The World's Oldest Recording:
Frank Lambert's Amazing Time Machine

By Aaron Cramer with Allen Koenigsberg

The name of Frank Lambert appears in no reference book and if you were to mention him to any phonograph or record collector, you might get the response: "Didn't he make those bright pink celluloid cylinders?" But that was, of course, Thomas Lambert! If you tried again, perhaps with a more advanced collector, you might learn that Frank was the inventor of the device which gave the G&T record company its expanded name from 1900-1907 (Gramophone and Typewriter). And that would be at least partly correct. But this isn't an article about his typewriters - it's about an unknown machine that he built, years before Bell and Tainter, which was able to preserve sound on the first permanent removable record. And in one of the most remarkable tales of chance discovery and subsequent sleuthing, the original machine has been found, and the recording, maybe the world's oldest, has been played and can be clearly understood! But let us begin...

At first thought to be an interchangeable sleeve for tinfoil sheets, the 2-5/8" engraved lead cylinder was found to contain the world's oldest playable record. The recording head is visible at left front, the reproducer at the rear.

Little is known about the life of Francois Lambert (for that was his original name) before he arrived in the U.S. in 1876. A small clue found in the postmark on a relative's letter has now been verified: he was born in Lyons, France on June 13, 1851, and was given the same name as his father, a maker of shawls; his mother's name was Vincelette Fayard. Lyons was a center of textile production (especially silk) and it was there that the Jacquard loom had been first invented. But it seems more likely that Frank, as he would soon become known, was apprenticed to a local machinist.

When he arrived in the U.S., the country was celebrating its first Centennial, and among the new devices exhibited in Philadelphia were Bell's telephone and the first practical typewriter. We can not yet prove that he attended the Philadelphia Exposition, but the inventive bug soon bit, and in a remarkable coincidence, he applied for his first patent (on a "Striking Mechanism for Clocks") the same week as Edison applied for his phonograph and received it (200,518) on the same day as Edison, Feb. 19th, 1878. Lambert shared this first patent with Walter Davies of Brooklyn, NY, who himself would gain a series of patents on striking mechanisms with others in his family, Henry and Edward. As a result of his affiliation with Davies, Lambert soon relocated to Ansonia, CT, a center of clock manufacture, but then moved back to Brooklyn by June 1880.

Sometime in 1880, while living on Grant Street as a boarder, Lambert met and soon married Jeanne-Marie Donval, a woman a few months older than himself. She already had a daughter, Julia Ida (born Dec. 27, 1876), a son Alexander (b. 1879), and may have been widowed from a furrier named John Simonet (the legal question of Julia's "adoption" came up years later in a court battle over his will). Frank and Jeanne-Marie had five children of their own: Eugenie, Frank jr, George, Martha, and Jeanne. Only Jeanne, who died childless in 1965, survived her parents.

While Lambert was living in Ansonia, he met a man who would have a profound influence on his life - John Thomson. In 1883-1884, Lambert assigned one half of his first typewriter patent to Thomson (who had originally been a clock inventor) and more prophetically, shared a patent with him on water meters in 1887. Lambert continued to improve fluid-measuring devices and together they formed the Thomson Water Meter Co., whose facilities were located at 100-110 Bridge Street in Brooklyn (now torn down). When Thomson was ailing in May 1925, the company was sold outright to the Neptune Water Meter Co. and Lambert received $800,000 in cash.

The invention for which he was best known, at least to modern collectors, was the typewriter that bore his name. It did not use a standard keyboard but was rather classified as an Index style (typewheel). An American company was organized in early 1900 (Lambert had become a U.S. citizen in 1893), but the foreign rights (except for North and South America, the Philippines, and the Sandwich Islands) soon went to The Gramophone Co., Ltd. of London, in an idea of Wm. Barry Owen who planned to sell them at the same price as the "trademark" Gramophone; the company was renamed to include the typewriter. Despite the precaution of patenting it in Sweden, India and New South Wales (among other places), the well-made machine was not widely successful although Lambert did receive $20,000 in advance royalties. G&T discontinued regular sales by 1904 (after making approximately 10,000 units in England) and reverted to their old name in 1907.

The mid-twenties were pivotal years for Lambert; his invalid wife died on Sept. 20, 1925 and Thomson would die in June 1926. However, for a 74-year old, he bounced back with a great deal of vigor: on Jan. 13, 1926, he married (at the NYC Municipal Building ) Jeannette Justine Lawson Ebbets, who was 44 years his junior; she had originally sold sheet music in his son-in-law's office and learned the real estate trade. He then left on almost a year's honeymoon throughout Europe. When he returned, he appeared to throw off his Brooklyn roots, and at his wife's urging, moved from his large house (not standing today) at 192 Prospect Park SW to the 22nd floor of the Savoy Plaza Hotel in Manhattan. Her mother and two brothers accompanied them. His final years (after 1932) were spent with his second wife at the elegant Waldorf-Astoria on Park Avenue.

