Reed references

1. Birmingham & South Staffordshire, or Illustrations of the history, geology & operations of a mining district. W. M. Hawkes Smith. 1838 [not in Ottley]
2 Patent No 2599, 24 March 1802. Richard Trevithick & Andrew Vivian
3 Patent No 2632, 28 June 1802. Matthew Murray
4 Life of Richard Trevithick. Francis Trevithick. 1872
5 Patent No 3431, 10 April 1811. John Blenkinsop
6 Watson Collection; North of England Institute of Mining & Mechanical Engineers, Newcastle
7 Patent No 3632, 21 April 1812. Wm & E. W. Chapman
8 Patent No 3666. 13 March 1813. Wm Hedley
9 First in Practical Treatise on Rail Roads. Nicholas Wood 1825
10 In the Science Museum [Timothy Hackworth's notebook]
11 Observations on the Comparative Merits of Locomotive and Fixed Engines. Robert Stephenson & Joseph Locke. 1830
12 Patent No 3887, 28 February 1815. Ralph Dodds & George Stephenson
13 Patent No 4067, 13 September 1816. Wm Losh & George Stephenson
14 In the Science Museum: presumably J. Adamson. Sketches of our information as too rail-roads. Newcastle: 1828 (Ottley 264)
15 Early Wooden Railways. M.J.T. Lewis. 1970.
16 Locomotive Profile No 25
17 Trans. of Newcomen Society, Vol VII, February 1927
18 Archiv fur Bergbau und Huttenuiesen, Vol XIX, 1829; translated and reproduced in part in A Century of Locomotive Building 1823-1923. ]. G. H. Warren, and in Trans. Newcomen Society 1953.
19 Notebook of John Urpeth Rastrick. 1829
20 LIfe of Robert Stephenson. Vol I. J. C. Jeaffreson
21 Locomotive Profile No 7
22 An Account of the Liverpool & Manchester Railway. Henry Booth. 1830
23 Practical Treatise on Locomotive Engines. G. de Pambour. 1836
24 Technics and Civilisation. Lewis Mumford
25 Description of Patent Locomotive Steam Engine. W. P. Marshall. 1850. (Written 1838)
26 Locomotive Profile No 15
27 Patent No 7745, 26 July 1838. John Gray
28 'Diaries of David Joy'. The Railway Magazine June 1908
29 The Railway Magazine August 1899, p. 104
30 A Century of Locomotive Building 1823-1923. J.G. H. Warren. 1923
31 Patent No 8998, 23 June 1841. Robert Stephenson
32 Railway Machinery. D. K. Clark. 1855
33 Patent No 9261, 15 February 1842. Thos R. Crampton
34 Patent No 14107, 25 October 1884. David Joy
35 Act 7 Geo IV cap 49
36 Society of Engineers 1862
37 The Engineer, 1 October 1858, p. 255
38 Glasgow Institution of Engineers, 1862
39 Inst. of Mechanical Engineers, January 1861 40 Proc. Inst. Civil Engineers, Vol XVI, 1856
41 'Coal Without Smoke', paper to Soc. of Engineers 1862
42 A Study of the Locomotive Boiler
43 Proc. Inst. Civil Engineers, Vol XVI, 1856
44 Proc. Inst. Mech. Engineers 1866
45 Patent No 6484, 7 October 1833. Robert Stephenson
46 Locomotive Profile No 1
47 To patent No 11086 of 1846; George Stephenson & Wm Howe
48 Report to canal shareholders Present & Future Prospects of the Monmouthshire Canal. Jas Brown. 1847
49 From 'Railway Accidents'; paper by Mark Huish. Proc. Inst. Civil Engineers, Vol 11, 1851-2
50 Locomotive Profile No 15
51 Locomotive Profile No 27
52 The Stirling Singles. K. H. Leech & M. G. Boddy
53 The Engineer, 5 September 1862, p. 142
54 Engineering, 11 May 1894, p. 611
55 Locomotive Profile No 31
56 The Decapod Locomotive of the GER. W. O. Skeat. Trans. Newcomen Society, Vol XXVIII, 1952-3
57 Proc. Inst. Mech. Engineers, 14 December 1945
58 Locomotive Profile No 22
59 'A Modern Locomotive History'. E. S. Cox. Journal Inst. Loco Engineers, 1946

