The Idea Factory: Bell Labs and the Great Age of American Innovation (5 page)

B
Y THE TIME
K
ELLY ARRIVED
at AT&T, the U.S. government had begun to concur with Theodore Vail’s arguments for his company’s expansion. A group of senators issued a report noting that the phone business, because of its sensitive technological nature—those fragile voice signals needed a
unified and compatible infrastructure—was a “natural monopoly.” A House of Representatives committee, clearly sympathetic to the prospect of simply dealing with a single corporate representative, complained that telephone competition was “an endless annoyance.”
⁵
In the Willis-Graham Act of 1921, the U.S. Congress formally exempted the telephone business from federal antitrust laws.
⁶

By then, the so-called natural monopoly had grown even larger. Indeed, the engineering department at West Street had become so big (two thousand on its technical staff, and another sixteen hundred on its support staff) that AT&T executives agreed in a December 1924 board meeting to spin it off into a semiautonomous company. They chose the name Bell Telephone Laboratories, Inc. Some of their reasoning remains obscure. A short notice about the new labs in the
New York Times
noted that “the new company was said to [mean] a greater concentration upon the experimental phases of the telephone industry.” The spin-off, in other words, was justified by the notion that scientific research at Bell Labs would play an increasingly greater role in phone company business.
⁷
Frank Jewett’s private memos, meanwhile, suggest that the overlap between the AT&T and Western Electric engineering departments was creating needless duplications and accounting problems. By establishing one central lab to serve two masters, the phone company would simply be more efficient.
⁸

On January 1, 1925, AT&T officially created Bell Telephone Laboratories as a stand-alone company, to be housed in its West Street offices, which would be expanded from 400,000 to 600,000 square feet. The new entity—owned half by AT&T and half by Western Electric—was somewhat perplexing, for you couldn’t buy its stock on any of the exchanges. A new corporate board, led by AT&T’s chief engineer, John J. Carty, and Bell Labs’ new president, Frank Jewett, controlled the laboratory. The Labs would research and develop new equipment for Western Electric, and would conduct switching and transmission planning and invent communications-related devices for AT&T. These organizations would fund Bell Labs’ work. At the start its budget was about $12 million, the equivalent of about $150 million today.
⁹

As president of Bell Labs, Jewett now commanded an enormous shop. That an industrial laboratory would focus on research and development was not entirely novel; a few large German chemical and pharmaceutical companies had tried it successfully a half century before. But Bell Labs seemed to have embraced the idea on an entirely different scale. Of the two thousand technical experts, the vast majority worked in product development. About three hundred, including Clinton Davisson and Mervin Kelly, worked under Harold Arnold in basic and applied research. As Arnold explained, his department would include “the fields of physical and organic chemistry, of metallurgy, of magnetism, of electrical conduction, of radiation, of electronics, of acoustics, of phonetics, of optics, of mathematics, of mechanics, and even of physiology, of psychology, and of meteorology.”
¹⁰

From the start, Jewett and Arnold seemed to agree that at West Street there could be an indistinctness about goals. Who could know in advance exactly what practical applications Arnold’s men would devise? Moreover, which of these ideas would ultimately move from the research department into the development department and then mass production at Western Electric? At the same time, they were clear about larger goals. The Bell Labs employees would be investigating anything remotely related to human communications, whether it be conducted through wires or radio or recorded sound or visual images. At the opening of the new U.S. patent office a few years after Bell Labs was set up, Frank Jewett, whose speeches were often long-winded and hyperbolic, found a way to explain the essential idea of his new organization. An industrial lab, he said, “is merely an organization of intelligent men, presumably of creative capacity, specially trained in a knowledge of the things and methods of science, and provided with the facilities and wherewithal to study and develop the particular industry with which they are associated.” In short, he added, modern industrial research was meant to apply science to the “common affairs” of everyday life. “It is an instrument capable of avoiding many of the mistakes of a blind cut-and-try experimentation. It is likewise an instrument which can bring to bear an aggregate of creative force on any particular problem which is infinitely greater than any force
which can be conceived of as residing in the intellectual capacity of an individual.”