Although there is no record of any invention credited to him after 1925, his other interests continued. He had long been an avid collector of shells, gems, and minerals. He owned several rental properties in Brooklyn, and some undeveloped land on Long Island , and when he died in 1937, his estate amounted to over a million dollars. He may not have been well-known (no obituary appeared), but he was certainly not the stereotype of the impoverished inventor.

Several years after Lambert started his new life at 74, he apparently rewrote his will, leaving the bulk of his estate to his second wife. Only the income from a $20,000 trust was left to his daughter Jeanne and the interest on $30,000 to his only grand-daughter Martha Emily Gillott. This aspect of his life gained journalistic attention as the NY Times trumpeted (8/19/1937): "Lambert's Millions in Will Contest." Jeanne and Martha were each offered a small outright settlement in addition as a result of the legal action.

Apparently the widow (who never remarried) had a head for business, for when she died in 1975 (but without a will herself), her estate was valued at over $15,000,000 (at least on paper)! She had been something of a recluse at the end and court documents speak of wading through piles of papers "several feet deep". Her niece, Jeannette Veronique Minturn (a landscape designer at Rockefeller Center), would eventually die in reduced circumstances (ca. 1991), leading to the curious events surrounding his lost invention.

The machine in question (see illustration) was acquired from an antique dealer who had bought the contents of a sealed storage room (at auction) from the Public Administrator when the fees could no longer be paid. Most of the contents, ranging from a grand piano to letters, paintings, and books, had been placed there between 1927 and 1937. And most enticing of all, according to his niece's diary were the models of his inventions - water meters, typewriters, gyroscopes, etc.

At first, I thought the heavy, oddlooking machine incorporated an obvious improvement over the Edison tin foil phonograph, that is, the foil could be wrapped around a removable (pre-grooved) metal sleeve. But when I brought the machine to Peter Dilg at the Baldwin Antique Center, he immediately guessed that the grooves (80 per inch) were not a guide for the foil at all, but themselves contained a recording, directly cut into the lead! Peter then used the steel recording point to make a new groove on a separate brown wag cylinder. When he had fitted a reproducing stylus to that track, he knew he had one that would fit the lead sleeve. He soon devised a delicately-adjusted electrical pick-up, and for the first time in over 110 years, as we rhythmically cranked away, these unexpected words eerily came forth:

One o'clock, Two o'clock, Three o'clock, Four o'clock, Five o'clock, Six o'clock, Seven o'clock, Eight o'clock, Nine o'clock, Eleven o'clock, Twelve o'clock.

("Ten o'clock" was omitted, perhaps because Lambert was still a relatively new immigrant).

There are other words too, including even a track going in the opposite direction, but they cannot be clearly understood at the present time. Some of the grooves appear to be damaged, or were perhaps over-cut, as this is the first phonograph equipped with a shaving device to erase a previous recording. The recorder used a tin diaphragm .008" thick and 2-3/8" in diameter and the reproducer diaphragm would be the same, presumably. The recording stylus is a finely ground steel cutting tool, with a draft at the edge to allow removal of the lead swarf. The top enclosures seem to be gutta percha and are threaded into brass mounting rings.

The machine base itself is solid 6" by 11" milled steel, and the length of the shaft and winding wheel is 15" the overall height is about 7". A quick release lever allows the main threaded shaft to be repositioned immediately after recording. The total weight is about 22-1/2 pounds.

I invited Allen Koenigsberg to look over the machine and together we tried to get into the mind of Frank Lambert. Allen agreed that the device was very early (1879-1880), probably made not much after the first tinfoil phonograph. It made no sense for Lambert to use such heavily machined materials after the introduction of wax cylinders later in the decade. And why did Lambert continue his testing beyond a traditional "one, two, three"? It could only have been done for the full sound track necessary for a talking clock! This supposition was somewhat confirmed when Allen compiled a detailed list of Lambert's 60 patents. There, at the beginning of his inventive career (in the late 1870's), were two for clock devices.

Another line of research recently opened up. It is relatively easy to consult published patents, but what happens to the ones that were never granted in the first place? I found out over the summer when my wife Thea and I visited the National Archives in Suitland Maryland. The ungranted applications are generally not saved, but since they were accompanied by a down payment of $15, Patent Registers were maintained, listing serial number, name of inventor, date, title of invention, etc. In the time available, we were only able to go through the year 1880 (the names are not fully indexed alphabetically for that period). But there, two intriguing possibilities emerged, two Lambert applications that were never granted: the first was for the vaguely-worded "Clocks" (filed on June 17th) and the second was for "Screw Cutting Lathes" (filed on Aug. 2nd)! Could it be that these cryptic references hid what Lambert was really doing?