Some Fundamentals
Apart from the first four of all, steam locomotives from the beginning in 1803 to the end of new construction for Britain in 1960 were pre-eminently creatures of their environment. Constant reference back to this cliche explains many restrictions, many novelties, and the actual trend of development and construction at different periods, which were governed more by manufacturing and mechanical engineering possibilities, and by the vagaries and uncertainties of financial-economic practices, than by true railway requirements.
Throughout the whole period locomotive design, construction and operation in Britain were more of an art than a science. Some consummate artists came forth; many were scarcely pavement artists. At the 'beginning a different situation was hardly possible; yet even in the 20th century scientific thought developed from fundamentals was scantily applied, and personal preferences based on incomplete thinking and pure emotion were prevalent, so that in the end the results fell short of what might reasonably have been expected from 150 years of development and the immense range of resources and facilities available to a nationalised railway system from a whole nation.
Only development of early primitive steam locomotives permitted the transmutation of the old waggonways, tramways and rail-roads into a country-wide railway system. Having brought something greater than itself, the steam locomotive, or rather the locomotive engineers, continued to control development and operation much as in later years holders of £10,000 worth of founders' shares controlled £10 million commercial enterprises. Often they restricted changes to more suitable methods in the fashion of. a debenture-holders' committee.
Little basic adaptation of steam power to suit widening railway requirements was made, and gradually railway systems in fully-developed countries became unbalanced, financially and technically.
Financial unbalance came from the general monetary system that did not, and does not, permit the automatic elimination or costless writing off of capital when the physical items it represents are no longer in existence. Within the capital structure of British Railways on its formation in 1948 was the legacy of Stockton & Darlington Railway capital of 1820-30, when nothing was left to represent it other than Locomotion on its plinth at Darlington station and a few small exhibits in York railway museum.
Two locomotive causes substantially helped the technical unbalance of railways in Britain and other highly developed countries, though they became of major importance only during this century, and then to increasing tempo. First, the continued construction of conventional steam locomotives of everincreasing size and weight perpetuated low power: weight ratios against total moving weight when, conversely, ever higher speeds and increasing density of traffic were calling for just the opposite. Such matters were not of such moment in George Stephenson's day; they became decisive during the 20th Century.
With conventional steam locomotives high power: weight ratios could be provided only with inconvenience. For example, 160 tons oflocomotive and tender weight and an axle load of 22 tons were required on the LNER to maintain a 70mph schedule over a level route with the 240-ton Silver Jubilee. Low power weight ratios promoted potentialities for unpunctual working; but for many years only in Britain was this supplemented by that far greater promoter of unpunctuality — the unbraked freight train.
Secondly, continued development of reciprocating steam locomotives forming no more than five to eight per cent of the wheeled stock of a railway gradually dictated the use of rails twice as heavy and bridges twice as strong as those needed by the other 92-95 per cent of the wheeled stock. This could not be called efficient or economic; but no attempts were made to develop high tractive effort, acceptable speed and great horsepower on axle loads low enough for substantial savings to be made in track cost and bridge construction. By the time the Garratt locomotive was well developed only railways 'up country', or those with exceptional drawgear, could derive proper benefit from it. Apart from a few examples in North America from World War I years, not until the 1950s was deliberate application made of wagons with laden axle loads equal to those of the locomotives that handled them.
This unbalance was aggravated in Britain by the perpetuation of three-link loose couplings and unbraked freight trains.
The dominant position, and even more the restricted capacities, of the locomotive engineer were accentuated also by the 'vested interests' that grew up in the British railway world and prevented free interchange and common cause, and admitted no new ideas from outside, taking only those that arose inside the particular closed ring. This was one major reason for what may not unjustly be called the great brake scandal. These factors remained almost unabated to the end of steam. The operation and operating ratios of British Railways since 1 January 1948 have been shaped greatly by the locomotive engineer and his limitations, which have prevented successful adaptation to rapidly changing conditions; and the huge 'paper' losses have been shaped by continuance of financial pundits equally unable to adjust themselves to the times.
One might well ask how a seemingly crude and ineffective machine could gain, and retain long after its day, the enormous interest, and even affection, given to the steam reciprocating locomotive. Possibly the reason is because of all man's engineering productions the steam locomotive is likest unto man — and woman. Within the large general framework of environment it showed the same immense range of capacity and characteristics; it was almost unpredictable in its performance except in large generalised terms; its day to day performance was affected profoundly by its physical condition; it had every shape, size, formation, colour defect and protruberance (warts and all), and almost every human characteristic from gentleness and urbanity to viciousness and irresponsibility; it proceeded with gaits that varied from the smooth and dignified to the rolling and unsteady; it could often limp or stagger home unassisted after misfortunes or spewout of potations that had been too deep; and, as one of the kindliest presidents of the Institution of Mechanical Engineers once remarked, it was human in being easy and delightful to conceive but painful and difficult to deliver.
Governing condition for all steam locomotives was the ferrous wheel on the ferrous rail. The only change in these frictional conditions from the time of Trevithick was the machining of the wheel tread and the smoother rail head. The effect of this was marginal, as was the effect of multi-cylinder propulsion. The only attempt at modification on any serious scale was the rack railway, for the application of rubber tyres did not get beyond railcars. Limitations of the steel wheel on the steel rail acted equally in traction and retardation, and as the railway system matured, braking became at least as important as tractive effort and horsepower. To this day the need for friction to give high accelerative and decelerative performances has to be carefully balanced against the need to reduce friction to give low resistance to movement and reduce the needed power output.
Effectiveness of the steam reciprocating locomotive arose primarily from the direct drive between piston and wheel tread, by which the starting of the prime mover meant automatically the starting of the locomotive and its train. In this it differed from the internal-combustion engine, and much effort in the early days of diesel traction was devoted to attempts to reproduce the direct-drive effect of the steam locomotive.
Moreover, the commercial effectiveness of the steam reciprocating locomotive depended on less than half a dozen essentials. They were:
(1) two or more cylinders with cranks at different angles to ensure starting in any position and give reasonably constant torque;
(2) the multitubular fire-tube boiler to ensure adequate steam generation within permissible weight and size limits;
(3) the blast pipe, which was necessary to get the potential generating capacity out of the multitubular boiler, and to give automatic regulation of the steam-generating and steam-utili sing portions of the whole machine; and
(4) expansion valve motion to give fuel economy and permit high speed.
Compounding was no more than an extension of the expansion principles, though its practical advantages were in other ways. Within the present century came the addition of something fundamental in a thermodynamic sense — the superheater. It was essential only in permitting further enhancement of power and giving additional fuel economy.
The probable adequacy of the relatively smooth iron wheel on the relatively smooth iron rail for light loads was shown by Trevithick's locomotives, and three of these also included blast-pipe exhaust. In 1812 came the first application of two cylinders at right angles; but not until 1829 were more than two of the four essentials embodied in anyone locomotive, and as a result the steam locomotive then made a sudden bound forward. Only in 1829-31 did the steam locomotive take on the basic form it maintained for the next 130 years, and justified the term, often used up to the end, of the 'Stephenson' locomotive. That term denoted Robert at least as much as George.

