Share This Article

German soldiers train to cope with enemy gas attacks along the Western Front. American troops received no such training. (SZ Photo/Scherl/The Image Works)


Late in World War I, the U.S. Army undertook development of its own deadly chemical weapons. Its experiments still haunt a Washington, D.C., neighborhood.

BY THE TIME THE UNITED STATES ENTERED WORLD WAR I, the Europeans had been gassing one another on the battlefield for two years. The American army had no experience of chemical weapons. It should have worried about defense against gas attacks—training officers and individual soldiers, providing decontamination gear, and familiarizing its medical staff with treatment of gas casualties—but it did not, and American soldiers would suffer as a result. The army did supply the men with some protective gear against gas attacks, and it began a crash program to develop its own poison gas, a secret weapon that would force Kaiser Bill to his knees.

The Americans, as they entered the war, were novices at the game of chemical warfare. When the first U.S. troops arrived in Europe in June 1917, the American public knew little about poison gas. The British had cut the German transatlantic cables in 1914, two days after the start of hostilities, and almost all war news that came to America came via the Allies. While the French and British deplored the Germans’ gas attacks in their propaganda, they were vague about the effects of gas, in part because they did not want to scare off the Americans. By the spring of 1917, the Allies imposed a total news blackout on gas warfare because, in the words of Britain’s assistant secretary of war, it might result in an “unreasonable dread of gases on the part of the American Nation and its soldiers.”

The American public’s ignorance of chemical warfare was more understandable than the American military’s failure to prepare for it. During a 1916 congressional hearing, a senator asked a colonel on the army’s General Staff whether the Europeans were using poison gas. He replied, “The papers say so, but we have not had any actual reports from our observers.” Not until February 1917—two months before the United States entered the war—did the army quartermaster general ask who would supply gas masks should the need arise. The Bureau of Mines was assigned the task. In May the War Department ordered 25,000 gas masks resembling the British box respirator mask. When the masks were shipped overseas, the British immediately found them to be useless against Germany’s latest gases.

Even had masks been available, no one in the military knew enough to construct a training program for their use. By mid-July, 12,000 U.S. troops were 30 miles from the front with no defensive gas training and no gas masks, even though General John J. Pershing had on July 5 established a Gas Service section within the American Expeditionary Forces. He appointed Amos Fries, a combat engineer who held only the rank of lieutenant colonel, as its chief. By September, Pershing realized the urgency. He cabled Washington: “Send at once chemical laboratory, complete equipment and personnel, including physiological and pathological sections, for extensive investigation of gases and powders.” The majority of U.S. troops entered the European fight during and after the German spring offensive of 1918. The Germans had a field day gassing the green American soldiers, whose casualty rate was extremely high.

Gas defense was not the whole of the American story. The United States employed gas offensively in the same way as did its allies and enemies—firing artillery and mortar shells that were loaded with chlorine, phosgene, and mustard, all gases that had been developed by German chemists. But back in Washington, the army’s chemists were preparing an American surprise for the Germans.

Immediately after the United States entered World War I, American University in Washington dedicated its campus to the war effort. The army had no experience in chemical warfare, so the Bureau of Mines, which knew about poisonous gases in coal mines, planned to use the campus for research and pilot production of chemical weapons. The campus consisted of only a single building at the beginning of the war; by its end, 1,200 scientists would occupy 153 buildings on American University’s 509 acres. No one seemed to question whether the nation needed to pursue its own poison gas development or whether Washington was a suitable site for such research.


ON SATURDAY, AUGUST 3, 1918, Senator Nathan Bay Scott was taking the morning air on his back porch with his wife and sister-in-law. Scott had represented West Virginia for two terms. After losing his seat in 1910, he had remained in Washington, gone into banking, and grown rich. Sipping his coffee, he looked out over the American University campus, which adjoined his house. He had watched 65 buildings go up in the last 16 months, including barracks, mess halls, and training facilities for 4,400 troops. The construction noise and dust had no doubt been annoying, but the senator likely accepted the sacrifices as necessary in wartime.

