Faces were saved in 2000 when a deal was struck with VIA Rail, operators of Canada’s passenger trains. The purchase price, at $130 million, was widely considered a bargain. The Nightstars went off from Newport docks, still with their multi-currency cash tills in place in the catering cars. Now, vehicles that were meant to allow the sleeping Glaswegian to pass the sleeping Dortmunder somewhere in the darkened Low Countries make their way across the spacious landscapes between Toronto and Halifax, Nova Scotia.
Meanwhile, we can still travel on the remaining sleeper trains. Can, and should: for there is nothing else like them. It is not simply the singular experience of having a little bedroom on a train – a stranger feeling still, now that ordinary seated compartments have disappeared. Nor is the appeal necessarily much to do with convenience, especially now that the cheap-flight merchants are criss-crossing the skies. There is a certain caravan-like charm in the little space-saving devices of the compartment, such as the fold-down lid to the washbasin that doubles as a sort of window-sill, but that is only a small part of it. It would also be difficult to claim that the sleeping compartments are especially peaceful spaces, with their background roar from the air conditioning, and the mysterious creaks, squeaks and groans when in motion.
No; what makes sleeping and awakening on these trains so special is the promise of magical translation to an utterly different place. Recalling Scottish excursions, the critic and essayist Max Beerbohm (1872–1956) considered that ‘nothing is more mysteriously delightful than this joint consciousness of sleep and movement’. The mood of grateful surrender to mechanical power is beautifully caught on celluloid in Powell and Press-burger’s I Know Where I’m Going (1945): the heroine’s night journey to Glasgow is represented not by stock footage of a speeding train, but by a dream-sequence in which a tiny model version glides through mountains made of tartan blankets, as an unseen voice sings a lullaby. Sleep abolishes time; it follows that sleeping through the night on a moving train abolishes distance as well as time. This abolition of time and space was the most astonishing thing about the railways when they were new. To slip into dreams somewhere past Watford or Rugby and to raise the blinds on waking to a view of the stern rocky mass of the Boar of Badenoch at Druimuachdar summit, the highest point on the railway network, is thus to join hands across the decades with the first travellers by rail. So we may feel again as Thomas Carlyle did, of his first nocturnal railway journey of 1839, ‘as if some huge steam nightbird had flung you on its back, and were sweeping thro’ unknown space with you’.
There is more. Sleeper trains include a lounge car, a sort of joint descendant of the family saloon and the observation car, with leather sofas and moveable armchairs that positively plead with the traveller to gaze out of the big windows. There are no headrests or Pendolino-type blind spots here to block the morning or evening view. Passengers make their way to this saloon, as they formerly did to the dining car when a full menu service was available, or as they might move from cabin to saloon deck on board a ship, with a sense of joining a social space. The peculiar solipsism of railway travel – a solipsism that was observed in the earliest compartments, and which smartphones and tablet computers have made worse than ever – is laid aside. People make eye contact and exchange words. It is discovered that there is value in ‘this frail / Travelling coincidence’, to quote from Philip Larkin’s ‘The Whitsun Weddings’ (1964), one of the few railway poems in English that can give ‘Adlestrop’ a run for its money.
Even better: there is a bar at the end of the carriage. Better still: it will serve you an alcoholic drink after 9 p.m., the time at which drinking on trains that run wholly within Scotland became illegal on 20 July 2012, in the interests of public order. Lulled by a whisky – though not by as much of the hard stuff as Larkin was prone to put away – we can fall asleep to the intermittent rhythm of the rails. As for this book, it is time to leave the train, to explore the railway on which it runs.
PART II
DOWN THE LINE
– 9 –
THE PERMANENT WAY
Everyone knows the noise a moving train is supposed to make: a sort of diddly-dum rhythm, four beats close together, repeated over and over. Everyone knows it; and yet the music of wheels on rail joints is unusual now. Modern track uses continuously welded rails (cwr), made by fusing together many shorter lengths that have been rolled separately. Underground passengers were the first in Britain to benefit from the improved ride offered by welded rails, after London Transport adopted the technology in 1937. British Railways followed suit in the late 1950s, and by 1974 continuously welded rails had already been laid on almost a third of track-miles across the network.