2 From Trevithick to Stephenson

Practical measures for locomotives came first with Richard Trevithick's high-pressure stationary engines evolved 1799-1801, though before that some possibilities of travelling engines on Watt's low- pressure system seem to have been mooted1; and road locomotives on other systems had been tried experimentally by Cugnot (1769) and Murdoch (1784).

Conviviality ended Trevithick's own first road locomotive attempt in 1801, but the results were good enough for money to be found to get Trevithick from Cornwall to London and pay fees for a patent2 to cover high-pressure, or 'strong steam', engines as such and the application thereof to road and rail vehicles. The patent specification stated that in general the ordinary surface of the wheel would be sufficient for adhesion. No mention was made of blast pipe or any other kind of exhaust; but a claim was made for two cylinders with cranks at right angles.

Matthew Murray usually has been given the credit of initiating cranks at right angles in his patent3 perhaps because all Trevithick's locomotives had single cylinders, whereas Murray applied twin cylinders and cranks at 90° to the first locomotives he built, though that was not until 1812. The Trevithick patent also covered mechanically-driven pump feed to the boiler, feed water heating, and boiler lagging by exhaust steam.

Trevithick was responsible for four locomotives: (1) the Coalbrookdale locomotive 1803: (2) the Penydarren engine 1804; (3) the Gateshead engine 1805; and (4) the London locomotive 1808.

