INSPIRATION IS AN ELUSIVE THING . Sometimes it comes in a dream, as when Friedrich August Kekulé’s vision of a ring of dancing atoms, like a snake biting its tail, revealed the structure of benzene. Sometimes it is a simple mechanical analogy, as when the motion of a ship’s wheel led Samuel Colt to design his famous revolver. Much more often, innovators don’t know where their crucial insights come from. But, decades after he found a way to make light bulbs three times as efficient, William Coolidge remembered exactly where he got the idea: in his dentist’s office.

Thomas Edison’s light bulb, invented in 1879, was an epochal advance, the first electric bulb practical for home and industrial use. Still, it was far from perfect. One of the biggest problems was its filament, which was made (beginning in 1889) from carbonized cellulose and which gulped power at a wasteful rate of three watts per candle.

In 1905 General Electric’s pioneering research laboratory in Schenectady, New York, hired William Coolidge to attack this problem. European bulb makers had started using metal filaments, and Coolidge set out to improve on their work. Tungsten, which has the highest melting point of any metal, looked like a promising material. Unfortunately, it is brittle and exceedingly difficult to draw into wire. GE already had tungsten bulbs on the market, made with a licensed German process, but they were expensive and so fragile that they could not be used in railroad cars. A better solution was needed.

In 1964, at the age of 91, Coolidge recalled where the effort to make ductile tungsten stood when he arrived at GE. “A typical method started with tungsten in powder form, to which starch paste or other sticky organic material was added. The mixture was then squirted through a die into a fine thread. This thread was then heated, in a suitable atmosphere, to remove the temporary binder. Further heating to a higher temperature sintered the particles of tungsten together.” But there was a problem. “The organic binder … left a trace of carbon in the tungsten filaments; and this carbon, in the lamps, vaporized out and blackened the bulbs. To avoid this difficulty, what could we use as a binder which didn’t contain carbon?”

The answer lay in Coolidge’s recollection of watching his dentist at work. Dental fillings are prepared by dissolving, or amalgamating, silver in mercury, usually with smaller amounts of other metals. The result is a conveniently moldable substance that conforms to the shape of a cavity and quickly hardens in place. “Some little time before coming to Schenectady,” Coolidge explained, “I had watched my dentist prepare silver amalgam for one of my teeth, and had been impressed by its plasticity. Was it possible that some amalgam might serve as a temporary binder for the tungsten powder?”

After much development, Coolidge perfected a process in which tungsten powder was dissolved in cadmium amalgam and the resulting mass was forced through a small hole to form a wire. When the wire was heated in a vacuum, first the cadmium and then the mercury evaporated, and the remaining particles sintered together as pure tungsten. This could be rolled and pressed mechanically and then drawn into strong, ultrathin wire. Soon GE was using a modified version of this method to mass-produce inexpensive light bulbs that used one-third the power of the carbon-filament type. In the final version, pressed tungsten powder was sintered directly instead of starting with amalgam, but as Coolidge recalled, “it all followed, step by step, from the memory of that silver amalgam used by my dentist.”

Light bulbs were far from the only use for Coolidge’s ductile tungsten. It gave a great boost to the emerging technologies of radio and electronics. Coolidge went on to develop an x-ray tube that used tungsten wire for the cathode and a piece of sintered tungsten for the target. These advances made x-ray tubes cheap and reliable enough to be used in doctors’ offices. Finally, in 1923, things came full circle when the Victor X-Ray Corporation of Chicago, Illinois (which soon became part of GE), brought out the first compact, safe, and easy-to-use dental x-ray machine. As today’s patients sit in the dentist’s chair, then, they benefit from an inspiration that another dental patient had in the same situation nearly a century ago.