The Years of Experimentation 1920-39

By Stephen Siddle

During the first two decades of the 20th century, Gauge-0 was the smallest recognised size of model railway, and there was no British based manufacturer of commercial model railway equipment. The leading British name in the trade, Bassett-Lowke, commissioned often remarkably accurate Gauge-0 and Gauge-1 models from the great Nuremberg toymakers, especially Bing with whom relations became very close.

But after World War One there was a general revulsion in Britain against the purchase of German goods. As a result, in 1920 Britain's biggest toymaker, Meccano Ltd., launched a new range of Gauge-0 tinplate trains: Hornby Trains; anti-German feeling was one factor in this decision, but the British market proved well able to support a domestic manufacturer and over the next two decades the hobby in Britain was to benefit immeasurably from the availability of Hornby Gauge-0.

These new Hornby Trains were a clear threat to Bassett-Lowke's position in the 'Indoor gauges' and the next development was almost certainly intended as their response. According to a magazine interview in September 1922, Bassett-Lowke had been contemplating the introduction of a new, much smaller, gauge as early as 1914; indeed according to his son in law, Bassett-Lowke's engineer Henry Greenly had gone so far as to draw up a standard working sheet of principal dimensions, including a scale of 4mm/ft and a track gauge of 5/8". The outbreak of war had killed the project, but now it was revived.

In 1919 W. J. Bassett-Lowke's lifelong friend Stephan Bing had become managing director of Bing-Werke AG, and the following year Bassett-Lowke asked Bing to produce a new 'table-top' toy railway to approximately half the scale of Gauge-0 for Bassett-Lowke to market in Britain. The actual design of the table top railway system was the work of Henry Greenly, in collaboration with Oswald Fischer of Bing.

It is important to grasp that Henry Greenly was not a model-maker but a professional engineer who had made his career in miniature railways and model engineering. When Bassett-Lowke produced a live-steam engine or Greenly published a design in the Model Engineer, the principal requirement was that it worked well, not that it be an exact scale replica of a full-sized locomotive. Consequently many of Greenly's designs were more or less freelance, and when we reach his largest locos, for 15" gauge, it is difficult to say whether we are dealing with a freelance model or an independent prototype design.

Greenly seems to have carried this approach over to the design of the Table Top Railway, for which he specified wheels 5mm wide running on 5/8" tinplate track - that is track half the width of Gauge-0. The new system was far smaller than any working toy train hitherto produced and the grotesque wheel profile adopted was presumably cautious engineering intended to ensure that the new system worked reliably in the hands of children. The trains were stamped 'Foreign-made' and marketed under the Bassett-Lowke brand to avoid anti-German feeling. Publicity began in the autumn of 1922, and the first sets were available in the weeks before Christmas.

The Table Top Railway, forerunner of the world's most popular sizes of model railway, was only a moderate success. An electric version (centre three-rail with trip reverser) appeared in 1924, in which year Bing released the system under their own name in Germany, but although it was copied by JEP in France (under the name Mignon using a gauge of 16.5mm), and one or two Nuremberg toymakers, the range did not develop, and it does not seem to have made inroads into the market for Gauge-0. In August 1932, in the depths of the Depression, Bing-Werke's financial difficulties resulted in its collapse and production ceased. The tooling subsequent passed to Karl Bub who restarted production in 1934 for the German market but this in turn ceased at the outbreak of war.

However, by 1924 Henry Greenly had obviously decided that this new small size had significant potential. In his famous book,  Model Railways, published In May 1924, he writes in the section on scales: "Gauge No. 00, 'Table Railways':- This standard gauge has recently been introduced by the writer at the instance of Mr. W. J. Bassett-Lowke to provide for those who are limited in space to that of an ordinary dining-room table. Clockwork and electric locomotives are supplied. The actual gauge is 16mm (5/8") and the scale is 4mm to the foot."

The last statement must be regarded as putting down a marker for the future, rather than as an accurate description of the Table Top Railway, as the little tinplate 2-4-0Ts and their coaches were toys with only a faint resemblance to any prototype.

Why 4mm/ft ? The track gauge was half that of Gauge-0, so logically the scale should also have been halved: ie. 3.5mm/ft. Drawings for 7mm scale modellers were at that time being published to this 'half-size', while some modellers had experimented with 1/8in scale (3.175mm/ft) prior to World War 1. In subsequent correspondence Greenly was to state "I submit that the gauge is not the correct method of arriving at the scale". Yet in 1903 Greenly had done precisely that to establish scales, wheel, and track standards, appropriate to Marklin's Gauges 0, 1, 2 and 3. What caused this U-turn?

