The Lady of Lyons Escapade and Tuplin's imagination.

One of the more controversial incidents in Railway History is the high speed test of GWR 2903 in May 1906. This somewhat shadowy event has been subject to imaginative tales and wild exaggeration, and sometimes the wildest exaggeration has been from those who seek to debunk the tale completely.

Railway speed records are odd things anyway. There was no sporting body to independently authenticate records as there were with outright land, water and air speed records, and by and large the rail records involved hurtling down hills, and would not stand up against the conditions imposed for the outright records. So it's rather dubious to regard some records as authenticated and others not.

Anyway, back to Lady of Lyons. The raw facts are these: It was normal to give newly built and repaired locomotives a high speed test. What high speed was depended on the locomotive of course, but the general practice was to run the engine out from the works at a moderate speed, check all the bearings etc. for signs of distress, and then if all was well run back at high speed.

For some unrecorded reason on this particular day several senior works staff members were on board the test run of a new Saint, and they had decided to see if a new locomotive could be run at over 100mph. They succeeded in this, and timed the locomotive in the region of 120mph, but not in any way that would be satisfactory for claiming a record. And that's it, the whole story!

From this all sorts of castles in the air have been constructed.

The first thing to understand is the nature of speed timing. No-one ever really records an exact speed, it will always be a range subject to errors. The better the equipment then the smaller the range. When Rouse-Marten timed City of Truro down Wellington bank he was using ⅕ of a second stop watches against ¼ mile posts, so the very best he could do was record speed to about plus or minus a bit over 1 mph, even if he didn't click the stopwatch a fraction too late or too early and record the wrong ⅕ of a second, which would be plus or minus a bit over 3mph. When Mallard was recorded at 126mph they were using a paper trace on a dynamometer car, which is far more accurate, but even then there are those who consider that claiming over 125mph was debatable.

History doesn't record how Lady of Lyons was recorded, but we may guess it was perhaps a 1 second stopwatch, and it was certainly being operated standing up in a crowded and wildly shaking cab, in which case an error range of plus or minus 10mph is very feasible. It's easy to see why those on board didn't feel able to claim any sort of record. We should also note that 120mph is a very round number. Another part of the legend comes from the signal box timing of 2 minutes over 4½ miles. Signal box timings were to the nearest 30 seconds, so even if the clocks were perfectly synchronised the best that time gives is plus minus about 35mph, and a 30 second error makes it plus or minus 100mph.

But before we get onto the legend, perhaps it's best to go back to the source. This is the episode as it was first reported in the Railway Magazine of April 1932.

Two Miles a Minute

During January last a statement obtained wide currency in the daily press that Mr H J Robinson, then just about to retire from the position of Chief Locomotive Inspector on the Great Western Railway, had been responsible for driving a locomotive in this country at a speed of 120 miles per hour. It is needless to say that readers of The Railway Magazine who are familiar with all the speeds hitherto claimed as railway records, and in particular with the figure of 102.3 mph achieved down Wellington Bank of the GWR on May 9, 1904, which from that day to this has had an unchallenged supremacy, are interested to know on what authority this new claim has been made, as is evidenced by the extensive correspondence we have received on the subject. We therefore wrote to Mr C B Collett, the Chief Mechanical Engineer of the Great Western Railway, who communicated to us an interesting account of what actually occurred. It appears that in May, 1906, No 2903 - one of the newly introduced 2-cylinder 4-6-0 locomotives and herself fresh from the shops - was taken for a trial run light from Swindon to Stoke Gifford, with the intention, after running the engine round the Filton-Patchway triangle, of having "a sharp run" back. Signal checks were experienced, however, and No 2903 was then stopped at Chipping Sodbury until "line clear" had been obtained through to Wootton Bassett, after which she was re-started, and there was evidently some running of very startling order down the 1 in 300 from Badminton to Little Somerford.