Each welded rail may be a mile in length, far too long to be carried on a single train; instead, shorter composite lengths are carried to the site of use, to be welded together as they are laid down. To the passenger in transit these welding points are undetectable, and the use of tapering joints ensures that even the noise made when the train crosses from one continuously welded rail to the next is barely noticeable. Noise reduction is not the only advantage: trains benefit from reduced vibration and friction in terms of maintenance costs, lifespan and lower energy input, and the smoother ride allows much higher permitted speeds.
Whatever its length, each rail will expand and contract according to variations in temperature. The longer the rail, the greater the potential movement, always with the risk of buckling in the hottest weather. Continuously welded rails therefore have to be tensioned as they are laid down, either by mechanical stretching or by heating, before being fixed firmly in place. Any expansion after that is absorbed by the elastic tension within the rail, which remains both taut and firmly anchored to the sleepers. This tension explains the singing noise the rails make when a fast train approaches.
Rails of the traditional type are inert by comparison. With their sleepers, they are laid down in panels of standard length, which in Britain was fixed at 60ft in the second quarter of the twentieth century. Each join is secured by bolting pairs of bars, called fishplates, through holes drilled in the web, the name given to the thinner upright section in the middle of the rail. A gap for expansion is left between adjoining rails, and the holes in the fishplates are elongated so that the greased bolt-heads can slide within them as each rail gently shrinks or grows. The colder the weather, the shorter the rails and the larger the gaps, so that winter trains often make more noise than summer ones; thus the Great Western in the 1930s allowed fully half an inch as its upper limit. The familiar four beats – two close together, a slightly longer gap, then another two closely spaced – are made by the wheels of the twin-axled bogies at the end of each carriage as they cross joint after joint of this type.
The fourfold beat is really a twentieth-century sound, and not the usual rhythm the Victorians knew. As we have seen, most carriages then had four or six wheels, on axles that were more widely spaced. The rails were shorter too – in the first years of the passenger railway, very much shorter. In the form first perfected for waggonways around 1790, this type was no longer than the span between one sleeper and the next. Each rail therefore needed support at the ends only, and from the curved profile of the underside this type acquired the name of fish-bellied rails. Cast-iron manufacture of fish-bellied rails gave way by the 1820s to wrought iron, which was less brittle and crack-prone; the Stockton & Darlington and Liverpool & Manchester both used rails of this material.
The next step was to make longer rails of wrought iron, without the fish-bellied profile. These in turn yielded from the 1860s to rails rolled from steel, which promised to last at least four times longer. Steel rails were also heavier, although at first barely more than half the weight of the thick rails now used on main lines. Lengths have also increased; on the Midland Railway, the progression went from 20ft rails in 1850 to 24ft in 1875, 30ft in 1884, 36ft in 1896 and 45ft a few years after that. Rails of different lengths would therefore have been encountered on most journeys. Altogether, the rails and
wheels of the Victorian railway would have given a less jaunty rhythm, closer to an incessant thumping or banging, amplified by the basic suspension and noise insulation of the carriages of the period. The harsh, halting beats of today’s four-wheeled Pacer carriages on jointed track give a sense of how fatiguing the experience must have been.
