley’s main, if not his only, interest was meteorology; Hazen had many wide-ranging interests, but his deepest enthusiasm was for the genealogy of the Hazen family, which he discussed continually and at length. Hazen did not trouble himself to be anything other than contemptuous toward Finley, and soon they found themselves at war.
Their first skirmish was over a narrowly technical issue about tornadoes. Professor Hazen published a letter in The American Meteorological Journal claiming priority for a key principle of tornado formation. Lieutenant Finley, Hazen noted, had written in one of his papers that tornadoes formed in the southwest quadrant of a large region of low pressure. Professor Hazen himself, shortly beforehand, had been quoted in an article in The Washington Post identifying the quadrant as the southeast. He therefore deserved credit for being the first to identify the correct quadrant.
Finley found this claim petty and infuriating—and factually incorrect. “Southwest” had plainly been a transcription error, one of many that dogged that particular paper. He could prove it: several of his earlier publications had specifically stated that tornadoes formed in the southeast quadrant. He demanded that Hazen withdraw his claim. Hazen backtracked, but only partway. He published a follow-up letter in which he asserted a new claim. It was true, he conceded, that Finley might be technically correct, but Hazen himself still deserved priority, because of his own “definite mention and emphatic calling attention” to the southeast quadrant. “Of all the ways of retreating from an indefensible position,” Finley countered in his own follow-up, “the method adopted by Professor Hazen is both unique and original.”
Their argument soon broadened out to take in the nature of tornadoes in general. In fact it became a kind of claustrophobic replay of the great storm war of the previous generation—except that it was conducted over the distance of an office desk, close enough for the combatants to feel each other’s breath when their voices were raised.
The grounds of this quarrel were both personal and scientific. Finley took for granted that he was the corps’s tornado expert and that his views should naturally be taken as settled. Professor Hazen considered himself the only true authority on the subject—even though, or particularly because, his views were sharply divergent from Finley’s, and in fact from every other meteorologist’s.
Hazen was a figure recognizable in almost every bureaucracy: the in-house skeptic. No matter what the prevailing view was on any issue, he was certain to take the opposite position and argue it with cool condescension and implacable tenacity. With the subject of tornadoes, Hazen considered his role to be the necessary corrective to Finley’s air of certainty. As far as he was concerned, Finley and every other expert had gone off the rails about tornadoes, and it was his job to bring to the subject some long-overdue common sense.
Hazen’s working premise was that tornadoes were an exceptionally rare phenomenon about which there was almost no reliable information. He didn’t believe that any of the basic facts about them had been established. He had reviewed the history of tornado studies, going all the way back to the works of the principals in the storm war James Espy and William Redfield and had concluded that they had all been thoroughly wrongheaded. There was no evidence of an upward movement of air in funnel clouds, as Espy had claimed; nor was there any evidence that the funnel rotated, as Redfield had believed. So what did the wind do in a tornado, if it neither rotated nor ascended? Hazen thought the question was still entirely open.
He did concede one possibility. He thought there might be something worthwhile in the long-neglected views of the third combatant in the war, Robert Hare. It could be that tornadoes were primarily electrical. One strong piece of evidence, according to Hazen, was the roaring noise so often heard around tornadoes. Hazen didn’t think this had anything to do with the inrush of wind; it was more likely to be the sound of an electrical discharge like that of a gigantic generator. He also thought electrical discharges accounted for the mysterious flashes and glows of colored light seen within the funnel. But most important, electricity explained the extraordinary destructiveness of the tornado: as the funnel passed over buildings, it was blasting them apart with lightning bolts.
Hazen refused to accept that there was an area of low pressure at the center of the funnel. He thought it was more likely to be an area of high pressure, and the tornado’s winds were in fact flowing outward. (He did not accept that a rapidly falling barometer was a sign of violent weather, either.) He was equally certain that reports of tornadoes half a mile wide or more were nonsense. He suggested that all such reports be “thrown out” from the study of tornadoes. He was aware that there were reliable reports of such storms, the most famous being the mile-wide funnel that had destroyed Natchez, Mississippi, in 1840. He didn’t deny that it had happened, but he didn’t think it was possible that it was a tornado. It was some other type of storm, possibly a hurricane. His own studies of the subject had proven to his satisfaction that a true tornado was almost never more than a hundred yards wide and that an observer should be able to get within ten yards of it in perfect safety.