Greenly had spent much of the intervening twenty years designing live steam engines of all sizes and had become thoroughly used to playing fast and loose with scale and gauge to make them work. In his book on live-steam engines he stated that a gauge wider than scale was necessary to accommodate an adequate (pot) boiler, and at the other end of the spectrum it had become accepted that 15" Gauge locomotives should be. constructed to 1/3rd scale, not 1/4 scale as implied by the gauge.

The Table Top Railway pushed small clockwork and electric motors to their then limits. The point is made vividly by a drawing in Greenly's first 00-gauge article (Model Railway News [MRN], April 1925) which shows a 4-4-0 based on the ex-NER R1 class with its entire boiler occupied by a pair of Bing mechanisms. The reason for this arrangement is explained in the article: when testing the first Bing set with a load intended to equal 300 scale tons Greenly found the 6V motor overheating. Twin motors halved the current flow through each motor for the same power output. The penalty was a mechanism 9cm X 3cm x 2cm which would only just fit into the largest pre-grouping 0-6-0s and 4-4-0s. Only one British railway, the GWR, then had a large fleet of 4-6-0s and as Stewart-Reidpath noted in the very first MRN, GW engines had taper boilers "which at the smokebox end are really very small in diameter" and made it impossible to fit in a contemporary 00-gauge mechanisms.

Underlying all these problems were the facts that in 1925 less than half of British houses had mains electricity, and the only rectifiers then available were small versions of the big mercury-are rectifiers used by inter-war suburban electric railways. Consequently the normal method of supply for model electric trains was from "accumulators" : large high-current rechargeable 6V batteries which could be obtained and recharged at wireless shops. In the Thirties and Forties 6V car batteries were commonly used, supported (as new smaller low-current rectifiers became available) by a trickle charger. As with full-sized third-rail electrics, low voltage meant high current and interwar 00-gauge locos drew anything between 1 and 3 amps. These difficulties were compounded by the poor magnetic materials then in use, which necessitated massive magnets for small motors. Indeed in 1926 A. R. Walkley felt it necessary to demonstrate that one could actually fit a permanent magnet motor in an H0 tank engine, so great was the apparent problem. Building a sufficiently strong, small, controllable self-reversing clockwork mechanism seems to have proved just too difficult.

It is at this point that a new strand enters the story, in the shape of three modellers from the newly-formed Wimbledon MRC: A. Stewart-Reidpath, A. R. Walkley, and Michael Longridge. Sometime in 1923 they began experimenting with models roughly half the size of Gauge-0. In Stewart-Reidpath own words (MRN Jan 1925): "A scale of 1/8" proved to be just too small for efficient tractive power and 4mm or 3/16" scale revealed that the saving in space required (which is one of the main objects in the introduction of this new Gauge) would have been considerably less in proportion ... 3.5mm scale has proved to be the happy medium for this small gauge". There is no mention of Greenly or indeed of the exact track gauge, but there is mention - critical mention - of Greenly's wheel standards. "I cannot see the necessity of wheels having treads 5mm wide even for the German-made tin-plate sets ... It is not only unsightly, it is bordering on the ridiculous ... By using built-up permanent way, wheels having treads 1.5mm wide and flanges 1mm deep can be used with confidence, the only important point being that track must be well and truly laid" (the comparative BRMSB dimensions for EM were 1.5mm tread and 0.75mm flange!). Stewart Reidpath concludes with some a statements of principle: "Scale is a thing that matters and it is possible to work to it. Detail is worth the time and trouble it takes - lay your track carefully ... Always work to drawings, and see that they are good ones. And for the love of Mike, never say 'That's near enough'".

In short, we are dealing with the hobby's first fine-scale movement. Their relationship to the Table Top Railway is shown by an article by A. R. Walkley in the July 1925  MRN, which contains a photograph showing a train headed by what is obviously a Bing 2-4-0T, with the comment in the text, "the loco speaks for itself and will not have a permanent position on the line".

This movement quickly attracted support. Even the Rev. Edward Beal, who in later years preserved a firm silence on the 00/H0 controversy while building the most famous 4mm scale layout of the period, was bold enough to write in the Oct 1925  MRN "For practical purposes, however, in the writer's opinion, there is nothing to beat the 'half 0 gauge' scale". There now followed what  MRN dubbed 'The Battle of the Gauges', in which Greenly at first stood nearly alone for 4mm scale. In December 1926  MRN printed the results of a vote on the relative merits of 3.5mm and 4mm scales: "At the time of writing we have received 131 votes divided as follows: 3.5mm and 5/8" gauge, 106 votes; 4mm and 19mm gauge, [American 00) 13 votes; 4mm and 16.5mm gauge, 2 votes." In the light of subsequent history, it is an astonishing result.