Construction of (1) was undertaken by the customer, the Coalbrookdale Ironworks in Shropshire, a firm with long experience in tramway and waggon way materials and general ironwork, and noted for its cast iron cylinders for steam engines. For well over a century the actual completion of this locomotive was scarcely appreciated, and the drawing found later was thought to be that of Trevithick's next locomotive. Only on discovery that the Coalbrookdale way had been narrowed to a gauge of about 3ft and converted from an edge way to a tramway prior to 1800 came the realisation that this drawing represented the first locomotive of all.

Engine (2) was the most famous of the four, largely because it won a bet of 500 guineas for Samual Homfray, the ironmaster owning the Penydarren works near Dowlais, South Wales, where the locomotive was built under the supervision of Trevithick. It is the only one of the four of which no pictorial representation exists; all that is known about it is contained in letters of Trevithick and Homfray4. Between February and July 1804 it ran some trips over the 9½ -mile tramway from the works to the Glamorganshire canal, but broke too many of the thin angled tram plates, and was dis- mantled and parts used for works purposes.

No 3 was built at the Gateshead works of John Whinfield, the Trevithick agent for high-pressure engines in the north-east, to the requirements of Christopher Blackett, proprietor of the Wylam waggonway. Though completed and put through a few runs in the works yard it was not accepted, possibly because on completion Blackett realised its weight of 4½ to 5 tons on four wheels with a tread width of 1in was impracticable for the wooden edge rails with which the Wylam line was laid. It was the first locomotive to be built with flanged wheels.

All three had a single horizontal cylinder inserted in the boiler end, a wide transverse cross head encircling two cylindrical guide bars, gear drive, and a big flywheel to get the engine over dead centres on stationary duties. As in waggon practice of the time, the wheels were loose on the axles; and as the gears were on one side only, though there was a crank at each side, the drive was only on to the two left-hand wheels. Thus the adhesion weight was only half the locomotive weight, and there was a strong shouldering couple that forced the driving wheels against the sides of the rail.

The Coalbrookdale boiler had two internal flues of different diameters, connected externally at the end remote from the chimney; succeeding Trevithick engines, to gauges above 4ft, had rather larger diameter boilers in which a wholly internal return flue could be incorporated. Machinery arrangements of No 1 must have been dictated largely by the narrow gauge. In No 3, and possibly in No 2, the whole layout was reversed, the cylinder being at the end opposite to the chimney and firehole, and the crankshaft located between chimney and boiler end. Layout of the Penydarren engine is not known, but as the gauge was 4ft 2in between backs of tram plates it may well have been like that of the Gateshead engine.

No information exists as to how the exhaust steam was led to atmosphere in No 1, but Trevithick's letters show that the exhaust of No 2 was up the chimney and that he was beginning to appreciate the value of this. How the engine ran from Penydarren to the canal wharf is unknown, for there was a tunnel en route that was lower than the height of the chimney.

No safety valve seems to have been fitted to any Trevithick locomotive, but No 2 is recorded by Trevithick to have had a lead rivet in the flue top and this acted as a fusible plug - the first one known. Two, if not three, of the locomotives were multi-purpose machines to do both stationary and traction work and had a form of clutch to disengage the drive from the rail wheels when not on the tramway.

In all three locomotives the cast iron boiler acted as the frame, and to it were attached cylinder, crankshaft bearings, the bearings for intermediate gear wheels, axle supports, and whatever drawgear there was. Despite the provisions of the 1802 patent not one of these locomotives had pump or any other feed to the boiler, and once the contents were consumed the boiler had to be refilled by hand. Trevithick's fourth locomotive differed from the others in having a single vertical cylinder inserted in the boiler top at the back end, and it had direct drive down to crankpins in the rear wheels. It was the first direct-drive locomotive, the first single driver, the first 2-2-0, and the first passenger engine, and an express passenger engine at that, for speeds up to 20mph were contemplated, though probably not reached. It ran on a circular track within an enclosure on the site of the present University of London buildings at various times between July and September 1808; and in view of the speed and proposals for several hours of continuous running it was the first locomotive to have a footplate.

This engine was built complete to Trevithick's requirements by Hazeldine & Co of Bridgnorth, then under the technical leadership of John Urpeth Rastrick. Known as Catch me who can, it ran on tram rails laid on continuous baulks of wood. No dimensions other than its weight are on record, and nothing is known as to its fate. As with No 2 it was connected with a bet, Trevithick offering to back it to make a greater mileage in 24 hr than any horse in England.