/> Ordinarily, Hazen’s views might not have gotten much of a hearing. But they happened to coincide with a backlash against Finley’s ceaseless talk of the dangers of tornadoes. Over the previous few years, a number of people in and out of the Signal Corps had been growing increasingly unhappy with the way Finley had been traveling around the Midwest as an authoritative spokesman about tornadoes and other forms of violent American weather. Local government officials in the states and territories beyond the Mississippi, real estate speculators, business associations, and all those involved in the clearing and settlement of the prairies had begun to speak out against the pernicious idea that tornadoes were common in their part of the country and that they posed any significant risk to the population. The head of the Iowa Weather Service, Gustavus Hinrichs, published an article in The American Meteorological Journal (which Professor Hazen was then editing) dismissing Finley’s tornado researches as fantasies. Most of the tornadoes Finley had recorded in Iowa, Hinrichs wrote, had never happened; “they have never existed outside the archives and publications of the Signal Service.” Hinrichs’s own researches plainly demonstrated that Iowa rarely saw more than “one real tornado a year.” This, he wrote, “is bad enough, and needs no amplification by professional tornado manufacturers.”
The corps’s own section chief in Oklahoma published a letter in the Monthly Weather Review (also edited by Hazen) attacking Finley. “Long-range forecasters, through their ignorant predictions of tornadoes, are causing much unnecessary alarm to the inhabitants of Oklahoma,” he wrote. “The constant fear and excitement have certainly a tendency to cause nervous troubles.” He was quick to be reassuring to those sufferers: “Not a single tornado has occurred this year, very few in other years, and the Weather Bureau has not as yet issued a single forecast of ‘Conditions favorable for severe local storms.’ ” And yet, because of Finley, “every time a thunderstorm occurs, or a rain cloud appears in any part of the sky, accompanied by even moderate winds, every cave and cellar is filled with frightened men, women, and children.” He didn’t approve of the use of tornado caves anyway: “This undoubtedly causes more deaths as a result of exposure in these damp places than have been caused by all the tornadoes that ever occurred.”
Professor Hazen himself magisterially summarized the situation in an editorial for the Monthly Weather Review. “When the Bureau began to collect and publish fairly correct statistics relative to these storms,” he wrote, “the sum total of their number and the attendant destruction was so large as to be very impressive.” But the calm analysis of these numbers had shown that the dangers had been vastly overstated—twisted by “sensational writers” to give “a most alarming picture of the condition of the atmosphere in this country.” It was important, he wrote, that meteorologists “did not join in the popular cry ‘beware of the western tornado.’ ”
On the evening of March 27, 1890, an unusually violent line of thunderstorms came across Louisville, Kentucky. The barometric pressure fell so quickly that the gas streetlamps all over town were snuffed out. Then a rapidly moving gust front knocked down the telegraph wires. A torrential downpour began falling. In the midst of the fury a tornado roared through the heart of town. One witness said it was like a turnip. Another compared it to a balloon lit up from within by orange fire. It was exceptionally powerful and so tightly organized that even though it passed a few hundred yards from the Signal Corps’s weather station, the anemometer there barely registered a flicker of rising wind. Because the telegraph lines were down, the observers couldn’t send on a warning.
The tornado was only around seventy-five yards across when it was first seen. By the time it had cut through the residential neighborhood in the southwest side of Louisville and crossed into the downtown commercial district, it had swelled to more than half a mile wide. It wrecked warehouses, office buildings, churches, factories, and meeting halls. Most of the commercial and industrial buildings were empty at that hour, but the meeting halls were packed. The people of Louisville, as was common in the Midwest then, frequently passed their evenings by attending lodges, membership associations, and quasi-mystical orders. Downtown Louisville on that night saw gatherings of the local branches of the Independent Order of Odd Fellows (the lodges of both the white and the black members), the Roman Knights, the Knights and Ladies of Honor, the Daughters of Samaria, and the Sisters of the Mysterious Ten (these two last were lodges of African American women). At every one of these lodges there were injuries and deaths. In one hall fifty-five people were killed. By the time the tornado crossed out onto the surface of the Ohio River, more than a hundred people were dead.
