Out of the Past, April 1995

Delving through some papers the other day I came upon a dirty little notebook which appeared to have spent much time on a workshop floor, the sort of thing one’s wife wants instantly to throw away.

Apart from a rendering of the poem “The Green Eye of the Little Yellow God” and other verses, perhaps self-composed, the notebook contained short reports of road-tests carried out between September 17, 1919 and June 29, 1920.

Maybe these brief long-hand entries were later transferred to works report-sheets. Anyway, they covered 92 vehicles, all numbered, although these numbers did not run quite consecutively. The first number was 27, the last FT278. There were few gaps, except that the final number was preceeded by FC4 and FC6. These were chassis numbers, but the only other data recorded was confined to make of engine and its number, make of carburettor and magneto, and sometimes jet settings, followed by brief test comments, all in pencil. The name of the chassis erector was occasionally quoted, usually J Brown, and the tester’s initials appear to have been “FB”.

The puzzle was to try to decide the make of these vehicles. Clues were few. Alone on one page bore and stroke are given as 120 x 60, which 1 had to assume should have been 60 x 120mm, together with the tappetsetting (6 thou) and firing order (1243). The early vehicles had Perkins power-units, but from March 1920 these alternated with Coventry-Simplex engines. 1 began to think in terms of commercial vehicles, even railcars, when one entry referred to an “off-side rear axle-box” as rather noisy.

But as the tests took in hill-climbing and seemed to be especially concerned with noisy axles, I felt perhaps they were done on roads, not rail-tracks… Searching further, in the faintly pencilled-in comments interspersed with those poems of the period, I discovered that the chassis frames had been made by Thompson’s, that Zenith carburettors were favoured, with a very occasional SU, and that ML magnetos giving a fixed ignition-setting were universal, except for a Fellows on one engine. That the vehicles were small is obvious from many references to 35 to 38 mpg, sometimes as much as 42 mpg, and speeds of 38 to 40 mph, although some vehicles were faster and chassis 79, Perkins-engined, was entered as “very fast, 45 to 47 mph”, and fortunately this chassis was also “flexible, good on hills and brakes good”. In March 1920 a Colonial model was tested.

It had by now become obvious that the transmission was by friction discs, but this was hardly conclusive identification of the make, because more cars than might be imagined had this inexpensive, crash-free, infinitely-variable (except for the gearlever slots) means of obviating a conventional gearbox, at this period. Originally the discs on the cars we are investigating were of composite material but by 1920 grey fibre discs were being tried and also a new type of radius-rod axle and new brakes Noise from the back axle was the major problem apparently, but the discs could be “singy” and once a n/s front pivot-arm came loose and had to be brazed. We learn that the Coventry-Simplex engine, bench-tested, ran up to 2600 rpm and that a disc had to be changed on one test, a Perkins engine on another. Fierce brakes were noted on one car. But you get the gist?

As for deciding what cars were being tested, I was as far adrift as when I started looking through this little book, although the use of Bisham Hill might have provided a clue. Then I came upon the decisive, or nearly decisive, answer. Chassis No 169 “was rather noisy, but was passed by Mr Keiller”. Ah, so we are dealing with GWKs, for C M Keiller was one of the directors of that concern, with Arthur Grice and 1 Talfourd-Wood — hence “GWK”. If any keen GWK owner has a car with one of the afore-quoted chassis numbers and likes to drop me a line, possibly I can tell him how it performed on test as a new vehicle.