In 1883, aged 76 and after 29 years in the post of Locomotive Superintendent of the North Eastern Railway (NER), Edward Fletcher retired. Mr Fletcher was very well-liked by the enginemen who referred to him as ‘Father Fletcher.’ His successor was Alexander McDonnell, formerly of the Great Southern & Western Railway. In 1884, Mr McDonnell introduced the inside cylinder 4-4-0 type to the NER1 with his ‘38’ class locomotives. Unfortunately, these were especially unpopular with the enginemen, who did not take kindly to some of the innovative features included in the design, particularly the change from the traditional NER right-hand drive to left-hand drive. The enginemen claimed the new locomotives were no better than, and sometimes inferior to, those designed by Mr Fletcher. Matters became fraught and Mr McDonnell left the NER in September 1884.
Pending the appointment of a new locomotive superintendent, the NER General Manager, Henry Tennant, chaired a locomotive committee which was responsible for the design of a new 2-4-0 express passenger engine, the ‘1463’ class, usually known as ‘Tennants’. Much of the detailed design work was delegated to Wilson Worsdell, the company’s Assistant Locomotive Superintendent since 1883. Being of typical Fletcher style, the enginemen loved them. Fortunately, No. 1463 has been preserved.
In 1885, Wilson Worsdell’s elder brother, TW Worsdell, formerly of the Great Eastern Railway, was appointed Locomotive Superintendent and remained in office until retiring in 1890. TW Worsdell was an enthusiast for compounding on the von Borries two-cylinder system and introduced 2-4-0 (1886), 4-4-0 (1887) and two separate 4-2-2 classes (1888-1890). The re-appearance of 'Singles' was not a complete surprise as several British railways reverted to the type for a few years in this period. The adoption of steam sanding had made them more sure-footed and it was believed they were freer running than coupled locomotives. Patrick Stirling of the Glasgow & South Western and, later, Great Northern railways described a coupled engine as like, "A laddie running wae his breeks doon!"
Wilson Worsdell was appointed Locomotive Superintendent in 1890. It was clear that the NER’s policy with regard to express passenger locomotives was not fixed and, although some more compounds entered service, Wilson Worsdell favoured simple expansion. By 1893, the Board’s Locomotive Committee was also having doubts about the value of compounding and instructed that tests were to be arranged and a report provided to the committee. This work was done thoroughly by Ramsay Kendal and Vincent Raven and their report was endorsed by Wilson Worsdell. In April 1894 the Locomotive Committee decided, ‘Not to resume the building of engines on the compound system.’ With three exceptions, this was to be the policy for the remainder of the company’s existence.
The decision was not a problem for Wilson Worsdell as he had rebuilt some of his brother’s compound engines as simple expansion – the best known being the rebuilt ‘Singles.’ For new express passenger locomotives, he favoured the simple expansion 4-4-0 and built the ‘M’ and ‘Q’ classes between 1892 and 1897. ‘M’ class No. 1621 has been preserved. Apparently following the instructions of the Board and prior to the Locomotive Committee’s decision, one ‘M’ class engine, No. 1619, was built as a two-cylinder compound. This locomotive was severely damaged in a collision in 1898 and was renewed as compound, but on the three-cylinder system devised by WM Smith, the company’s Chief Draughtsman. No. 1619 remained a one-off on the NER, but the three-cylinder compound system was taken up enthusiastically by Mr Smith’s friend, Samuel Johnson, on the Midland Railway. Therefore, the famous ‘Midland Compounds’ and the later LMS development can trace their origin to the NER2.
A larger development of the simple-expansion 4-4-0, the ‘R’ class, was introduced in 1899 and sixty were built between then and 1907. These were truly excellent locomotives where the designer managed to achieve a good balance between the various components of the design. The coupled wheels were smaller than his earlier 4-4-0s, the cylinders were the same size, but the boiler was of a larger diameter and the working pressure had been increased to 200psi. Importantly, they were fitted with piston valves, rather than the slide valves fitted to earlier locomotives.
An 'R' class 4-4-0 No. 592 in beautiful NER 'Saxon Green' livery. A small production run of this variation of the Union Mills 'D20' was commissioned by Mr Jeremy Burrows, to whom I am especially grateful as it is a gorgeous model. There is more information about this class of locomotive in the Union Mills Locomotives on Poppingham page of this website.
