The Secret History of Lead
Enter Du Pont
No longer the shambling, anarchic outfit it had been under the inveterate risk-taker W.C. Durant, GM was now to be run in the militarily precise mold of E.I. du Pont de Nemours & Company of Wilmington, Delaware. Awash in a sea of gunpowder profits from World War I, the du Pont family had been increasing its stake in GM since 1914. By 1920 it controlled more than 35 percent of GM shares and moved to pack the board, installing professional management, with the du Pont faction taking control of the corporation's all-powerful finance committee.
Caught short by a margin call in the recession of 1920, Durant, GM's colorful founder, lost his stake and was forced by the du Pont family to walk the plank (he would spend his final days running a bowling alley). One of the clan's craftiest patriarchs, Pierre du Pont, was coaxed from retirement and named GM's interim president; Alfred Sloan, who had demonstrated the coldhearted allegiance to the bottom line the du Ponts revered, became executive vice president preparatory to assuming the top slot. The pressure on all concerned, including Kettering and his research division, was to make money and to make it fast.
Lest there be any misunderstanding, Sloan wrote to Kettering in September of 1920, alerting him to the du Ponts' new math: "Although [the Research Corporation] is not a productive unit and a unit that is supposed to make a profit, nevertheless the more tangible result we get from it the stronger its position will be.... It may be inferred at some future time...that we are spending too much money down there [in Dayton] and being in a position to show what benefits had accrued to the corporation would strengthen our position materially."
That time would come soon enough for Kettering to deliver. An air-cooled engine he'd championed--copper-cooled, he called it--would soon prove a costly disaster for GM. Fortunately for him, immediately after joining GM he had given his trusted assistant Midgley two weeks to find something to ignite the new management's interest in funding continued fuel research. Though it would take somewhat longer than two weeks to fire their masters' enthusiasm, "Midge" succeeded.
And the Winner Is...
The effect of this sudden time constraint was striking. As GM researcher and Kettering biographer T.A. Boyd noted in an unpublished history written in 1943, Midgley's main research in 1919-20 had been to make alcohols out of olefins found in petroleum through reactions with sulfuric acid. (Farm alcohol was one thing, but a patentable process for production of petroleum-derived alcohol--a possible money-maker--was quite another, one of considerably greater interest to the corporation.) "But in view of the verdict setting a time limit on how much further the research for an antiknock compound might continue," Boyd said, "work was resumed at once in making engine tests of whatever further compounds happened to be available on the shelf of the lab...or which could be gotten readily."
As noted earlier, Midgley tested many compounds before isolating tetraethyl lead in December 1921. In the early days, he would attribute the discovery of TEL's antiknock properties to "luck and religion, as well as the application of science." In a 1925 magazine article, he would recall false trails with iodine, aniline, selenium and tellurium before hitting upon lead. Curiously, his article omitted any reference to the alcohol-gasoline blend he'd patented just five years earlier.
Another oddity: The exact number of compounds tested prior to TEL's discovery varies dramatically in different accounts. As Professor William Kovarik of Radford University has observed, confusion reigns in part because the lab's day-to-day test diaries have never been released to the public by the General Motors Institute (GMI) archive. In the words of one archivist there, GM's lead archives have been "sanitized." One 1925 article in the Literary Digest put the number at 2,500 compounds tested, while The Story of Ethyl Gasoline, a 1927 pamphlet released by a company Midgley would help found, states that 33,000 were studied. Another time, he claimed 14,991 elements were examined, while a 1980 Ethyl corporation statement set the number at 144. This question is important because GM's discovery of lead's antiknock properties, which initially caused little internal excitement, would be hailed in popular media and later cited in polytechnical texts as a model of rational, orderly scientific inquiry that sought the single best answer to the knock question. A more realistic view of events is that TEL's re-emergence in the twenties was the result of a crude empirical potshot that was understood to promise a landslide of earnings over time.