March 15, 1941
TO: Director
FROM: M. A. Tuve
SUBJECT: Prowlers
At 2:45 a.m., March 15, Mr. Price the night watchman noticed two men in front of the Cyclotron Building. After watching them a few seconds he saw them go down the steps at the side of the building toward the driveway. He called the police then, Eighth Precinct, and they arrived some four minutes later and scouted the grounds and the first floor of the Cyclotron Building but could find no trace of the men. Mr. Price searched the top floor of the building but found nothing unusual. He notified Mr. Tuve at approximately 2:55 a.m.
The strangers caught snooping around Section T’s hub at the Department of Terrestrial Magnetism, the Cyclotron Building, had no good excuse to be anywhere near it, let alone in the quiet, dark early hours of the morning.
Tuve was anxious about security in general. “Confidential and secret information is being too widely dispersed in our organization,” he warned his team.
Beginning in late March, Section T staff were only allowed to discuss test results, project status, or the potential military use of the fuse on the second floor, where access was severely restricted. Fuse prototypes had to be stored there also, and covered when transported. If technicians required a special tool, they had to put in a purchase order rather than borrowing it from another group.
Tuve’s message was clear: “Keep out of other’s [sic] rooms.”
Any public discussions of their classified work—at, for example, the local Hot Shoppe restaurant—were to cease immediately.
“The oath of secrecy means silence,” Tuve reminded them.
As a general rule, workers were to assume their telephones were tapped. Staff had to present salmon-colored IDs to enter the Cyclotron Building, and visitors could enter only through the south door, and only after signing a register.
Trash marked secret or confidential had to be incinerated. Lab notes, drawings, and blueprints were locked up nightly, and shades were drawn.
The Cyclotron Building, a functional, two-story structure, today resembles an aging preschool or a high-school science building. But in 1941, the brand-new thirty-six-room construction was home to a particle accelerator meant to appease biologists who wanted to radiate fruit flies, frog eggs, and rats with neutrons.
The accelerator was incomplete, and the building poorly furnished.
Acoustics in the Cyclotron Building weren’t designed for keeping secrets. Narrow slits between the panels of glass bricks embedded throughout the edifice—the kind you might see striping the walls of an indoor swimming pool—allowed staff in one room to clearly hear chatter in the next. Tuve had separate fuse projects divided into groups, and scientists tackling one challenge couldn’t hear classified details of another. The ceilings had to be soundproofed, and the holes in the walls plugged.
To conceal their workbenches and experimental parts from visitors, the scientists needed some one dozen tarps, roughly fifty square feet each.
At first, Section T staff had no place to hang their coats. They needed chairs and stools. Simple metal chairs would do, Merle urged, in one of a series of pleading memos to the director of DTM, John Fleming.
Ceiling sprinklers—emergency nozzles for chemical fires—leaked.
There weren’t enough trashcans.
Merle, rather embarrassingly, was compelled to write Fleming to request “adequate supply of pencils, pads, and other stationary [sic].” A DTM administrator, he remarked, “requested a formal memorandum from me to you before he can do this.”
Merle wanted six Sears, Roebuck workbenches, three CONFIDENTIAL stamps, ten poles for opening high windows, and some linoleum-topped tables.
One of his team supervisors had no desk, so he held group meetings around a lathe. Even as Fleming dutifully fielded Merle’s requests, it was difficult for the DTM director to keep pace. Tuve was adding new recruits to the team weekly. By February of 1941, he employed some thirty physicists, radio engineers, and technicians, three typists and stenographers, and two watchmen.
Equipment to build and test fuse parts poured in: tool sets, tool grinders, drill presses, lathes, milling machines, blowtorches, safety goggles, calipers, clamps, soldering irons, and soldering pencils. Merle needed microscopes and micrometers for precision measurements, a “Vacuum Tube Voltmeter” to check the electrical characteristics of tubes, resin and beeswax to cushion the tube parts inside shells, electric hot plates and coffeepots to melt the wax, thermometers, batteries and battery testers, binoculars and a range finder for field tests, a radio receiver unit, squibs, resisters, sockets, plugs, cables, spools of wire of various tensile strengths and electrical properties, and many pounds of explosive powder.