The purpose of the run was to demonstrate that an engine taken straight from the shops could be run at over 100 miles per hour. Those on the footplate included Mr Collett, who was then Assistant Manager of the Locomotive Works, Mr G H Flewellen, who was the Locomotive Inspector, and the Foreman of the Erecting Shop, Mr Evans. The timing for some distance by the mileposts with a stop watch was given as 120 miles per hour, and the clocking between the signal-boxes of Little Somerford and Hullavington was booked as two minutes for the 4½ miles.

Mr Collett points out that, while the object of running a new engine on its first trip at over 100 miles per hour was achieved, the timing could not be regarded as accurate and that the 102.3 mph record of 'City of Truro' in 1904, made under the personal observation of one of the most careful recorders of his time - the late Charles Rous-Marten - with the aid of a chronograph reading to one-fifth parts of a second, must remain the best duly authenticated railway speed record that this country has yet witnessed.

This rather measured report is the fullest first-hand account we have. It's certainly much more than a "Big fish" story, but it's clear that no-one could claim any kind of record on this basis.

And now for the legend. The wildest stories come from W. A Tuplin, who was a trained and competent engineer in another branch of the profession, but was inclined to let his enthusiasm run away with him when confronted with a steam engine. This account is from his book "Great Western Saints and Sinners", and I have chosen to annotate it at regular intervals. To be fair to Tuplin he says at the beginning of his flight of fancy that its conjecture, and at the end that it's a guess, and yet it tends to be treated far more seriously than it deserves.

The next incident to be recorded in this series was one of the most extraordinary happenings ever admitted by a responsible official of a British railway. It is important in that it enables one to say with reasonable certainty that the highest speed by steam on the Great Western Railway was attained by two-cylinder 4-6-0 No. 2903 in May 1906. The figure quoted was 120 M.P.H. and this was confirmed in a very rough way by the recorded passing times at two signal boxes; these worked out at about 135 M.P.H. [Many critics take this mention of 135 M.P.H., surely an impossible speed, as evidence to debunk the whole episode. But as we have seen this timing has such a wide error range as to be almost meaningless.]

It was regular practice to give engines newly-built at Swindon a running-in trip from there to Stoke Gifford and back, a total distance of about 70 miles. This test-track was used instead of the Paddington-Bristol main line because there was less traffic on it and because the triangle of lines at Stoke Gifford was a convenient means of turning engines for the return journey to Swindon. It includes a straight descent of 9 miles at 1 in 300 to Little Somerford, followed by a rise of 6 miles at 1 in 300 leaving plenty of room for stopping. It was usual for every engine that behaved normally on the outward journey to be run pretty fast on the way back.

So far so good, but the 1906 incident started a rumour of 'two miles a minute' and this became so persistent a legend that eventually someone persuaded Mr C. B. Collett to admit that a group of 'high-ups' at Swindon had indulged in a bit of horseplay.[Tuplin's imagination is already at work: no-one said anything about horseplay!] The Railway Magazine for April 1932 states on p. 305 that, on the evidence of Mr Collett, [Here Tuplin reproduces the two paragraphs based on Collett' letter already quoted above. He then continues with his own interpretation/imagination ]

As GW engines at that time were never required to exceed 90 M.P.H. in ordinary service and were put on fast jobs only after running for a week or two on slow ones, there was no technical need to know whether a newly built engine could reach 100 m.p.h. So why was it desired to know it? Had anyone worked out at what speed the balance weights of a Lady would lift her driving wheels off the rails in every revolution?

They wanted to know whether the Lady could run at over 100 M.P.H. on her maiden excursion. But if that was all, why go up to 120 M.P.H. where out-of-balance forces were nearly half as big again? Had they developed a kind of mob-hysteria that led them to urge the driver to go 'all out' regardless of everything? By the time they had run more sedately on to Swindon and the disciplined serenity of the Works they probably agreed that not a word of this sporting venture should be whispered to anyone. No intention of this kind is ever more than a pious hope and every student of the steam locomotive must be glad that an official statement was eventually made.