Buried within Jewett’s long speech was a clear manifesto. The industrial lab showed that the group—especially the interdisciplinary group—was better than the lone scientist or small team. Also, the industrial lab was a challenge to the common assumption that its scientists were being paid to look high and low for good ideas. Men like Kelly and Davisson would soon repeat the notion that there were plenty of good ideas out there, almost too many.

Mainly, they were looking for good problems.

K
ELLY HEWED
to his vacuum tube work in the years after World War I. As he ascended steadily into management, his job came to include responsibility for developing the vacuum tubes built by Bell Labs for Western Electric, which was to say he saw himself as being responsible for improving the most important invention of his lifetime up to that point. Tubes could do much more than amplify a weak phone signal or radio transmission: They could change alternating current into direct current, making them a crucial component in early radios and televisions, which received AC from the power grid but whose mechanisms required DC to operate. What’s more, the tubes could function as simple and very fast switches that turned current on and off. Early in Kelly’s tenure, his tube shop made fifteen different models. There were large water-cooled tubes the size of wine bottles that were used in high-power radio broadcast stations; small tubes for public-address systems; and the famed repeater tube that had been Harold Arnold’s great contribution in bringing the transcontinental connection to bear.

Sometimes a vacuum tube was described as a cousin to the ordinary incandescent light bulb. In some respects this was true—for instance, both devices contained wires that were sealed inside a glass container. Yet the differences between them were far more pronounced. A factory could turn out tens of thousands of bulbs a day. But the daily output of, say, a repeater tube that allowed telephone conversations to be conveyed across
the country was at best in the hundreds. What’s more, the kind of tubes made in Kelly’s shop had to be forged with a jeweler’s precision. There was no room for error. If a light bulb failed, it would be easy to replace and not necessarily urgent; if a repeater failed, many conversations would, too. Money would be lost, maybe even lives.

Early in his career, Kelly wrote a long article explaining in meticulous detail how a vacuum tube was made. It reveals something of the nature of Bell Labs’ work in general at the time, which aspired to be at the leading edge of what any company in the world could achieve, both in conceptual and manufacturing terms. The tube shop Kelly described was located in a building a dozen blocks south of the West Street labs in Manhattan, at 395 Hudson Street. There, in a gritty industrial neighborhood, his workers, men as well as women, labored behind lab benches in large rooms outfitted for assembly and production. To explain the process, Kelly used the example of the repeater tube known as the 101-D. Its production began with a glass pipe about the size of a man’s pinkie. The pipe was heated and rotated on a machine so that its bottom opening could then be flared out. A different machine would take the top opening of the pipe, insert four long wires, and then heat the glass to seal the wires so they extended through the seal. The four wires resembled plant stalks poking through a hill of snow. This assembly was called the stem press.
¹¹

Next, a worker would attach, carefully and by hand, a solid glass rod atop the stem press, just behind the four wires that were already poking up. The glass rod was positioned vertically and was in turn superheated. A hand-operated machine was then used to insert in its hot, softened glass ten more wires. Two of these stuck out vertically, the other eight horizontally.

The assembly now looked more like a broken toy than an electronic device. It was a mass of wires, fourteen in all, poking out in all directions from a central glass core. But it wasn’t done. First, the tube’s glass core had to be heated and cooled and heated and cooled again in order to harden it. Afterward, a worker had to administer several chemical washes to remove grease and oil from the surface of the glass and wires. The
faintest trace of impurities raised the risk of failure. Finally it was time for a worker to arrange the functional parts of the vacuum tube—the parts that would amplify phone signals—around the glass core and wires. These parts were the cathode, grid, and anode. It had taken the Bell scientists years to figure out, in Edisonian, trial-and-error fashion, which materials worked best. The anode was a tiny flattened, hollow box of sheet nickel; the grid was a mesh fashioned from nickel wire of several different diameters; the cathode was a ribbon of metal, M-shaped, made from a platinum-alloy core coated with other trace elements. All of these parts were heated in an oven to 1,000 degrees centigrade to burn off imperfections.