Merle’s office was in room 101 by the entrance. Directly across the hall was the “Powder Room.” He notified his staff to “KEEP OUT” of it.
“Don’t handle explosives without proper instruction,” he cautioned. “Experts are never careless. Beginners must not be careless even once.”
When building prototypes, he reminded them, they needed goggles to work with any equipment “liable to throw particles or blow up.”
He asked Fleming to order a first-aid kit “of considerable size.”
Among the memos and warnings posted on the bulletin board were more mundane messages: apartments for rent, a request not to park by the neighbor’s house, and an edict from Director Fleming not to play baseball on the lawns.
Fleming would soon change his mind. If the Section T staff working on the secret project agreed to use the grass at the top of the hill, Tuve relayed, “a not too violent form of baseball will be perfectly all right.”
“P.S.,” Merle added, “Please don’t knock down the weather man’s equipment.”
Section T needed an airplane.
If the British were right, and some kind of “electrostatic” sensor might detect an aircraft by its electric charge, the first step was measuring the charge.
Tuve outsourced the job to a new consultant: head of the physics department at the University of New Mexico, the aptly named Jack Workman, a scientist in his early forties with a plump face, buzzcut, and streaks of gray over his ears.
Through a military liaison, Richard Tolman arranged for an Army Air Corps monoplane, an O-47—an aircraft just bigger than a Japanese Zero—to be loaned from Fort Sill, Oklahoma, along with a pilot.
In a series of experiments at the Albuquerque airport, Workman had the O-47 pilot do flybys over an “electrometer,” a procedure similar to—but more expensive than—waving a balloon near your hair after rubbing it on the carpet. The metal plane’s electric field, Workman found, varied dramatically. Snow and ice lent the aircraft a negative charge; in nicer weather, it held a positive one.
“In one test on a cumulus cloud,” Workman reported to Tuve, “the plane was negatively charged after passing through the base . . . and positively charged after passing through the top and returning through clear air.” After Workman installed a voltmeter on an aircraft wing, he discovered that the static charge on a plane over fifteen thousand feet was too weak to trigger a sensor.
So ended the electrostatic fuse idea.
Section T’s “acoustic” fuse required Merle to order—or ask Fleming to order—microphones that could withstand high heat and humidity. (Imagine the conditions in a Pearl Harbor cargo hold or a tropical munitions lot.) Initially, Tuve’s engineers tried to copy the British acoustic design for bombs based on “two photographs of poor quality,” one drawing, and “a meager description of a few tests.” That plastic prototype was abandoned, and another cylindrical design was constructed of metal, wood, and sponge-rubber rings.
Merle described it as a “flying toilet bowl.”
When the scientists attached the listening devices to falling dummy bombs, they learned that the rushing wind, the bombs’ “self-noise,” was too loud for the microphones to distinguish the roar of a nearing aircraft.
The “acoustic” fuse was dead, too.
Tests on a “photoelectric” fuse that reacted to small changes in light, and experiments on a radio fuse design, on the other hand, showed promise. Through early 1941, Section T assembled crude prototypes of both designs on Cyclotron Building workbenches. Constructing basic models that didn’t have to withstand the punishment of antiaircraft guns wouldn’t solve the rounds-per-bird problem, but it would allow Section T scientists to field-test the fuse circuitry.
James Van Allen, a twenty-six-year-old physicist, worked on both designs.
Shy, handsome, with a remarkably kind face, Van Allen was a farm boy from Mount Pleasant, Iowa. Like Merle, he was captivated as a child by radios and mechanical and electronic gadgets. He tinkered with old car parts and built a Tesla coil, horrifying his mother by demonstrating how the foot-long electrical surges caused his hair to stand on end and shoot out little sparks.
