Rolling a wheel along a prepared track or 'way' is an old idea, the Babylonians carved ruts in stone paved roads to guide wagon wheels along. This technique was also used by the ancient Greeks and after them the Romans, who built some roads in Britain with ruts at their standard wagon wheel spacing or 'gauge' of about five foot (1.5m).
In the early seventeenth century wooden tracked railways used wooden wheels with flanges running on rectangular section rails (technically called 'edge rail') appeared at British mines and reached quite high levels of sophistication. In 1992 a short section of wooden track including a set of wooden points, was unearthed at the site of an old iron works at Bersham near Wrexham. The track is believed to date from the early eighteenth century. The rails were made of elm attached to oak cross-ties or 'sleepers', the two being held together with wooden pegs, the gauge or distance between the rails was just over four foot (1.22m). The word 'sleeper' in this context dates from about 1789 and comes from Newcastle (where many local terms are based on old Norwegian words), it has the same root on Old Norse as the word 'slab'. The word 'rail' was already in use for things made from lengths of wood and this is probably the origin of the term railway (first recorded use 1681) or rail-road (first used as a description in 1702).
The term railway and tramway both mean the same thing, rail comes from the French 'raille', tram from the Germanic 'traam', both meaning a length of timber or plank. In Britain the term tramway became associated with light lines operating on or beside conventional roads, the tracks being commonly in-set into the normal road surface. The term 'wagon way' was used to describe some early lines but from about 1850 the British only used the term wagonway for light lines using horses or men to move small wagons and settled on railway for anything bigger.
Generally speaking the wooden track used at mines in the North East favoured a gauge of around three feet whereas those further south could have gauges of up to five feet. The difference is believed to be due to the types of coal mine, the deeper pits of the North requiring smaller wagons with necessarily smaller gauge.
In about the 1730's people started adding iron tyres and even iron spokes to wheels. Casting iron wheels or tyres with a flange proved difficult at first and an alternative was to use normal un-flanged wheels (usually fitted with iron tyres) running on L section iron rails. This kind of track became known as a 'plate way' and they were favoured by Benjamin Outram, one of the pioneers of tramways in the late eighteenth and early nineteenth centuries. Plate ways were initially used to transport goods short distances from mine or quarry to a nearby canal but some lines were up to thirty miles long. Outram favoured a gauge of four foot six inches, which was also the standard distance between cart wheels in his home area. He believed ordinary road carts could use his tramways but in practice the reverse happened with his flat wheeled wagons being dragged off the plateway for road delivery to the customer. L shaped metal track was common by the 1870's, reaching the Welsh quarries and mines in the 1890's.
The L section plateway rails were bolted or spiked down to wooden cross ties (or sleepers) to keep them a constant distance apart. This was building on existing technology developed for the wooden railways in mines and quarries. Unfortunately untreated wood had a tendency to rot and after about 1800 various people began using stone blocks laid in rows under each rail. There was nothing linking the two rails together but the blocks were heavy enough to prevent the tracks moving. Generally the L's faced outwards and the stone blocks were usually either square or roughly circular with a hole somewhere about the centre. A wooden plug was driven into the hole and the track was secured to this with a metal spike.
This L section track is of relevance to the modeller because it remained in use for 'wagonways' into the 1940's. These light lines were used for moving stone blocks from a quarry to a railway loading bank and for small trucks carrying coal or ore from minor mines to a nearby railway. Some of these lines used steam or petrol engined locomotives but most used men and horses to move the wagons. This kind of track can be represented using Plastruct 'fineline' (polystyrene) L section but this is really rather too heavy and it is better to build up lengths from Slaters 10x20 thou microstrip glued into an L section. Typically the rails would be perhaps three foot (1m) long, often with the stone supports only at their ends. The stone sleepers or setts were usually of fairly regular shape, typically a foot or so (30 cm) across and one option is to use a leather punch to obtain small discs from ten thou plastic card.
Note you cannot use long strips and form them into a curve, the rails were straight and curves were formed from a series of 'dog-leg' bends, presumably using rails of slightly different lengths. If you do make the rails from long lengths bend them at the 'joints' to form the curve. Normally, judging from photographs, on long straight sections the sleepers seem to have been set in pairs directly opposite each other.
Fig___ L shaped track on stone sleepers
In the early 18th century a firm in Shropshire produced cast iron flanged wheels for use on wooden rails. In the 1760's an iron foundry, again in Shropshire, started laying iron bars along the top of their wooden rails partly to protect the rails and partly to stockpile the iron plates. The men who did the work were called 'plate layers', a term which remains in use today for staff employed in laying and maintaining railway track.