Washington in August is often unbearably hot, but the weather was pleasant that morning—70ºF, a clear sky, and a breeze from the east. That breeze, the senator noticed, was pushing a yellow cloud toward his house. Perhaps, the senator thought, someone was burning brush. Then he and his family choked as the cloud attacked their throats and eyes and burned their faces. They stumbled inside and closed the house as best they could. It was a narrow escape.

The Washington Post ran a story the next morning: “While experimenting with what is said to be German mustard gas,…[soldiers] came very near killing Senator Nathan B. Scott and several of the women members of his family….A number of chickens, birds, and some small animals were killed by the fumes.” That was the army’s cover story; in fact, the gas was not German mustard but a secret American weapon, Lewisite.

An accident at the pilot plant that was manufacturing the poison had resulted in a spill of about eight pounds. Two years later, after the end of the war, the New York Times would call Lewisite “the deadliest poison ever known,” a substance so deadly that a “drop poured in the palm of the hand would penetrate to the blood, reach the heart and kill the victim in great agony.”

Lewisite’s origins are unclear. The substance was likely first synthesized by the Reverend Julius Nieuwland, a Catholic priest at the University of Notre Dame, in 1904. But Captain Winford Lee Lewis, an army chemist, is the undisputed father of Lewisite as a weapon, developing it in early 1917. While mustard gas had proven to be an extremely effective blistering agent, it was considered too persistent to be used on the offensive—it hung around so long that it would poison the attacker’s own troops as they moved into territory that the enemy had abandoned. It had another disadvantage: Its physiological action is delayed for hours, like a particularly hellish poison ivy, so enemy troops were often not immediately aware they had been gassed and would continue fighting. Captain Lewis was asked to find a poison gas that would outdo mustard, one that was “(1) effective in small concentrations; (2) difficult to protect against; (3) capable of injuring all parts of the body; (4) easily manufactured in large quantities; (5) cheap to produce; (6) composed of raw materials that were readily available in the United States; (7) easy and safe to transport; (8) stable and hard to detect; and, most importantly, (9) deadly.” A colleague suggested that Lewis take a look at Father Nieuwland’s doctoral dissertation, in which the chemist-priest had described combining arsenic trichloride and acetylene. The result had made him deathly ill. When Lewis repeated Nieuwland’s experiment, he found that the results matched his goal—immediately painful, more toxic than mustard, and less persistent than mustard because it decomposed in water. But he insisted that his compound was not the same as whatever Nieuwland had made, which Nieuwland had described as a black, gummy mass. Lewis was adamant that he deserved credit for the weapon’s discovery, and he was successful in gaining recognition. “Lewisite” was the name that would be used, perhaps because “Nieuwlandite” was just too difficult.

Once he had established Lewisite’s properties, Lewis passed the compound to his supervisor, Captain James Bryant Conant. Conant had been a 24-year-old Harvard organic chemistry instructor when he offered his research skills to the country during the war; he worked as a civilian for the Bureau of Mines at first and was later commissioned in the Army Sanitary Corps.

Conant was a thin, bespectacled Yankee from Dorchester, a working-class neighborhood of Boston. He had been awarded a scholarship to Harvard, where he had majored in chemistry and graduated Phi Beta Kappa in 1913. He was both secretive and ambitious; his son would say, “Anything that he wrote or he said he always weighed the political consequences.” When Conant’s autobiography appeared in 1970, John Leonard reviewed it in the New York Times and commented that it seemed to have been written “in an airless room on some other planet.” “Mr. Conant doesn’t examine his life,” Leonard continued, “he reports it as might an obituary writer.” Despite hailing from Dorchester, Conant had broken into the circle of the Harvard elite, making contacts that would serve him in later life—as president of Harvard and high commissioner in Germany after World War II.