Track with bullhead rail, as explained in I-Spy on a Railway Journey, 1963
Jointed rails may still be found all over the modern railway, especially in sidings and alongside the platforms at stations, where maximum speeds are low. Much of this is of flat-bottomed outline like that of the welded rails, with a profile that spreads out below. This is a very old form, and a few railways adopted the type as early as the 1830s, but it became the national standard for new running lines only in stages from 1949. The ‘bullhead’ form of rail that was near-ubiquitous before then is not hard to spot. It is more symmetrical in profile from top to bottom, like a swollen letter I of which the upper cross-stroke is thicker and fatter than that below. As this lower part is not wide enough to sit upright, the rail depends for its support on heavy cast-iron chairs that are screwed or bolted to the sleepers. To anchor the rail firmly within these chairs, fat wedge-like fastenings known as keys – originally of oak or sometimes teak, latterly of folded sprung steel – are hammered in between the chair and the outer side of the rail, against the web. Together, the chairs and keys hold the rail with its upper surface at a slight angle to the horizontal (the standard figure on straight track is 1 in 20), to offer a continuous surface to marry with the tapered steel tyres of railway wheels.
Most of the sleepers used in Britain over the years have been of wood. When railways were young, however, sunken stone blocks were used instead. They were drilled with holes for fixing the chairs, and each one sat beneath a single line of rail. Cross ties were often dispensed with in track of this kind; the weight of the blocks themselves was considered enough to stop the rails from spreading. For the horse-drawn routes on which such blocks were first used, there was a clear advantage too in leaving a level pathway down the middle. Stone blocks also had the attractions of durability, outweighing the cheaper initial cost of timber sleepers, and rigidity. The last point is particularly important: engineers around 1830 imagined the railway as a sort of giant horizontal machine, trains and track together, in which hardness and firmness were just as much at a premium as they were within a stationary engine. The London & Greenwich even set its granite sleepers in concrete, to make doubly sure.
This policy was a mistake. Quite soon, it was realised that the shock-absorbing qualities of wood were a positive advantage, reducing wear and tear on the rails and offering a smoother ride. The case was demonstrated in 1837 on the Grand Junction Railway, named because it would connect two early inter-city lines into a single trunk route. These lines were the Liverpool & Manchester and the London & Birmingham, each of which had mostly employed stone sleepers on its tracks. The Grand Junction’s engineer Joseph Locke used a mixture of stone and temporary timber sleepers, intending to upgrade the timber sections in due course. Once the differences could be compared, it was the stone blocks that were discarded. Other lines followed suit, which meant that a great many redundant stone blocks of similar size became available. Many were reused within the railway network, with its constant need for materials of every kind. Second-hand sleeper blocks thought to have come from the Liverpool & Manchester turned up during the demolition of a coal depot at Shepherds Bush in 1974, when the railway historian J. B. Atkinson rescued a few at the price of 50p each, payable to British Rail. Perhaps the oddest instance of reuse is still passed by thousands of unwitting railway travellers every day: the prodigious neo-Gothic church of St Walburge, just north of Preston station, built in 1867 using limestone sleeper blocks discarded by the Preston & Longridge Railway in its lower walls.
The drawbacks of rigid running were also discovered by Isambard Kingdom Brunel on the Great Western. Independent-minded as ever, Brunel chose to make his broad-gauge track of a form known as baulk road, in which each rail was supported by continuous longitudinal timbers laid on a compacted bed. Cross timbers were still needed to keep the rails at the correct distance of 7ft 0¼in apart, and these were placed at intervals of 15ft. But Brunel wanted to make his track firmer still, so the cross timbers were anchored by piles driven 15ft into the ground. It did not work. The earth and gravel ballast around the track subsided with time, and the impact of passing trains forced the piles further into the earth at uneven rates. The timber framework also sagged between the points supported by the piles, and the ride became a queasy dipping up and down. So the piles were quietly removed and new broad-gauge routes for the Great Western and its associated lines in the West of England and South Wales did without them. Released from the technological cul-de-sac of the pile, the broad gauge delivered what Brunel had promised, a smoother ride for passengers than stone-sleepered standard-gauge lines. The sounds of travel were different too: the ends of broad-gauge rails were not aligned, doubling the number of joints encountered by the six-wheeled carriages and turning the noise and vibration to an irregular clatter.