At this time, the earlier enthusiasm for speed, as evidenced by the ‘Races to the North’ of 1888 and 1895, had receded and the emphasis was moving towards enhanced passenger comfort, which meant train weights were increasing. Wilson Worsdell’s response, also in 1899, was the first British 4-6-0 passenger locomotive, the NER ‘S’ class. These ten engines resembled a stretched ‘R’ with smaller coupled wheels and outside cylinders. Unfortunately, most of them reverted to slide valves. They were a disappointment on express passenger trains, with the smaller ‘R’ class being significantly better performers. The 4-6-0s were more at home on fast goods services and 30 more of the class were built, specifically for this work, between 1905 and 1909.
Undeterred by the uninspiring performance of the ‘S’ class, Wilson Worsdell introduced the ‘S1’ class in 1901. These engines had larger diameter coupled wheels and all five built were fitted with piston valves. Although the ‘S1’ class’ performance was better than the ‘S’ class, with Charles Rous Marten, the leading train timing and performance expert of the period, being impressed with their running, no more were built and NER express passenger locomotive design proceeded in a different direction. Meanwhile, the ‘R’ class continued to perform truly excellent work.
The reason for the change in direction was a delegation of NER officers visited the USA in 1901 and Wilson Worsdell was impressed by the work done by ‘Atlantic’ locomotives there, particularly on the Pennsylvania and Philadelphia and Reading railroads. On his return to Britain, he gave instructions for a 4-4-2 express passenger locomotive to be designed. Although he had a forceful personality, Mr Smith, by then, was in indifferent health and was away from work when the new locomotives were being designed, with his deputy, George Heppell, taking charge of the work. The first NER ‘V’ class 4-4-2 appeared in November 1903 and a further nine were soon in service.
Mr Smith let it be known that he was not at all impressed by the 'V' class, which led to an exceptionally gracious response from Wilson Worsdell. With approval from the Board, Mr Smith was given authority to design his own ‘Atlantic.’ It is scarcely believable that this happened and the two locomotives which were built were credited to WM Smith rather than Wilson Worsdell. One can barely imagine what would have happened if Dugald Drummond had been Mr Smith’s superior!
It appears the design was worked out on a ‘when time permits’ basis between 1903 and 1905. Mr Heppell, being unhappy with Mr Smith’s criticism of the ‘V’ class, declined to be involved and Ralph Robson produced most of the drawings. The two locomotives, Nos. 730 and 731, appeared in 1906 and were four-cylinder compounds3 with, uniquely on the NER, Belpaire fireboxes. No 730 had Stephenson link motion and No. 731 had Walschaerts valve gear. On test, No. 731 had superior coal consumption. Interestingly, as part of the test, coal consumption figures for the ‘V’ class locomotives, considered to be heavy on coal, were compared and, although the simple-expansion engines used more coal, there wasn’t that much in it. There was not a clear advantage for the four-cylinder compound over the two-cylinder simple-expansion locomotive. Nevertheless, the construction of ten more compound ‘Atlantics,’ based on No. 731, was authorised in 1907.
In a fascinating, twist, Mr Smith had been granted two patents for features included in the four-cylinder compounds. However, he died in October 1906 and it appears that his estate demanded royalties if further locomotives were built incorporating these features. Not surprisingly, the NER cancelled the authorisation and no more of these locomotives were built. And, at this point, the story takes another interesting turn!
The NER was a forward-thinking railway and had a dynamometer car and, in the autumn of 1907, tests were carried out between York and Newcastle using four locomotives: a new compound 4-4-2, a ‘V’ class 4-4-2, an ‘S1’ 4-6-0 and an ‘R’ class 4-4-0. Unsurprisingly, as it was the largest engine, the compound had the highest drawbar pull and the others’ drawbar pull was ranked in accordance with their size – in descending order, ‘V,’ ‘S1’ and ‘R’. But, when the results were tabulated as drawbar horsepower (DBHP) per ton of locomotive weight, the order was: ‘R,’ compound, ‘V’ and ‘S1’. Similarly, ranking as DBHP per square foot of grate area resulted in a draw between the ‘R’ and the compound, followed by the ‘V’ and ‘S1’.
The especially efficient performance of the ‘R’ class in these tests might well be the reason for the NER deciding to revert to a 4-4-0 for express passenger work and 10 massive simple expansion locomotives of the ‘R1’ class were introduced in 1908. These engines were of approximately the same power as the compound ‘Atlantics’ and did some good work, but they lacked the sparkle of the ‘R’ class and no more were built. Not for the first time did an engineer learn that bigger is not necessarily better.
As a result of the experience with the 'R1' class, the next express passenger locomotives built by the NER were 10 more ‘V’ class 4-4-2s in 1910. Wilson Worsdell retired in 1910 and was succeeded by Vincent (later, Sir Vincent) Raven. He finally resolved the express passenger wheel arrangement trilemma in which the company had found itself, by favouring the ‘Atlantic’ type, with his excellent three-cylinder simple expansion ‘Z’ class, of which 50 were built between 1911 and 1918. Apart from the first 10 as built, these locomotives were fitted with superheaters and proved to be a worthy successor to the ‘R’ class.