Trevithick's locomotives were before their time; there was no pressing economic or engineering need for any of them. Not one did useful or concentrated work yet they were not without influence on later practice. All four operated with the adhesion from normal treads on normal rails or tramplates; but with only half the locomotive weight adhesive they did not prove the adequacy of the smooth wheel on the smooth rail for economic haulage, and the doubts that arose led to the peculiar form of the first post-Trevithick type, which was intended to do a job of work and to counter the increasing cost of horse traction (up to £80 a year 'working' expenses per horse) due to the long-continued Napoleonic wars. This first commercial application of steam traction was the Blenkinsop type working from Middleton colliery to Leeds, about 2½ miles. From the beginning loads above 10 times the locomotive weight were hauled along the level. Consideration of loads such as these led Blenkinsop to make certain of daily operation by using a rack.

Blenkinsop's patent5 was for a railway system, not a locomotive, but he did 'declare that a steam engine is greatly to be preferred to any other first mover'. More than any other man of his time who was in a position to have locomotives built, Blenkin- sop realised the possibility of immediate applications to lines other than his own. Over the next few years, while retaining his position as Charles Brandling's manager and partner at Middleton, he did intensive sales work in the north of England, and by 1816 more than half of all steam locomotives at work were of his type.

Design and construction of the locomotives were left to Matthew Murray of the firm of Fenton, Murray & Wood at Leeds (price £350 to £400 each), and the Middleton way was relaid with cast iron edge rails, one of which had semi-circular cogs on the side which meshed with the final gear wheel on the locomotive. The four road wheels were flang- ed. Blenkinsop paid the Trevithick patent-holders £30 a locomotive licence fee for the use of the high- pressure principles, and this would be the reason for the use of Trevithick's four-way plug-type valves for the cylinders in preference to Murray's own short flat-D outside-admission slide valves.

The distinct advance of the Blenkinsop-Murray locomotives was the use of two cylinders with cranks at right angles. The exhaust did not go up the chimney and so did not influence the fire, though the boiler was only a single-flue type and so had not the potential evaporative capacity of the Trevithick return-flue form. Cast iron was still the boiler material, but the shell was oval, possibly to accommodate both the vertical cylinders plus a 14in flue. No means of boiler feed were fitted at first, so again the work done was limited to one boiler filling, which lasted about five miles. Blenkinsop had a scheme for filling against pressure from a high water tower, but this was not put into effect. Recorded pressure was 50-55psi. A defect was the rack on one side only, which gave a twisting effect against the rails, but it was retained on the grounds of simplicity and expense though the patent covered a rack on each side. A central rack would have prevented horse traction, which was used on the Middleton way and other lines where racks were put down whenever the locomotives had to be laid off.

The first locomotive was ready and tried in June 1812; it and a second one were put to regular work in August that year. The two vertical cylinders were let into the boiler, with the valve gear and an open exhaust outlet between them, and they drove separate intermediate shafts through the connecting rods. Gear wheels on these shafts engaged with a central gear below the boiler, and on the left-hand side of that shaft was the large toothed wheel that meshed with the rack. An advance on Trevithick practice was a separate wooden frame that carried the boiler through four fixed brackets on the upper side and the four axle bearings on the under side.

Two more locomotives of the same type were built for Middleton in 1813 and worked another section. From 1814 eccentric drives to the valves and wrought iron boilers with return flues were adopted gradually, and by 1815 all Leeds-built engines had been given a silencer box between cylinders and final exhaust, a spring-loaded safety valve, and a mechanically-driven boiler feed pump that lifted water from a frame-mounted tank at the back end. Engine weight at first was about 5 tons but the Middleton engines themselves in after years scaled 6¼ tons. From earliest days trailing loads of 70 to 85 tons could be drawn along the level, and later records show up to 140 tons, or 24 times the locomotive weight. Such loads justified the cost of the rack for they were far above the possibilities of plain adhesion wheels on the light tracks at that time.