Crude cast iron rails were first used by Richard Reynolds in 1762 but the ancestor of the modern rail was devised by William Jessop in 1789 (Mr. Jessop also introduced the first workable iron points at the same time). Jessop favoured the flanged wheel with flat topped 'edge' rail over Outram's L shaped rail and although he was in partnership with Benjamin Outram the two seemed to have agreed to differ on this point and their firm produced both types of track depending on who was in charge of a particular project. Jessop's iron rails were made in three foot (1m) to six foot (1.8m) lengths, there were side plates cast onto the rail through which bolts secured the rail to stone sleepers and the centre of the rail was of deeper section than the ends. This is called 'fish-belly' rail and it was used for a number of light wagonways from the 1780's.
Fish bellied rail was not as common as the L shaped rails on minor tramways and wagonways feeding full-size railway loading points but it remained in use in some remote locations up to the 1940's mainly carrying horse drawn or man-pushed mineral wagons. It is less easy to model than the L section type and probably not a practical proposition in British N although you might manage to glue short lengths of 10x10 thou strip to the bottom edge of 10x20 thou strip. These rails were only made in straight lengths so again if laying a curve from a single length of strip remember to put it down in a series of 'dog legs'.
Fig___ Fish-bellied rail laid on wooden sleepers
This sketch taken from a short length of preserved tramway track, each rail is only about four foot long (1.3m).
In 1797 edge rails were cast with no built-in feet or fixing holes, these were mounted in cast iron brackets called 'chairs' and the early examples used wooden sleepers. This system of rails and chairs became the norm for railways built after the beginning of the nineteenth century.
In 1820 an ironmaster called John Birkinshaw patented an improved rail with a thicker section at the top and bottom, forming an I shape and this basic design became the norm for railway track. The rail lengths were laid in the brackets secured to either stone slabs or wooden sleepers and a wooden wedge was driven in to hold the rail in place. In early days the wedge was driven in on the inner face, after about 1900 everyone changed to driving it into the outer gap. This shape of rail and method of fixing became the norm on British railways. In the 1970's the wooden chocks were replaced with spring steel chocks.
Fig___ Light ' I ' section rail
At the rail ends Birkinshaw added special chairs to hold the rails together but in the 1840's someone came up with the 'fish plate', these comprise two short lengths of metal bolted across the join at the end of the rails. The plates are just over a foot long (30cm) and have four bolts passing through them. Two-bolt fish plates were tried in the late 1930's so that the sleepers could be placed closer to the rail ends and provide greater support, this did not prove to be a success however.
The joints are where the most problems occur in track and the unsupported 'fishplate' joint suffered greater wear than Birkinshaw's joining chairs but fishplates and bolts remained the standard system for joining rails. Welded rail (discussed later) is now standard for all main line tracks but fishplates are still used in yards or where points or other complex track is joined.
Fig___ Photo of bullhead I section rails joined with fishplates
Most of the wear on a rail happens on the top surface but experiments with 'reversible' rails that could be turned over were not a success as the 'chairs' which hold the rail wore on the underside in use. This lead to the development of 'I' section rails with a thicker section at the top called 'bullhead rail' which became the standard in Britain until after the Second World War.
The design of rails has evolved quite slowly as technology improved. Cast iron tends to fracture and there was much debate as to the best cross-sectional shape. One of the pioneering railway engineers, a chap called Vignoles, designed an I section rail with a flat bottom and a thickened top edge for the Liverpool & Manchester line in the 1830's. Vignoles wanted to bolt the flat bottomed rail down onto to timber strips running along the under-side to support it with wooden cross sleepers to keep the rails the correct distance apart. This arrangement for supporting the rail is called 'baulk road' and in the South West I. K. Brunel used the baulk road idea for the Great Western Railway. Brunel used an inverted U shaped rail called 'bridge rail' for his track.
By the 1840's most British companies were using bullhead rails carried in cast iron 'chairs' but in many parts of the world the Vignoles flat-bottomed rail was preferred, held down with spikes hammered into the sleepers.
The very first steel rails used in the UK were laid at the Midland Railway's works at Derby in 1857 but they were thrown away after sixteen years service. There was some initial doubt about using steel for rails because although it lasted much longer at the time it had no 'scrap' value. This soon changed however and from the 1870's rolled rails of 'Bessimer' steel had become the norm for all new track.
Rail is supplied to the railway company in straight lengths and is transported in this form through the system. The standard rail length stabilised at about thirty foot (9.1m) until after the First World War when forty foot (12.2m) and forty-five foot (13.7m) rails appeared as steel rolling technology improved. By the 1930's sixty foot (18.3m) rail lengths were standard on main lines and rails of up to a hundred and fifty or so feet were supplied to the LNER and LMS companies. With Nationalisation the sixty foot rail was adopted as the standard. Longer rails required longer wagons to transport them and bogie designs were introduced by several companies as rail lengths increased.
Rail taken from Peco or other track is not suitable for representing rail loads on wagons as it is way over scale. The best option is to use the finest 'I' section from the Plastruct range (reference BFS-2), do remember that all the rails would be of the same standard length and take care when cutting.