After high school, he’d hoped to enroll at the U.S. Naval Academy. But he failed the physical due to flat feet, bad eyes, and poor swimming.
When Van Allen was recruited by Merle in late 1940, he was a promising young Fellow at the Carnegie Institution with a stipend of some two hundred and fifty dollars a month. He lived in a single upstairs room in a house just down the hill from the campus.
In the early months of Section T, as the Luftwaffe pummeled London, he noticed that his colleagues were drifting off to other areas of DTM to work on some secret project that seemed exciting, challenging, and important.
He felt, he said, that he was “fiddling while Rome burned.”
At first, he worked on a new photoelectric design. Like solar panels, the “photocell” sensors converted light to electricity. If the light shifted, the change in illumination translated to a jump in electric output—which could be used as a signal to trigger.
His trouble was wiring a sensor so it could detect a 1 percent change in light when it was dark and during the day, when it was two thousand times brighter. A logarithmic circuit, which reacted only to relative changes, solved the problem.
“If we had the whole place full of bright lights,” Van Allen said, the circuit triggered if you simply “waved a fan.” When he turned the lights off, and you “could barely see in the room,” he explained, “I’d wave the fan again and it triggered.”
The “rudimentary” radio circuitry was just as sensitive. All kinds of reflective surfaces would bounce the signal back to the device. The prototypes would trigger, Van Allen said, “if you brought a fly swatter near the machine.”
The radio model was “a funny kind of circuit.” Normally, a radio that sent and received signals was wired to be insensitive to the objects around it, but with such a simple design, Van Allen found that it was almost impossible to build a prototype that didn’t have a strong sensitivity to nearby objects. “You could have someone open the door and the thing would kick off.”
By the summer of 1941, field tests of the two fuse types made it clear which had the most promise. Fitted into dummy bombs and tested at a nearby air base, the light-sensing model proved inconsistent. The lens opening needed to be of different sizes depending on the weather, and the flight tests had to be flown “away from the sun” to avoid glare catching the lens and triggering them early.
The radio fuse was the most promising design.
And yet, it remained impossible unless Tuve’s group could produce tiny electronics capable of surviving a cannon blast. Just as problematic, they hadn’t yet solved the “microphonics” riddle of keeping the whole damn journey quiet enough for a miniature flying radio to send and receive its signals clearly.
For those puzzles, Merle needed a specialist, an expert not in radio circuitry but in advanced structural engineering: a scientist who understood how materials react under great stress, how they bow and bend, and when they break.
Ray Mindlin, a thirty-four-year-old assistant professor at Columbia University, had a taste for Chopin, dry humor, and sports cars. He was a specialist in materials science. Confident, stylish, with dark bushy eyebrows, he looked like the type of professor who engaged in profound discussions, in hallways, with his hands in his pockets. He ran track, and was fast off the blocks. According to a friend, “If there had been such a thing as a five-meter race, he would have won it all the time.”
Asked how things were going, he would reply: “Fair to Mindlin.”
The engineer had studied stress analysis and problems like the “torsion of structural beams” and the weight distribution around tunnels.
He was recruited in January of 1941, by Western Union.
According to a Section T staffer, Mindlin had just finished a consulting job to help explain why the Tacoma Narrows Bridge collapsed in late 1940. Battered by forty-two-mile-an-hour winds, the bridge had wobbled surreally, twisted like a trampoline, and flung chunks of concrete into the air “like popcorn” before crashing dramatically into the waters of Puget Sound nearly two hundred feet below.
Engineers blamed the failure on “excessive oscillations.”
At first, Mindlin didn’t have security clearance. So Section T lied to him about the tiny, rugged tubes that he was to help build. The vacuum tubes were meant for meteorological balloons, they said, and had to withstand a long fall to earth.
Led by Henry Porter, the bulky Chicagoan, Section T’s “rugged electronics” team now had a bevy of centrifuges at their disposal that could simulate thousands of times the force of gravity. As revised hearing aid tubes arrived, Mindlin tested their electrical properties, spun them in centrifuges, and had samples fired in shooting tests. He also pulled engineering tips from strong tube designs, including a Canadian tube that protected its thin wires, which ran vertically up the narrow tube, by threading them through small platforms made of mica.