It also shows that the traditional view of the adoption of 4mm scale 16.5mm 00-gauge in Britain - 'that it was All Greenly's Fault' - will not hold water. Greenly was a very authoritative figure, but one man's authority does not overturn a general consensus in the hobby in favour it of an apparently logical exact scale position. Something else is required. In fact, we have to explain two contrasting phenomena: the virtually complete abandonment of H0 in favour of 00 in Britain between 1926 and 1939, and the universal adoption in Continental Europe and the USA of a scale (3.5mm) invented by a group of South London modellers.

One reason for the demise of British H0 - the difficulty of fitting the only readily-available electric mechanism into a loco - has already been touched on. A second, and probably more important, reason was raised by Greenly in the June 1925  MRN. "In fixing a scale it is important to get sufficient space between the outside of the tyre and the outside of the vehicle. Don't worry about the exact scale equivalent of the gauge ... the width over the outside of the tyres should be correct". His accompanying Fig.3 shows he was now proposing 2.5mm treads for 4mm, a value which is comparable to  present day Hornby (2mm) and British Lima (2.9mm) wheelsets.

Greenly makes a second point: while it is possible to get away with a great deal in a model of an inside-cylinder engine, when Walschaerts valve gear has to be modelled, the restricted width available in 3.5mm between the wheel face and edge of a scale-width British footplate would make the task almost impossible. Indeed Sydney Pritchard of PECO, himself an H0 modeller between the wars, later maintained that certain types of British outside-cylinder locomotives simply could not be built in H0.

These were to prove formidable problems within the confines of British loading-gauge steam. They still trouble today's P4 and S7, where they have been resolved only by the adoption of large radii curves, small layouts, and fully compensated chassis. The accepted solution in US and Continental H0 is to let the wheels spread outwards to accommodate the wider treads; as steam engines in these countries were generally built with high running-plates and exposed wheels there are no splashers to foul the wheels, whilst the bigger loading gauge results in wider vehicles with plenty of room to fit valve gear and a generous overhang to veil over-wide bogies in a decent obscurity. 

The spread of H0 beyond Britain is difficult to trace. It is almost certain that the Americans imported the concepts from Britain in the period 1926-8. The name H0 and the existence of 'American 00' (4mm scale/19mm gauge) are clear evidence for this - as both originated in Britain in 1926-7;  MRN reported in September 1927, "A name for 3.5mm gauge ( sic) is coming into use in some corners of the model railway world. This is H0 gauge which means half '0' gauge, to distinguish it from 4mm scale, which is adopted in the trade for '00' gauge".

The first British mention of 16.5mm gauge (as opposed to 5/8" or 16mm comes in a letter from a French correspondent in July 1925. JEP had adopted 16.5mm in France, but the real reason this figure became established is believed to be the profile of the wheels of the Bing Table Top Railway. These had a huge root radius which was fine on round-topped tinplate rails, but when modellers started making hand built track , using 1/8" four-square or 14x22 gauge brass bar for the rails, it was necessary to spread the gauge slightly to ensure the flat part of the tread was running on the flat rail surface.

Although H0 may have appeared later in the US than in Britain, it was to spread more rapidly. In its first Reader Poll in 1936,  Model Railroader reported that 36% of respondents used H0, and another 2.1% American 00. Judging by the magazines, the comparable British figure would have been under 25%. Two-rail electrification was already common in the US in the late Thirties, a time when It was regarded as a controversial and potentially unworkable novelty in Britain.

By 1925 the new gauge had aroused considerable interest but only isolated experimental models had been built. The first significant layout in the smaller scales was A. R. Walkley's 'Layout in a Suitcase' described in the June 1926  MRN, this was the first ever portable layout (and the very first exhibition layout), fully scenic, depicting a small goods yard, built on hinged, folding boards with backscenes, and (almost unbelievably at this date) two-rail.