Conant had continued in Harvard’s graduate program and received his doctorate in organic chemistry in 1916, just before the United States entered the war. He and two chemist friends could see that many organic chemicals were selling at very high prices because of the war. They decided to manufacture benzoic acid, but they found that producing chemicals in large batches was not the same thing as working in laboratory flasks: They burned down one building and used the insurance settlement to move the business to a second.

When Roger Adams, an instructor in organic chemistry at Harvard, moved to the University of Illinois, Harvard offered Conant the open faculty position. He accepted, and the move to Harvard was timely, as the benzoic acid business ended in catastrophe with a second fire two months later. In his autobiography, Conant says, “The laboratory work on which the new process was based (and which yielded a patent eventually) proved to be incomplete.…A fire demolished all but the concrete walls of the plant. Stanley Pennock [one of Conant’s partners] and a plumber…who was repairing the piping were killed almost at once. Loomis [the other partner] escaped with minor injuries.” Conant did not mention the plumber’s name (Max Stein), nor did he remember that a mechanic, Samuel Welte, also died.


WHEN CONGRESS DECLARED WAR in 1917, Conant considered enlisting, but an MIT professor convinced him that he could best contribute by offering his chemistry skills to the government. He was put in charge of a research group at American University in Washington. In September 1917 he was given the task of devising an industrial process for the synthesis of mustard gas, which the Germans had begun using only two months earlier. His efforts resulted in a process that yielded 30 tons of the gas a day.

In May 1918, a year after the American entry into the war, Conant was assigned the task of turning Lewisite into a weapon, a job that included both developing an industrial process for its production and determining its toxic qualities on animals and human volunteers. One of the test volunteers, Sergeant Temple, later said that after a small drop of Lewisite was applied to his forearm, extremely painful inch-high silver blisters formed that didn’t heal for eight weeks; his forearm was still scarred when he was interviewed almost 50 years later.

The animal test subjects were not treated as gently as Sergeant Temple. Dogs and goats were attached to stakes in nearby fields, exposed to Lewisite bombs, and observed as they struggled and died. “Their nostrils clogged and they coughed excessively. Many died at this stage. If the dogs continued to live, they sneezed violently with a continuous flow of watery fluid from their nostrils. More dogs died during this period.”

The government had difficulty convincing chemical companies to produce poison gases. The work was dangerous, and the only customer—the government—would immediately discontinue purchases when the war ended. The army constructed its own plants at Edgewood Arsenal in Maryland for chlorine, phosgene, and mustard. The accident rate among plant employees was appalling; at one point mustard gas production had to be shut down for lack of workers. After the war, the Journal of Industrial and Engineering Chemistry paid tribute to those “who were killed, not in the thrill of battle, not under the glory of a charge, but back here in the steady grind of preparing material for the men at the front.”

In the spring of 1918, the army pushed to take control of all chemical warfare operations, including research and production within the United States. On June 28, President Woodrow Wilson established the Chemical Warfare Service (CWS). Although General Pershing had earlier removed Major General William Sibert from command of the 1st Infantry Division before it was deployed in combat, he recommended him to command the CWS, with Amos Fries, now brigadier general, reporting to him and running things in France. Because Lewisite was to be America’s secret weapon, it was not produced at Edgewood but assigned its own production site in Willoughby, Ohio, about 30 miles from Cleveland. The similarities between Willoughby in World War I and Los Alamos in World War II are striking: Willoughby was called “the mousetrap,” because soldiers could get in but not out—no one assigned to Willoughby was transferred until after the war, and soldiers were told they would be court-martialed if they revealed what was being manufactured or even where they were stationed. The plant’s mailing address was a post office box in Cleveland, and all mail was censored. If Willoughby was the Los Alamos of World War I, Conant was its Oppenheimer. By late July 1918, Conant, then working at American University before the project moved to Willoughby, had devised a five-step industrial synthesis for Lewisite. He dreamed that this would be the weapon that would win the war.

By that time, the army had taken over the poison gas effort, and Conant had been commissioned as a captain. He was promoted to major and sent to Willoughby to produce Lewisite in quantity—3,000 tons would be needed for an offensive planned for spring 1919. His departure for Willoughby was as timely as his earlier departure from the benzoic acid business: Two weeks later, the accidental release of Lewisite in his American University pilot plant gassed Senator Scott and his family.