Afterward, the tube shop workers welded these parts to the ragged wires sticking out from the glass core. The contraption no longer looked disheveled. It looked like a device with a purpose. Every part was now tidily connected and wrapped tight. At this point an employee would insert what they had in front of them—the assembly of welded wires and metal plates anchored to the glass rod—into a round glass bulb roughly the size and shape of a conventional light bulb. Then they would heat the bottom of the bulb to create a closed seal.

A vacuum tube couldn’t work without a vacuum inside. So the pumping began. It was a complex, four-stage process requiring several different pumps, all of the machines designed within Kelly’s tube shop itself. The goal was to eliminate the air inside—to “approximately one-millionth of an atmosphere,” as Kelly would explain—through a hole on top of the bulb. But afterward a few other steps still remained: The inside of the bulb, for instance, had to now be heated to about 800 degrees centigrade for further improvements in the vacuum. And the hole on top of the bulb had to be sealed. Finally, a worker would connect four wires dangling from the bottom of the vacuum tube to a small cylindrical base and fasten the base to the bottom of the bulb. At last, after this final step, one could admire the finished vacuum tube, the 101-D, and get the impression of looking at a large but fabulously complex light bulb with an intricate miniature architecture of metal plates, posts, and wires inside.

Kelly called the tubes “miracle devices” that would usher in a great
age of electronic communications. But he knew better than anyone how difficult they were to make: labor-intensive, complex, expensive. He knew they soaked up vast amounts of electricity to operate and gave off tremendous amounts of heat. Most of all he knew they had to be perfect, and often they weren’t. “They were awfully hard to make and they broke all the time,” his wife would recall. “He was always hoping there would be something.” Something else, in other words, that could do what only tubes could do.
¹²

I
N THE LATE
1920
S
, work at the tube shop, as well as at Bell Labs, boomed along with the rest of the American economy. In the months after the stock market crash of 1929, when the black depths of the Great Depression weren’t yet apparent, Kelly and a few other colleagues belonged to a buoyant “three-hours-for-lunch” club, a group of Labs employees intent on trying the newest Manhattan speakeasies (Prohibition was still in force) before the police could shut them down. But the business climate grew ever more dire. The astonishing drop in manufacturing jobs and the unrelenting misery in the American farm belt drove down phone subscriptions—and with them AT&T’s revenue. In the course of three years, between 1930 and 1933, more than 2.5 million households, most of them Bell subscribers, disconnected from the phone grid. In 1932 alone, the number of telephones with Bell service dropped by 1.65 million. Western Electric laid off 80 percent of its workforce. The Labs, which had typically hired a few hundred young employees every spring, sending out a team of recruiters to speak with professors at colleges around the country in search of graduate students who might be well suited for industrial research, stopped hiring. And then, with a straitened budget, Frank Jewett, still the Labs’ president, instituted pay cuts and a four-day workweek.

And then Harold Arnold died.

Jewett’s research deputy, forty-nine years old, suffered a heart attack at 3 a.m. on a July morning at his home in Summit, New Jersey. Jewett soon appointed a successor: a tall, thoughtful, experimental physicist
named Oliver Buckley who had spent much of his career at the Labs trying to address the special problems that affected “submarine” cable—that is, cable that went under water, connecting islands to the mainland, and was susceptible to a range of stresses that didn’t affect ordinary underground phone cables. Buckley’s dream was to run a transatlantic cable from North America to Great Britain, a project that the Depression and various technological challenges had placed on an indefinite hold.

Not long after Buckley moved up, Mervin Kelly did, too. Through his work in the tube shop, as both a researcher and production chief, he had extended the life of the Western Electric telephone repeater tubes from 1,000 hours to 80,000 hours, an impressive and cost-saving feat. In 1936, Kelly was appointed director of research. The Bell Labs hierarchy was now established for the next decade: Frank Jewett on top, Buckley below him, then Kelly. Though Kelly was not technically in charge, that mattered little. As events would show, he would lead regardless of his rank or station.