Interestingly, in 1918, as part of the NER’s response to the government’s request for designs for national standard locomotives, outline designs were produced for another 4-6-0, effectively a six-coupled version of the ‘Z’ class. None of these were built and the NER’s final express passenger design was Sir Vincent Raven’s ‘Pacific,’ the first of which appeared in shop grey right at the end of 1922. The remaining four entered LNER service after the Grouping of the railways in 1923, but it was HN (later Sir Nigel) Gresley’s ‘A1’ ‘Pacific’ that was chosen for future LNER construction.
The NER was the third largest British railway company by route mileage4 in the period we have been discussing and was a very well-run concern. Although they give the appearance of being haphazard and lacking a firm sense of direction, I think the events I have described are better seen as its reasonable response to a time of great development in locomotive design, spanning the period from the second flowering of the single-driver type to the coming of the ‘Pacific.’
Some common themes are well demonstrated by the NER story: the usual excellence of the British inside cylinder 4-4-0, the years of indecision between the 4-6-0 or the 4-4-2, the marginal (in the British context) benefits of compounding and, finally, the triumph of superheating.
From the early Twentieth Century, the Great Western Railway (GWR) led the way with British locomotive design. Mr Churchward’s detailed tests demonstrated to his satisfaction that the simple-expansion 4-6-0 was better suited to his railway’s needs than an ‘Atlantic’, whether compound or not. His early adoption of superheating (albeit low degree) and long lap/long travel valves put the GWR at the forefront of British locomotive performance.
Probably the next best company for express passenger locomotive policy in the Edwardian period was the Great Northern Railway, with Mr Ivatt’s splendid large-boilered ‘Atlantics’ and, later, the first Gresley ‘Pacifics.’
Many other companies during the period from 1900 to the Grouping were struggling to move from their usually very good 4-4-0 designs to larger express passenger locomotives. For instance, the Midland didn’t try, the Great Central couldn’t make up its mind5 and reverted to the 4-4-0, and, after some early success with 4-6-0s in Mr McIntosh’s time, the Caledonian appeared to lose its way, with Mr Pickersgill’s three 4-6-0 designs of 1916-1922 being particularly indifferent performers.
Regarding other ‘Edwardian’ 4-6-0s, honourable mentions should be made of the GER ‘1500’, the LNWR ‘Prince of Wales’, the Highland ‘Castle’ and G&SWR ‘128’ classes.
It is worth noting that the next step forward in locomotive design occurred in 1917 when REL Maunsell introduced his ‘N’ class 2-6-0s on the South Eastern & Chatham Railway. With a taper Belpaire boiler, a high degree of superheat, two outside cylinders and Walschaerts valve gear operating long lap/long travel valves, these were genuine Twentieth Century engines.
In preparing these notes I have referred to many valuable sources but any errors or omissions are mine. I have not cluttered up the post with references but these are, of course, available on request.
1 The Stockton & Darlington Railway, part of the NER from 1863 but managed independently until 1876, had three designs of outside-cylinder 4-4-0s dating from 1860, 1862 and 1872.
2 Walter M Smith’s son, John Smith, was a draughtsman at the Midland Railway’s Derby works, becoming Chief Draughtsman there in 1901.
3 It is indeed strange that WM Smith, famous for three-cylinder compound locomotives, changed to the four-cylinder type late in life. John Smith had accompanied the Midland 4-2-2, No. 2601 Princess of Wales to the Paris Exhibition in 1900. Also exhibited was the first Nord compound 4-4-2 and John Smith took a keen interest in this engine. Presumably, he discussed it with his father who decided to use the same system for his compound ‘Atlantics.’ Mr Churchward, on the Great Western, was also impressed by the Nord compound 4-4-2 and No. 102 La France entered service on the GWR in 1902, followed by two larger French compound 4-4-2s in 1905.
4 In terms of route mileage, the NER was behind the GWR and LNWR, but ahead of the Midland. Next were the NBR and the GER. This 'top six' all had route mileages above 1,000. These figures are for 1913.
5 In the 1900-1921 period the Great Central built four 4-6-0 express passenger classes, two with outside cylinders, one with inside cylinders and one with four cylinders. There were also five types of mixed traffic or fast goods 4-6-0s, with the same mix of cylinder configurations. This was in addition to three types of express passenger ‘Atlantics.’ But it was the ‘Director’ 4-4-0s of 1913 and the later ‘Improved Directors’ that were the definitive GCR express passenger locomotives.