Conant asked for a thousand men at Willoughby, and by November 1918 a total of 542 enlisted men and 22 officers were stationed there. They worked to repair an old automobile plant on the site, lay railroad sidings, and build the production facilities. It is uncertain how much Lewisite Willoughby produced before the armistice on November 11; Winford Lewis claimed that 3,000 tons were synthesized, while Conant said that only pilot production runs were accomplished, although he gave no figures. Other accounts mention 150 tons produced with peak production of 10 tons a day.

The disposal of the Lewisite is equally hazy. According to one account, the material had been loaded into shells and was halfway across the Atlantic when the order came to dump it at sea; according to another, 364 55-gallon drums of Lewisite were shipped by train after the war’s end from Willoughby to Baltimore, where they were loaded onto barges and sunk 50 miles off the coast.

The armistice ended work at Willoughby, and America’s secret weapon was not used. Lewis said the joy the soldiers at Willoughby felt at the Allies’ victory was mixed with disappointment: “We of the Chemical Warfare Service felt strangely punctured, depressed and irritable next morning after the celebration.”

The American experience with gas was different from that of the British or French. Americans troops were first deployed at the front in October 1917, just when—as we have seen—the Germans began to use mustard, which was far more effective than the other gases; 23 percent of American casualties were from gas, although only 2 percent of these were fatal. While the high casualty rate was certainly due in part to inadequate American preparation, the British gas rate spiked at the same time: Of the 188,000 British gas casualties inflicted after the first chlorine attack on April 22, 1915, about 161,000 occurred after July 1917. Most were from mustard.

The CWS, like the rest of the army, rapidly demobilized after the war’s end. The regular army officers detested gas and, along with the general public, wanted gas warfare abolished. General Peyton March, the army chief of staff, testified before the Senate: “When I was in France I saw 195 small children brought in from about 10 miles from the rear of the trenches, who were suffering from gas in their lungs, innocent little children who had nothing to do with this game at all.” A congressional bill, supported by the War Department, proposed placing a greatly diminished CWS under the Corps of Engineers.


TODAY, THE AREA AROUND THE AMERICAN UNIVERSITY SITE is an affluent Washington neighborhood called Spring Valley. In 1990, construction workers dug up laboratory equipment and a 55-gallon drum there, and then headed for the emergency room complaining of burning skin and eyes. In 1993, utility workers uncovered rusted bombs, and the Army Corps of Engineers excavated and removed 43 shells suspected of containing poison gas. The Corps then announced that the site was safe.

In 1996 workers planting a tree were overcome by odors and suffered severe eye burning. They dug up broken bottles and glassware containing liquids; an environmental firm assayed the soil and found arsenic, a component of Lewisite. The army ignored that finding, but in 1999 agreed to test the South Korean ambassador’s residence, where it found high levels of arsenic. Shortly after that, another excavation unearthed 12 bottles of Lewisite and eight of mustard gas. In 2004, the army estimated that another four years would be required to fully decontaminate the American University site.

As recently as 2012, the work was continuing, with the Army Corps of Engineers demolishing a house that sat on top of a disposal pit. Sergeant Charles Maurer, stationed at American University in World War I, had photographed the area, and on the back of one photo, he wrote: “The most feared and respected place in the grounds…Death Valley. The hole called Hades.” Late in life, he would not go near the spot.


Patrick Coffey is the author of Cathedrals of Science and American Arsenal. Excerpted from American Arsenal by Patrick Coffey with permission from Oxford University Press USA, © 2014 by OUP.


This article originally appeared in the Summer 2014 issue (Vol. 26, No. 4) of MHQ—The Quarterly Journal of Military History with the headline: American Gas

Want to have the lavishly illustrated, premium-quality print edition of MHQ delivered directly to you four times a year? Subscribe now at special savings!