The rails on the baulk road were rolled to a distinctive up-and-down section, with horizontal flanges on each side. Holes in these flanges allowed the rail to be fixed to the timbers. The type was known as bridge rail, and versions of it are still used on some railway bridges, where the rails are fixed directly to the structure. On other lines, our traveller of 1862 might also have encountered the Barlow rail, which had a profile like a broad inverted V with the sides curving outwards to a distance of thirteen inches. This width allowed the rails to be buried stably in the ballast without any sleepers at all, and with only iron cross ties to link them. When a train passed in dry weather, these rails tended to throw up a lot of dust; when waterlogged, they were known for squirting out mud.
In another crucial respect, the physical form of the railway in the 1860s was much less diverse than it had been around 1840. In those years, not every railway engineer was yet thinking of how the various lines might be linked in a national system. The Eastern Counties Railway is a case in point. By 1843 its main line from London was open as far as Colchester, to a gauge of 5ft: incompatible with the standard gauge of 4ft 8½in, without offering any compelling advantages over it. When the railway was being planned in 1836, the exact figure must have seemed unimportant; not so by 1843. During the course of one costly and laborious month in the year following, the company therefore converted all its tracks and rolling stock to the standard width. Its neighbour in the East End, the London & Blackwall Railway, did the same in 1849; the original gauge here was 5ft 0½in. North of the border, ‘Scotch Gauge’ (4ft 6in) was a common choice for horse-drawn mineral railways of the central belt, and the first Scottish passenger railway followed local usage: this was the Garnkirk & Glasgow, opened in 1831. Irish gauges were all over the place too in the early days, ranging from the standard British width to an indulgent 6ft 2in for the first railway built in Ulster. These liberties ended in 1843, when the Board of Trade decreed that Ireland should adopt 5ft 3in as standard. The gauge remains in use, and visitors will notice that Irish tracks have a spacious look by comparison with those in Britain and on the Continent; the difference also explains why there have never been train ferries across the Irish Sea.
A decision on Ireland having been reached, the British nettle stood next in line for grasping. The question was settled by Act in 1846, a decisive early episode of state intervention in the operation of the railways. The Act took account of the deliberations of a Royal Commission that sat during the previous year, to which most of the senior figures within the industry had presented their thoughts. What was under scrutiny by then was not the residue of near-standard gauges, which could be left to the companies to sort out themselves, but the disparity between standard gauge and Brunel’s version, fully 2ft 3¾in wider.
Months passed while the country waited to learn how far the Commission’s findings would be given legal force. Me
anwhile, the pamphlets flew. Some of these came from the pen of the civil servant and design reformer Henry Cole (1808–82), one of those energetic, unclassifiable figures who pops up everywhere in the Victorian world. In 1843 he had commissioned the world’s first Christmas card for sale; in 1850 he got the Great Exhibition under way. Helping to settle Brunel’s hash was another of Cole’s projects. His pamphlet entitled Inconsistencies of Men of Genius, Exemplified in the Practice and Precept of Isambard Kingdom Brunel, Esq. was unsparing. Much of its text consisted of self-incriminating material from Brunel’s own statements, especially from a sticky interlude in 1838–9 when a faction of Liverpool shareholders in the Great Western had tried to jettison their chief engineer and his clever ideas. Cornered, Brunel had been forced to admit that the broad gauge ‘could have no connexion with any other of the main lines’. As Cole put it, ‘we have every practical railway engineer denouncing the Broad or seven feet Gauge, and only Mr Brunel attempting to defend it’.
He was not far wrong; the consensus against Brunel’s gauge was decisive. The engineer enjoyed a few good moments when giving evidence, and even managed to persuade the Commission that trials should be held to show the relative merits of the gauges. The Great Western’s trial train easily vindicated Brunel’s claims for superior speed and comfort, but that was not really the point; and the outcome was easy to foretell. Under the ensuing Act of 1846, new passenger lines within Britain were to be of standard gauge unless by special exception. The broad gauge was allowed to continue, subject to some lines being converted for joint operation, and to make extensions within its existing territory. Thus the Great Western and its allies in the western lands were spared the immediate disruption and expense of forced conversion.
The Railways Page 32