By 1816 at least six Blenkinsop rack locomotives were or had been at work at Kenton & Coxlodge colliery (Newcastle), Willington (Tyneside), Orrell (Wigan) and Whitehaven, and two had been built in Germany. Those at Wigan and Whitehaven were built locally; those in the north-east were made by Murray, and the two for Newcastle in 1813-14 show the first known attempt at standardisation of parts, for on 14 October 1814 Murray wrote to Watson, who was handling the business: 'In your new engines we will endeavour to make the four cylinders one exact size, you can then have a pair of spare pistons to fit either engine.6

All engines except the German pair did useful work over some years. Pronounced drop in the price of horse fodder, and the remaining locomotive being almost worn out, led to steam traction being given up entirely at Middleton in 1835 after a period of growing horse traction when, according to a table published in 1837, the average trailing load of a locomotive was only 33 tons. This change is not surprising, for according to the investigations of Rastrick and Walker for the Liverpool & Manchester Railway, the coal consumption was enormous at 2¾lb/ton-mile.

The commercial success of the Middleton locomotives involved additional capital investment of around £700 a mile for the rack rails, which might be warranted for a considerable daily ton- nage, but for other routes two questions were still unanswered: (1) what was the effective adhesion between a smooth iron rail and smooth iron wheel?, and (2) was it possible to devise a machine to take advantage of that adhesion usefully and which would suit the light edge rails and tramplates then prevalent? The two men who set out in 1812 to solve these questions were William Chapman and William Hedley, who, though working separately, must have had some interchange of ideas.

Chapman began in 1813 with a locomotive built for him probably by the Butterley Iron Co, which pulled itself along a chain or rope laid between the wooden rails of the Heaton waggonway at Newcastle. As far as is known it had four wheeels, and two vertical cylinders sunk within the boiler which drove the chain rollers by side lever mechanism and a two-speed gear. A more important part of Chapman's 1812 patent7 was that covering the equalisation of the load on the road wheels of springless six- and eight-wheel locomotives and to enable them to pass round curves, for Chapman was convinced that such multi-axle types were needed to get adequate traction capacity on the existing light rails. This led to the first bogie locomotive, an eight-wheeler arranged in two trucks which gave longitudinal but not transverse equalisation, and which was built for Chapman by Phineas Crowther of Newcastle.

This locomotive was set to work on the Lambton waggonway in Co Durham in December 1814 and a contemporary description includes the first record of locomotive driving wheels slipping under high trac- tive effort. It was the first locomotive with two cranks at right angles on the same shaft, and as such was a further small step towards what became even- tually the conventional direct drive. Chapman's first locomotive did not last long in chain form, but seems to have been rebuilt as a six-wheel adhesion locomotive with two of the axles arranged in a bogie, and in that form worked for some years into the 1820s at Heaton, but not with a long haul at first as much of the Heaton waggonway continued with wooden rails until 1821

. In 1808 the Wylam line was relaid with cast iron tramplates to a gauge equivalent to 5ft to replace the wooden rails that had precluded the trial of the Trevithick-Whinfield locomotive in 1805, and in 1812 Blackett sanctioned the construction of a locomotive. Years later William Hedley claimed that as viewer, or manager, at Wylam he began ex- periments in October 1812 with a hand-operated vehicle which had satisfied him as to the efficacy of normal adhesion and which settled that question generally. Nevertheless, his patent of 18138 deals almost entirely with means of increasing friction such as teeth and flanges, and apart from that it really covers nothing at all.

Wylam Puffing Billy of 1815 as it was around 1864; the two sons of William Hedley alongside to left.

The first locomotive to run at Wylam was a single-cylinder Trevithick type built in 1813 by Thomas Waters of Gateshead, who by then was the north-eastern agent for Trevithick-type engines generally. It had gear drive, flywheel and a single-flue boiler, and was the first Trevithick-type locomotive to do useful work, for it was in operation, at least intermittently, for a year or so. Presumably it had exhaust blast up the chimney; if so, the Wylam people did not see the advantage.

During its operation a two-cylinder four-wheel locomotive later to be known as Puffing Billy was put in hand at Wylam. This locomotive was completed around March 1814, and soon after was supplemented by a generally similar unit now known as Wylam Dilly. A third was built subsequently. Possibly Puffing Billy was the first locomotive to have a wrought iron boiler and the first to have cylinders outside the boiler. Corresponding with the facilities at the Wylam shops the cylinders were of plate, and though at first they were lagged with wood they seem to have been un- lagged in later years. At first the exhaust went straight from cylinder to atmosphere; after a short time a silencer or exhaust box was inserted and the outflow from that taken to the chimney.