Over the next few years a number of specialist firms emerged to serve the new gauges. The first, Marshall-Stewart (later Stewart-Reidpath Ltd) were H0 specialists; their cast metal 0-6-0T body was almost the only widely available loco, and Edward Beal kit-bashed them into everything from an a 0-4-0T to an 0-8-4T. By the mid-Thirties Stewart-Reidpath were also advertising 00-gauge material, a first sign that British H0 was fading. Holtzapffel, who merged to become Walker's & Holtzapffel, and who as W&H Models remained London's largest model shop until their demise in 1994, began advertising 00-gauge components in 1928. Bonds of Euston Road were advertising a hand built 0-6-2T in 00 from April 1929, whilst another London firm, Hamblings, proclaimed themselves 'The Home of 00 Gauge' and stocked nothing else, and Edward Exley became famous for their hand-built coaches. Most of these firms combined a model shop with the production of a range of their own components and a small range of hand-built 00-gauge locomotives and rolling stock available to order. By 1939 it was possible to obtain a surprisingly wide range of good 00-gauge models form such sources - albeit at a price, often a considerable one.

The big name in British 00-gauge modelling in the Thirties was the Rev. Edward Beal (1889-1985), a Church of Scotland minister, whose freelance West Midland Railway was the first demonstration of the potential of 00-gauge on a grand scale. Begun in 1930 and rebuilt and enlarged three times by 1937, the West Midland featured a long series of firsts: a double-deck girder bridge, an operating hump yard, a working Beyer-Garrett, a coaling plant with working wagon hoist, and many more, all written up in  MRN with fine perspective drawings, along with a seemingly endless stream of designs for realistic operational and lineside features. He published a book,  West Midland, A Railway in Miniature, in 1952 giving the full story of this remarkable railway. Indeed during the 1930s Beal was the main and at times only source of the  MRN's 00-gauge articles, as well as the author of what became the standard books on the hobby,  Railway Modelling in Miniature (1st edn 1935) and  The Craft of Modelling Railways (1937).

Beal was a highly competent scenic modeller, and in the early thirties he and Sir Eric Hutchinson founded MERCO of Dundee who produced brick papers and wagon and coach litho papers in 4mm and 7mm scale. By the a standards of the day these were very realistic; they were also modestly priced and opened up many possibilities to modellers who could not afford hand-built models from the London shops.

Beal was a freelance modeller: all his models are essentially 'generic', and they lack the intense sense of a very specific time and place which has come to characterise British modelling; but they were always under-pinned by a very thorough understanding of the operating principles and working practices of the contemporary mainline railways. His vision of a really big layout depicting the operation of a modern mainline system has been taken up in America although not in Britain, but as Beal was both keenly aware of U.S. developments and a friend of Linn Westcott, editor of the Model Railroader, this is perhaps not inappropriate. Nevertheless Beal inspired the post-first world war generation of British modellers by showing them what could be done in 00, and he provided them with their main source of practical advice on how they could do it.

In 1927, Stephan Bing resigned from Bing-Werke after a disagreement with his Board. He subsequently purchased a small Nuremberg toy company, and along with several other former Bing-Werke staff began to develop new products under the trade name 'Trix'. By 1932 he had established a UK agent, Trix Ltd, with W. J. Bassett-Lowke as one of the directors, and this company set about manufacturing Trix products at Northampton.

As the Depression began to ease in Germany, Trix started to develop a small-scale toy railway system; the design team included Oswald Fischer who had worked on the Table Top Railway. The gauge adopted was 16.5mm; the new trains had a credible DR look to them, but were not scale models so it is difficult to specify the scale exactly. They operated on a 14V AC 3 rail system which allowed two trains to be controlled on the same track.

The new Trix-Express was launched in Germany in March 1935 and marketed in the UK for Christmas 1935 under the name 'Bassett-Lowke Twin Train Table Railway'. British outline stock appeared in November 1936; this was made at Northampton using chassis and parts from Germany and was again toy-like. Meanwhile Marklin had launched a competing table top range in Germany, also on 16.5mm track and using 14V AC but at a scale of 1:85, under the name '00 gauge'. These two German ranges, Trix and Marklin, mark the appearance of H0-gauge RTR in Continental Europe.

1937 saw the first 'scale' RTR trains from Trix, a two-car Diesel Flyer and a DR Pacific with Walschaerts valve gear; both had readily identifiable DR prototypes although they were never so advertised. The scale was approximately 1:90. Trix's French agent also received the first (non-scale) French outline model. The following year saw the Pacific chassis used to produce an LMS Princess, an LNER A3, and a US Pacific, followed in 1939 by the culmination of pre-war British Trix, an LMS Coronation. Appropriate lithographed coaches were produced for each engine.