The real trouble was the fine electronics themselves.
In early 1941, Mindlin produced a sequence of diagrams, “design curves” that broke down how tough each tube part was. It wasn’t so different from how engineers approach bridge design. The tubes even had miniature cantilevers inside, like bridge supports. Mindlin worked out the warping and final “yield” points of dozens of minute components: grids, grid wires, and grid posts, getters, which kept the vacuum tubes sealed, “presswire welds,” and mica support “spacers.” He laid out “the bending moments at the supports and at the midspan and the twisting moment at the supports” for a “grid lateral.” He pored over blueprints of tube parts and compared yield points to the muzzle velocities of antiaircraft guns.
Or, as Porter put it, “Mindlin appeared one morning with a series of graphs.”
Working closely with the subminiature tube contractors—Hytron, Raytheon, and Bell Telephone Labs—the rugged electronics team tested, refined, and retested a series of prototypes to fix the myriad design weaknesses. One of the most daunting roadblocks was the filaments. Like carbon filaments in a light bulb, these wires heated up when electrons flowed through them. Each an inch long and 0.00075 of an inch wide, they were the centerpiece of the tubes and the most delicate component. At Hytron, the filaments had to be mounted using magnifying glasses by teams of women hired for their “nimble fingers.”
When tubes were spun at high velocities, as they would be when fired from antiaircraft guns, four bad things tended to happen: the electrical characteristics changed, the tubes short-circuited, the filaments broke, or the filaments bowed and vibrated so badly that the microphonics problem was untenable.
To keep the fine wires from bending, the miniature “cello strings” would have to maintain a very high tension. Designing a filament support to keep that tension would pose problems, but the first step was to find a material that had a naturally high tensile strength even at hot temperatures.
Bob Brode, a forty-year-old physicist (who was one of the top experts on Tuve’s original list of recruits), reached out to manufacturing companies in Chicago and Newark. He requested samples of tough, fine wires made of silver-clad platinum and nickel and chromium alloy.
Mindlin and Brode found that tungsten (or wolfram, the W on the periodic table) fit the bill. Tungsten has the greatest tensile strength of all metals and the second-highest melting point of all elements, at over 3,000 degrees Celsius.
With Mindlin’s design curves, and tungsten filaments, and a ton of trial and error—bit by bit, test by test—the tubes were becoming quieter and stronger. After Raytheon’s president called regarding tube production, Merle recorded his reply in his notebook, referring to himself in the third person.
The tube design that emerged was stacked, with platforms like floors in a high-rise that were strengthened by miniature columns. Mindlin didn’t seem to care how minute Section T’s devices were. To him, these weren’t glass tubes the size of paper clips. They were narrow, three-story buildings.
Gardiner Means was worried about his sheepdogs.
The cannon fire on his Virginia estate frightened the animals and rattled their nerves. Several times, the canines “lit out for parts unknown” and did not return for some time. Also, as Means informed DTM’s Director Fleming, he was expecting a hay crop soon in the test field, and depending on how frequently Section T planned to use the site, the men might trample the hay or, in any case, “They may have some difficulty in recovering whatever it is that they are shooting.”
The truth was, the old howitzer was too big for Means’s farm. Even one-pound rounds, falling from fifteen thousand feet, were quite dangerous, and Section T had already moved on to firing larger ammunition at military firing ranges.
For a short time, Tuve’s group utilized a Navy range in Virginia Beach. But the test rounds, shot vertically, kept drifting unpredictably in the breeze. The Navy decided that the shells were landing too close to its torpedo station.
At a proving ground in Maryland known as Stump Neck, Section T shot six-pound rounds out of a 57 mm gun. Accessed by a narrow causeway, Stump Neck was easy to guard, and “drifters” landed in the water. For protection while the sizable metal capsules plummeted back to earth, they erected a four-inch plate of steel on posts and covered the top with sandbags.
