Histories name the scientists. The machines remember the makers. Kelvin sketched; James White’s Glasgow workshop and the Légé firm in London cut metal to thousandths of an inch — and a predictor is only as honest as its worst gear mesh. This is the craftsmen’s story, from the 1872 prototype to No. 2.
White and Légé: teaching brass to add
Kelvin’s 1872 design demanded ten exact gear ratios — compound trains approximating the moon’s awkward 12.4206-hour period. Workshop notebooks record repeated re-cutting: a wheel that looked right but ran minutes off per simulated month went back on the lathe. The builders also invented much of the detailing — the slotted crank pin, the chain guides, the pen carriage — that later machines copied wholesale. When I call the prototype a Légé machine, I mean it the way sailors name ships for their yards.
Fischer: ambition in tons
E. G. Fischer, who turned Rollin Harris’s specification into America’s 37-constituent giant, faced the opposite problem: not delicacy but mass. Thirty-seven trains, thirty-six floating pulleys, a frame that must not flex as the room warms. His answers — jewel bearings, a cast-iron backbone, motor drive — set the template for every later large machine, including the German Gezeitenrechenmaschinen.
The anonymous bench
Behind the named makers stood turners, fitters, and wire-men whose initials survive only in setup logbooks. They lapped bearings, trued pulleys by ear, and knew each machine’s personality — which pinion complained in winter, which wire stretched after oiling. Operators inherited this lore apprentice-style; no manual captured it all.
Three plates I keep coming back to
Nameplates are small biographies. Three have stayed with me from years of museum photographs and catalogues:
“J. White, Glasgow.” James White’s workshop built for Kelvin when the whole idea was still a wager. White’s men had to invent detailing as they went — nobody had ever cut a 73-tooth wheel to stand for the moon before. Look at surviving White gears and you see hand-finishing under the blackening: the prototype was argued into existence as much as engineered.
“Légé & Co, London.” The production partner that turned Kelvin’s second design into the template every later builder copied — the slotted crank pin, the chain guides, the pen carriage described in my prototype reconstruction. Légé instruments have a recognisable hand: compact, dark-lacquered, built like the marine chronometers coming off neighbouring benches.
The Coast Survey plate. America’s No. 2 carries government brass — office, year, contract number — and reads like a declaration of independence from European tide tables. Fischer’s name sits beside the survey’s, maker and office sharing credit the way White and Kelvin once did.
Each plate answers the FAQ below about why nameplates matter better than any paragraph of mine could.
Margins of the manual
Why do maker nameplates matter? They settle arguments: office versus maker, date versus rebuild. My museums guide asks you to photograph every plate for exactly this reason.
Could a good design survive bad building? No — Doodson understood this: a 60-constituent theory needs a 60-constituent-quiet machine, and quiet is built, not drawn.
Are the firms still alive? Mostly absorbed or gone; their ledgers, where they survive, sit in museum archives and repay digging.
Read the nameplate
Every surviving machine carries a small brass plate with two names: the office that ordered it and the firm that built it. Histories quote the first; the machine itself honours the second. Next time you face one — the museums guide tells you where — photograph the plate before anything else, because dates and makers settle more arguments than textbooks. Then watch their work move in Kelvin’s prototype, and meet the women computers whose ledgers completed what the workshops began.