The Trix Twin Railway had begun to affect sales of Meccano's Gauge-0 Hornby Trains. Meccano responded by developing their own 00-gauge range, Hornby Dublo "The general scale of the models will be 4mm to the foot, but the track will conform to the usual 16.5mm gauge. In general our '00-gauge' material will be as near to scale as possible" ( Meccano letter, 2/5/36). Hornby Dublo was launched in September 1938 in three-rail 12V dc and clockwork, with wagons for all four mainline railways, LNER coaches and scale models of an LNER A4 Pacific and N2 tank. All models were very close representations of well-known prototypes and comparable to or better than the best commercial Gauge-0 of the day. (Indeed the locos were to remain in production into the 1990s under the Wrenn banner). Hornby Dublo was just over a quarter of the price of Trix Twin, it was better engineered, and it was backed by Britain's biggest toy manufacturer; it was an immediate success and additional wagons, track and accessories quickly followed. Then came World War Two.

It is sometimes said that Hornby Dublo killed off British H0. In truth by the mid-Thirties British H0 was already at a very low ebb, surviving mainly in the letters columns of the two modelling magazines. In 1936, one letter to the  MRN pointed out that more than 80% of components available to 16.5mm gauge modellers were to 4mm scale. And in 1939, after another bout of H0 v 00 correspondence, the editor of the  Model Railway Constructor [MRC] proposed a National Register of H0 modellers (perhaps the first attempt to establish a 'scale society' in Britain). Fewer than a dozen modellers replied expressing interest.

Why did British H0 fail, and why did 'American 00' (19mm gauge) vanish without trace? The size of motors alone was not the issue, P. O. W. Chubb, later proprietor of the  Constructor and a member of the later BRMSB, wrote in the July 1936  MRC:  No.1) Can one build strictly to scale? No. 2) Can one reduce external lines and details to scale? Yes. if one is prepared to go the 'Whole Hog' and reduce all working condition to scale. 3) (That means) (a) scale curves: in Gauge-0 15' or over, (b) sprung axles...if standards of absolute accuracy throughout are laid down the very conditions themselves would kill the hobby". And in a later letter ( MRC Dec 1941) "The reason 3.5mm failed is that it was much too difficult for the man of average ability to build anything satisfactory that would work". George Mellor of GEM wrote in his 1938/9 catalogue: "It is impossible to employ exact scale wheel-treads and flanges...as these would be so small that the slightest error in aligning the track or the suspicion of a warp on the baseboard would be sufficient to cause derailment...Goods of our manufacture are ALL built to 4mm scale 16.5mm gauge, the original and only practical 'Double-0 Gauge'".

As we have seen, exact scale wheels are needed to fit inside the splashers of a scale British outline steam engine. All the Trix engines, including their British Pacific, had their wheels below and clear of the bodywork, meaning seriously under-diameter wheels on British prototypes. But even with exact-scale wheels there is no escape for either British H0 or P4 from F. W. Chubb's point three above. In a period when layouts were built at home and compensated chassis were still 30 years in the future, that spelt death to British H0.

One further pre-war development must be noted. In 1935 the National Model Railroad Association was formed in the USA, to bring order and compatibility to the chaos of proprietary standards then existing in the US. After another round of H0 v 00 correspondence in the  MRN the following spring the April 1936  MRC announced the formation of a 'Bureau of British Railway Modelling Standards', which it stated would establish standards for British modelling and which it claimed the trade were pledged to support. The next MRC saw an article headed 'Proceedings of the BRMSB' and signed 'The Editor' which reported that a pair of exact scale 16.5mm gauge wheels had been made and the flanges had shown no sign of breaking off despite extensive testing. The following month, under the same heading came a set of wheel standards for 16.5mm gauge including a 15mm back to back measurement.

With the exception of one article on an unrelated topic nothing more is heard of the BRMSB for five years. Nowhere are its members named, the  MRN for the period is totally silent on the subject of the 'Bureau', and in its 1950 pamphlet the BRMSB itself stated that it was founded in 1941! The  Model Railway Constructor at this period was a 'one man band', published, edited, largely written (under various pseudonyms), and even briefly printed and bound by its founder E. F. Carter. It seems very likely that this first BRMSB was another of Carter's pseudonyms, and represented an attempt to bounce the hobby into adopting an ultra fine-scale set of standards for 16.5mm gauge. That attempt failed, but it planted a seed which was to grow dramatically in the next decade.



Sources and Authorities for Part 1

Return to the Double O Gauge Association web site