The procedure for test shoots was the same as before. They potted various fuse parts in metal cylinders using beeswax and resin, fitted them into hollowed-out shells, fired the rounds straight up, and tried to hear and see where they landed. Using clamshell diggers normally used to make fence-post holes, they excavated the pods for analysis. It took nearly a minute for a shell to speed thousands of yards high, slow, apex, plunge, and finally bury itself four feet in the dirt.
Stump Neck’s caretaker had a Chesapeake retriever, Curly, who was not at all frightened by the gun blasts. When the rounds hit the ground, the dog would run out madly and dig into the earth or splash hopefully into the water. Curly seemed to imagine that the gigantic gun would bring down many ducks.
For some early tests, to make sure that the landing impact wasn’t damaging the test parts, they used modified signal flares. A “Star Shell,” shot from an antiaircraft gun, normally released a parachute holding an illuminating “candle” that could light up the night sky. Tuve’s scientists simply replaced the candles.
At an Army proving ground in Aberdeen, Maryland, Ray Mindlin watched as the miniature parachutes floated his test pods gently back to earth.
Mindlin’s tubes were withstanding gun blasts at an encouraging rate. In a March test shoot of Bell Labs samples, two-thirds of the prototypes survived.
On April 20, Jack Workman, in from New Mexico, fired a very special fuse part at Stump Neck: an oscillator for sending and receiving radio signals. The transmitter used a tube similar to the one developed by Mindlin as an amplifier, only wired differently and connected to an antenna. Workman turned on a radio receiver, and heard, through headphones, a live radio signal coming from a shell in flight.
A radio fuse was possible.
On May 8, Merle traveled down to a Navy proving ground in Dahlgren, Virginia, to witness the breakthrough firsthand. Accompanying him was the head of NDRC’s Division A, the pipe-loving Richard Tolman.
Dahlgren was home to one of Section T’s strongest advocates, Lieutenant Commander Deak Parsons; years earlier, he had recognized Tuve’s work as a basis for radar. With sharp brown eyes and a wrinkled brow that suggested an abnormally large brain, the thirty-nine-year-old wasn’t a typical military man. He was, his biographer wrote, “a new breed of officer.” An admiral later described him as “the finest technical officer the Navy has had in this century.”
Parsons was deeply interested in—and well informed about—radio engineering. He also happened to be a former gunnery officer and ordnance expert.
At Dahlgren, the artillery was a marked upgrade from the outmoded guns and one- or six-pound shells Section T was used to. The base housed the massive five-inch guns common on Navy ships, guns that weighed two tons apiece and that were in the most desperate need of smart fuse ammunition. These were the kinds of gun that could actually take down enemy aircraft from a great distance. The firing range here was aquatic, stretching twenty miles down the river toward Chesapeake Bay. Parsons often helped Merle arrange shoots toward the target area five miles downstream. During firing, barges and fishing boats were cleared out of the danger zone.
To view that day’s test, Tuve and Tolman boarded an observation boat and drifted down the Potomac. The river bottom was littered with shells. Seagulls squawked overhead. Over two miles away, a gun blast echoed and a shell whistled past above their boat as the physicists listened intently to their radio receiver for a signal from inside the round. Thousands of yards downstream, tall white plumes of frothy spray shot up like geysers as the shells splashed down one by one.
They heard signals in three of seven rounds.
For Merle—and for the U.S. Navy—the test was a “go” sign. Van Bush informed President Franklin Roosevelt of the results: “Success is now in sight.”
The little tubes weren’t quite tough enough yet, and the microphonics puzzle wasn’t fully solved, but the basic components of the fuse were in place. “It is now clear that such devices can be made,” Merle wrote, on June 11. Urged on by Parsons, he drafted a plan to expand Section T from a staff of sixty-five to a hundred and twenty, and to double his monthly budget to sixty thousand dollars.
DTM’s Director Fleming had to order more desks and aluminum chairs.