Even though the Signal Corps still didn’t make tornado predictions as a matter of policy, it was widely blamed for not giving any advance warning. What made this criticism most painful to Greely and his forecasters in the Washington City headquarters was that they actually had decided to break their own rule and issue a warning. They had been monitoring the progress of the storm system as it crossed Kansas and Missouri, and they had recognized the strong probability that it would spawn tornadoes. At some point during the late afternoon of the twenty-seventh, the situation looked so dangerous that somebody—possibly John Finley himself—had persuaded Greely to issue the old public warning for “severe local storms.” But it arrived in Louisville too late to make the evening newspapers. The earlier forecast was still posted around town, as a kind of inadvertent monument to Weather Bureau ineptitude: “The indications for today in Kentucky are fair weather, followed in the western portion by rain, easterly winds, and stationary temperature.”
But the corps’s reputation took a much worse hit the day after the Louisville tornado. That was when the chief observer for New York City, Sergeant Elias Dunn, took upon himself to warn the people of the lower Mississippi valley of imminent disaster. Dunn happened to be a well-known figure in New York; he was called Farmer Dunn, because New Yorkers thought that only a farmer could care as much as he did about the weather. On March 28, Dunn gave interviews to several newspapers about what had happened in Louisville and what was likely to come. When the tornado had crossed the Ohio River, he said, it had stirred up a gigantic tidal wave that was even now rushing downstream. His best guess was that on April 1 it would reach Cairo, Illinois, at the confluence of the Ohio and Mississippi Rivers, and over the next several days travel down the Mississippi to the delta. He estimated that it would reach New Orleans around April 15. The wave then, he said, would be fifty feet high.
This prediction appalled and outraged the citizens of New Orleans and the lower Mississippi River valley. The corps was obliged to issue a public disavowal of Dunn’s views. No such tidal wave had been observed, the corps declared, and none was expected. Dunn had been speaking only for himself and had no business associating such a pronouncement with the corps.
There was one practical result of Dunn’s warning and the storm of criticism that followed. Chief Signal Officer Greely was forced to revisit the whole issue of tornadoes and attempt to settle the conflict about them within the corps once and for all. He believed that the most urgent practical necessity was ending the war between Finley and Professor Hazen. He did so by following the example set by his predecessor, General Hazen, whenever Finley got into trouble: he sent Finley away.
Finley was reassigned from the Washington City office to the most remote point Greely could find: San Francisco. There Finley was put in charge of forecasting for the entire Pacific region. Finley proved to be a reliable and popular forecaster who immediately won the trust of San Francisco’s newspapers and the public; one of his first public pronouncements was to correctly forecast the end of a severe drought that had been ravaging Northern California. But he said nothing at all about tornadoes: there had, after all, not yet been a single confirmed report of a tornado west of the Rockies.
Back in Washington City, meanwhile, Greely made several public statements about tornadoes. All of them echoed, or even parroted, Professor Hazen’s views. Their gist was that tornadoes
were not as dangerous as Finley had claimed and as was popularly believed. The official view of the Signal Corps, as articulated by Greely, was that tornadoes were a minor and rare phenomenon that was not worth the resources for extensive study.
It was a bad position to argue. Given the destruction the tornado had left behind in Louisville, Greely’s remarks came in for open ridicule and contempt in the press. He was frequently reminded that he had earlier praised Finley’s tornado work. So why had he changed his mind? Greely had no answer. But it didn’t matter anyway; he was soon forced to backtrack.
The weather that spring and summer proved to be unusually bad across America. In the months after Louisville, several other very destructive tornadoes were reported from the Midwest and the South. There was widespread speculation in the nation’s press about why. The general consensus was that the American climate was growing dramatically worse; the most widespread popular theory was that it was caused by the deforestation of the eastern half of the continent. At the end of the year, Greely responded to the public concern by announcing in his report to Congress that he was reopening Finley’s tornado research project. But this time he was determined to make sure it would have responsible leadership. Greely wrote: “Impressed with the number and violence of destructive tornadoes during the past year, it is believed that an investigation of phenomena of this kind on their numbers, area devastated, lives lost, and other such information might be of current interest. This work was intrusted to Professor H. A. Hazen, who has given much time and attention to these phenomena.”