It was most imprudent to allow so many officials to participate in what was undoubtedly a risky exploit, and these sporty souls would naturally pick a day when Churchward was away from Swindon

Had he gained any hint of such intention he would have forbidden it for one obvious reason in some such terms, as 'If you go and get yourselves all killed, where the bloody hell am I? [All the above is imagination! Mob hysteria? Secrecy? Churchward kept ignorant? There's not a shred of evidence to support any of that. We should remember that even 3440's high speed run was kept quiet from the public]

What really did happen on this extraordinary occasion? No one concerned would say a word while the incident was fresh in any- one's mind although it was admitted that a speed of about two miles a minute was reached. One is justified in resorting to conjecture and the fact that the engine had pole reversing gear, which is not safely adjustable at speed, makes one wonder whether this was a factor.[Conjecture is always fun, and makes for a good story to sell books, but how much conjecture is really justified? Key to this flight of fancy is lever reverse, but I am informed that on a piston valve fitted locomotive, as 2903 was, the lever does not spring into full gear as described. This is apparently only a phenomenon of slide valves]

It is possible that the driver had found that his original pole- setting was not producing 100 M.P.H., and was thereupon persuaded by one of the officials to try a little later cut-off. If he were rash enough to attempt this (or if someone else said 'Here, let me do it') the engine might well drop into full gear and accelerate like mad. Everyone might then be so appalled by the exhaust noise of such running as to do nothing at first to neutralize it, until by the time the driver had recovered from the shock of having the pole pulled out of his hand and had got round to turning the blower on and shutting the regulator, some unabashed spirit said, 'No! Leave her at that. Let's see what she'll do."

Or did the driver try to close the regulator and find that he couldn't? A big flow of steam from the boiler will sometimes cause such a pressure drop from one side of the regulator-valve to the other that the valve becomes very hard to move and if an engine without a train is running very fast downhill with steam urging it on, its brakes won't stop it quickly, if at all.

If anything of this sort happened on Lady of Lyons, already doing nearly 100 M.P.H. down 1 in 300, there could have been plenty of lively apprehension on the footplate. With driving wheels leaving the rails eight times per second, with the engine using water so fast that the fusible plugs were at risk and with the impossibility of stopping in any distance less than about six miles, someone had to do something. There were plenty of people there, and so two could pull at the reversing pole and two at the regulator handle while someone else made sure that both injectors were working and then, still remembering what they had come for, took some mile-post-passing times. [All this flight of fancy is founded on the belief that the lever reverse could slam into full gear, but as noted I am informed this doesn't happen with piston valves.

Evidently they did get things under control at last, with 100 M.P.H. well and truly exceeded and everybody on board still shaking. The sensible ones would realize that they'd only had what they'd asked for and had been lucky to get away with it. After the white faces there would be metaphorical red ones, and no desire to admit to anyone what danger they had produced for themselves and the engine.[I submit that the piling of imagination upon supposition upon speculation has reached comical proportions!]

This is only a guess.[to say the very least!] What a pity that they didn't have anything like the black box that aircraft now carry to take a record of what went on during alarming last moments! The nearest equivalent to it was provided by the signalmen at Hullavington and Little Somerford. Their evidence of roughly 135 M.P.H. tended to show that Collett's report of about 120 M.P.H. was not an exaggerated one.[I hope I've demonstrated that signal box timings show nothing of the sort.] but he himself admitted it and disclaimed it in the same sentence. An interpretation of this is that the speed was attained but in circumstances that did not do much credit to those responsible for the exploit.[A very imaginative interpretation]

I think this is enough to demonstrate that even though not much about this incident is known we need to be very careful about the mythology. I have seen Tuplin's faintly ridiculous flights of fantasy reported as facts.

Jim Champ, 2021

Return to the GWR Information Index Page.

The original articles on these pages are mainly contractions from my book, "An Introduction to Great Western Locomotive Development", a study of Great Western Railway locomotive classes, which is published by Pen and Sword Books. You may order it from here.

To contact the writer please use this form, or if you prefer use Linked In or even Facebook.

© Jim Champ, last edit 12 Oct 2021

Search the site for with

HTML check . CSS check