With the purchase of the machine, I was also able to obtain many of Lambert's personal effects, including some original patents, business letters from his patent attorneys (including pioneer Edith Griswold), die stocks, scrapbooks and photos. Amongst nearly 2000 newspaper clippings of poems and poetic sayings, I found a curious one from about 1896. It was a two-inch announcement about a talking clock, but apparently that of the Swiss inventor Casimir Sivan, who employed vulcanite discs to announce the time. Did this stir his memories?

At this point, Allen recalled that Edison himself, in the optimistic days shortly after the invention of the phonograph, predicted that his creation would be used in talking clocks and said so in a published article in the May-June 1878 issue of the North American Review. As a matter of fact, he even arranged a contract to produce one (on Jan. 5, 1878) with Daniel Somers and Henry Davies. That must have been the link, for it was to the same Henry J. Davies that Frank Lambert assigned one half of his first granted patent!

Edison had planned a copper-plate, lever-wound spring phonograph which was successfully built in Ansonia and described in a letter to Alfred Mayer on Feb. 11, 1878. Would it be impossible that either Davies or Eugène Pastre (a friend and co-patentee of Lambert also) asked Lambert to make a cylinder model as well? Both Davies and Pastre worked for the Ansonia Clock Co.

And now, as we go to press, we have located a little known court case through the kindness of researcher Ray Wile. It was filed against the Edison Phonograph Works by the American Graphophone Co. in 1892 and contained some startling testimony from several experts in the field (U.S. Circuit Court, Dist. of NJ, In Equity, No. 3500). Summoned to the Edison table in early April 1896 as a defense against Bell and Tainter's patent 341,214 was none other than the President of the Thomson Water Meter Co., one Frank Lambert. In answer to several detailed questions, he testified under oath that he had indeed built a recording machine "around 1879", and that he had actually engraved the sound vibrations on a permanent metal sleeve. All of our guesses and suppositions had miraculously been proved true!

It is now clearer than ever that when Bell and Tainter embarked on their phonographic endeavors at the Volta Laboratory in Washington, DC, they may have had another motive for delaying their patent applications and sealing their original work in a metal chest (Oct. 20, 1881) deposited with the Confidential Archives of the Smithsonian Institution. The Volta group may have feared Edison's legal wrath over the meaning and use of the word "indenting" as they converted an Edison/Bergmann tinfoil phonograph to a wax covered Graphophone; or they may have even have doubted their contractual rights to a non-telephonic invention; but how could they deal with a lead-sleeve phonograph which engraved sound vertically, and was built prior to their own? Although Lambert said he could not locate his original machine in 1896 and had destroyed his only duplicate, Columbia found it in their best interests to settle and cross-license Edison in December 1896.

There might be other candidates for older "recordings", like Leon Scott's fabled phonautographic tracings of Abraham Lincoln at the White House (1863) or the still-extant foil strip of the aged voice of scientist Joseph Henry (1878).

But primitive sheets of sound like these have never been restored. Until that day arrives, the rhythmic words of the modest French immigrant who settled in Brooklyn constitutes the oldest playable record in the world.

Frank Lambert died in New York at the age of 86 (from pneumonia and uremia) on June 21, 1937, and was buried in a solid copper casket at Green-Wood Cemetery in Brooklyn. He lies there in good company today, among the likes of Samuel Morse (telegraph), Elias Howe (sewing machine), Walter Hunt (safety pin) and Louis C. Tiffany (stained glass). But there is a final mystery, even in death, surrounding him. Although Green Wood has verified that Frank Lambert is indeed buried in Lot #28721 and his first wife and three of his children have their names clearly engraved on the handsome marble tombstone, his own name is not to be found at all. An omen, no doubt, of the obscurity into which he would soon unjustly fall.

Our thanks to Ruth Edge at EMI Archives, Peter Dilg, Allen Koenigsberg, Thea Cramer (above and beyond the call of duty), Daniel Marty, Jean-Paul Agnard, the National Archives (James Cassedy, Greg Bradsher, Robert Morris, John Silardo), Ray Wile, the Musee de Lyon and Theresa La Vianca at Green-Wood Cemetery.

This article first appeared in Antique Phonograph Monthly, Vol.X No.3.
We are grateful to Allen Koenigsberg for allowing us to reproduce it here.

This issue of APM, and a few limited others, are still available from Allen.
Contact him via his web page at

Image Copyright © 2006 Peter Liebert

Back to main Aaron Cramer Tribute page