From the beginning these 5-ton locomotives broke too many tramplates, and during 1815 they were rebuilt as eight-wheelers. The illustration often reproduced since 18259 has purported to show the Wylam locomotives in their eight-wheel form, but is almost certainly Chapman's drawing of his own eight-wheel double-bogie Lambton engine. The Wylam locomotives must have been rebuilt on Chapman's principles, though neither Hedley nor his family in later years gave Chapman any credit.

In 1828 the Wylam way was again relaid, and in the process was reconverted from a tramway to a railway with edge rails of cast iron. The eight-wheelers were then converted back to four-wheelers, but other substantial modifications were made coincidently, additional to the provision of flanged wheels. In this revised form two locomotives worked intermittently until about 1862, when they were held for preservation; one is now in the Science Museum at South Kensington and the other in the Royal Scottish Museum in Edinburgh.

The one thing definitely known about performance at Wylam is contained in Timothy Hackworth's notebook10 under date August 1828. From this it appears that one eight-wheel locomotive could handle about 10,000 chaldrons a year in 820 journeys at a total cost of £372, and that all the gear wheels and half the road wheels normally needed renewal within the year.

Killingworth locomotive rebuilt with new wheels and other parts, as it stood for many years on the High Level bridge, Newcastle.

George Stephenson first appeared on the locomotive scene at the time Chapman's and Hedley's two-cylinder locomotives were coming on to the rails. He began to build his first locomotive Blucher in the autumn of 1813 at Killingworth and it was in steam at the end of July 1814, Therefore he could have had little of that help from visits to Wylam always claimed by protagonists of Hedley and Hackworth, for in 1813 all Wylam had running was the single-cylinder Trevithick-type locomotive, and Puffing Billy was on the rails only some four months before Blucher. His useful pre-knowledge must have been gained from inspections of the Blenkinsop rack locomotives on the Kenton-Coxlodge and Willington waggonways, and perhaps because of this his initial locomotive does not seem to have had blast-pipe exhaust when built. Writing in 1830 Robert Stephenson and Joseph Locke recorded11 'This [augmenting of fire temperature] was affected shortly after the first Locomotive Engine was tried on the Killingworth Railway by conveying the steam to the chimney where it escaped in a perpendicular direction up the centre.'

Blucher was the first flanged wheel adhesion locomotive to do any work; otherwise it was not of very notable design, and had Murray's type of drive to two gear shafts with cranks at right angles, from which further gears led to two axles in place of Blenkinsop's rack drive, The boiler was of single-flue type. Apparently there was a water chamber round the chimney that acted as a feedwater heater, and from which a mechanically-driven feed pump drew the water for the boiler. According to Nicholas Wood it also had a chain drive from one axle to the axle of the tender or convoy carriage to gain extra adhesion weight. This engine pulled six times its own weight up 1 in 450, but was recorded to be very noisy and rough, as would be the Blenkinsop, Chapman and Hedley locomotives.

How slow, cautious and uncertain were the steps from Trevithick's first locomotive to the later conventional steam locomotive drive is shown by Stephenson's second locomotive, completed in March 1815, which incorporated improvements outlined in his joint patent of that year12 Though the patent included the driven tender the mechanism was not applied, as experience with Blucher had shown it was not worthwhile, The distinct mechanical advance was the deliberate elimination of gear drive and rack rails and the substitution of two vertical cylinders driving down direct to crankpins in the wheels; the pins in one wheel-and-axle group were at right angles to those in the other group. Thus the unpatented direct drive of Trevithick's Catch me who can was resuscitated, but this time in a coupled locomotive with cranks at right angles.

To couple the two axles the patent named alternatives of inside sprockets and chains or inside cranks and coupling rods. Which was used in this second locomotive is uncertain, though chains seem probable and were used in later Killingworth locomotives. A separate frame supported the boiler and took the axle bearings and drawgear, and along this frame the leading axle bearings could be adjusted slightly to take up chain stretch. This meant the cylinder line was then not on the bearing and axle line. An increase in flue diameter raised the evaporative capacity; as a result, according to Wood, the blast exhaust was not always considered essential, and at one time some of the later Killingworth engines seem to have worked temporarily without it. Nor did all Killingworth locomotives retain pump feed, and as late as January 1825 a demonstration given to committees of the Liverpool & Manchester and Birmingham-Liverpool Railways had to be limited to 14 miles when the boiler had to be refilled with 200gal of fresh warm water, probably by a hand pump.

All questions of adhesion and of the evaporative capacity of single-flue boilers down to after the time of the first public railway to be worked by steam were at their simplest, because every railway or tramway to which locomotives were applied had a down grade in favour of the load. Thus steam trac- tion was established under conditions that were easiest for it.

Another Stephenson joint patent of 181613 included strengthened cast iron wheels, the applica- tion of separate tyres shrunk on, and the well-known steam springs. These were intended primarily as weight equalisers on four- and six- wheel locomotives, and were Stephenson's version of the Chapman ideas. Initial application was to the centre axle only of the first six-wheeler, the chain-coupled 0-6-0 of the Kilmarnock & Troon Railway in 1816. The longitudinal adjustment could not be repeated where steam springs were used on more than one axle, as they were on some of the subse- quent Killingworth four-wheelers, but with the coming of acceptable plate springs from around 1822 they were gradually given up for new construction. Stephenson built only four or five locomotives for Killingworth over the years 1814-21 with successive detail improvements. Another five to the same general design were supplied to Hetton colliery in 1822. No authentic record of the builder of these five remains; probably parts were supplied by Killingworth and by Burrell in Newcastle, and erection done on site. Hetton was a new pit opened in 1822, and the two nearly level sections of its railway route down to the River Wear formed the first railway to be built specifically for operation by steam locomotives.

The Stephenson engines just detailed established the 'Killingworth type', and more appear to have been built at that place by Nicholas Wood. They had the characteristics of two vertical cylinders let into the boiler, direct drive to crankpins in the wheels, coupling of the two axles by internal chains, and single-flue boilers. In general the cylinders were about 9in bore, of cast iron lined with copper sheet, and with valves operated at first by a tumbler, though later Wood introduced loose eccentric actuation. Some time after 1825 outside coupling rods were adopted for new Killingworth engines and are believed to have been applied to some of the older engines in place of the chains, though chains were still in use at various places in 1827. According to J. Adamson's Book14 the performance of a Killingworth locomotive was '126,000 tons conveyed one mile in 312 days. The performance at the Hetton colliery during the same period amounted to 198,000 tons conveyed one mile. The difference arises from the greater regularity of the line in the latter case.' Adamson also records the first suggestion for banking power: 'The use of two engines on such slopes, one acting in front of a train of waggons and the other behind them, has been proposed by Mr Stephenson of Newcastle upon Tyne, and where the inclinations are of considerable length, would form a most convenient method of surmounting them.

George Dodds on the Monkland & Kirkintilloch Railway in 1831 resuscitated the Killingworth type with outside coupling rods and other improvements including one not initiated at Killingworth: a flue furnace with a few small tubes from the front tube plate to the chimney base; there was no separate smokebox. The two Monkland locomotives were the first built in Scotland other than Timothy Burstall's Perseverance entered at Rainhill in 1829, and were constructed by Murdoch & Aitken in Glasgow. George Stephenson's greatest claim in regard to locomotives as distinct from railways is that from 1815 to 1825 no steam locomotives other than by him were built in Britain. Moreover, he continued those 10 years in growing faith in the steam locomotive as the motive power of the future and urged it whenever he could, as at Darlington from 1821, through a time when everyone else had given it up for new construction. At the same time he made but trifling improvements in the locomotive itself and permitted what was even a deterioration in design just when an improvement was needed most decisively.

He was not the father of the locomotive; he con- tributed few salient developments, and was never responsible himself for any really noteworthy design, though he held back more than one promising development through caution and obstinacy. From his appointment as engineer of the Stockton & Darlington he began gradually to visualise a country-wide system of railways to one rail gauge with steam locomotives as the motive power and, apart from very steep grades, rejected the central power plant as represented by cable haulage and self-acting inclines.

When Stephenson went to Killingworth the waggonway had been in operation some 40 years to a gauge of 4ft 8in. It was not altered, but when Stephenson grew beyond Killingworth and engineered the Hetton and Springwell ways, the SDR, the Bolton & Leigh and the Liverpool & Manchester, he merely kept the gauge to which he was accustomed. As M.J.T. Lewis put it succinctly15 : 'If Stephenson had found his mission in life at Heaton, our standard gauge would probably be 4ft 3in; if at Wylam, about 5ft. It was largely a question of chance.'