Where curves and complex track such as points and crossings were needed they were made on the spot from standard rail lengths. Teams of men cut the rail with hacksaws and bent the rail with levers, chocks and pulleys, power tools did not exist so every hole had to be drilled in the metal using hand drills.
As the train passes through the curve the most wear occurs on the inside face of the outer rail. To help reduce the wear on very tight curves some companies added a felt pad, laid against the rail and with its bottom edge sitting in a trough of oil. As the wagons passed they pressed the felt, picking up some oil on their flanges as they did so. An alternative system employed small plungers, operated by the wheel flanges, which squirted grease onto the rail side. Even with the lubrication the rails tended to wear and standard practice was to exchange the inner and outer rails every few years to even the wear.
Rails are defined by their weight per foot or per yard and the weight of rails has risen steadily over the years. Early lines used iron rails of as little as forty pounds per yard, quite soon steel rails of eighty pounds per yard appeared and by the turn of the century everyone was using ninety odd pounds per yard for new rail. In the early twentieth century, following an accident at a major main line junction, it was found that rails ranging down to forty pounds per yard were still in use, prompting the companies to check and replace the older track. In 1921 the Government laid down a national standard of ninety four and a half pounds per yard plus or minus half a percent (this was in fact slightly lighter than the GWR's standard rail at the time). These days the rail weight is not nearly so standardised I contacted Railtrack on this in 1999 and they advised me that although they could quote for specific sections the data for the heaviest and lightest on the network was not directly available.
I enquired on the newsgroup uk.railway regarding the actual weight of rail and received the following replies:
I thought Railtrack standardised on UIC60 rail (60 kg/m) some years ago, at least for new main-line schemes. However, I don't know whether they continued to buy earlier specs for replacements on other lines.
UIC 60 (60kg/m) only seems to be used on the higher speed main line routes, the only places I've noticed it is on WCML renewals, although it may be used elsewhere. Otherwise most renewals I've seen still seems to be using Flat Bottom 113A (113 lb / yd) on F 40 sleepers or EF 28 sleepers on Southern Region. (It seems the use of "zone" has been discontinued)
All the UIC 60 I've seen has been laid on G 44 sleepers with pandrol fastclip fixings, but as I've never actually re layed any I couldn't say if it can be laid on F 40 type sleepers for standard re railing or whether its use requires G44 sleepers and therefore is only practical for use on renewal projects. (There may also be load gauge issues associated with its use for re railing as it is a deeper profile.)
To put these comments in context with the earlier figures 60kg is 132 lbs and 1m is 1.09 yds so UIC60 track weighs in at 121 lbs per yard. UIC60 rail is a heavy duty rail and has been approved by the International Union of Railways (UIC) for use on high-speed lines throughout Europe.
The payload of goods vehicles has increased steadily since the Second World War, in 1948 the maximum weight per axle was seventeen and a half tons. This was increased in 1962 to twenty two and a half tons giving rise to the 'big' forty five ton wagons such as the Peco long wheelbase tank wagon. In 1966 the axle weight was again increased, to twenty five tons, but initially this was only on specific routes. By the 1990's most of the network was allowed to carry twenty five tons per axle and on some routes greater loads are permitted, allowing the very large one hundred and two ton bogie tank wagons to operate.
In the days of steam hauled trains the heaviest element in the train was normally the locomotive, differing weights of rail and the strength of viaducts and bridges under the track all contributed to restrictions on the use of the heavier locomotives. Each line was classed in a table of 'route availability' (usually abbreviated to RA) on a scale of one to ten. The locomotives were similarly classified and were allowed to operate only on lines which were able to bear their weight. Typically the locomotive would have a coloured disk painted on it to indicate the minimum RA class on which it could safely operate.
Post privatisation Railtrack retained the RA classification system, although it is normally the freight vehicles these days which push the limits of the per-axle weight.
Track rated at RA1-6 is the most common and can carry up to 20.3 tonnes per axle, RA7-9 allows axle weights of up to 24.1 tonnes and RA10 can carry up to 25.4 tonnes per axle. in the 1990's RA7-9 was used for a network of freight lines around London, for the West Coast Main Line up to Carslile and in areas where there are industries which generate heavy freight flows. RA10 was confined to a couple of lines in Scotland, a similar number in Wales and a number of short sections which regularly handled very heavy freight.
As noted above the joints in a track are where most of the wear occurs to both rails and wheels, nearly half the work done on conventional track is in maintaining the joints. Modern practice favours the use of long lengths of welded rail, made up of standard sixty foot (18m) lengths with their ends welded together to create continuous rails of up to a mile or so (1.6 km) in length. Welded track offers not only a smoother ride but also, thanks to the reduced number of joints, requires much less maintenance.
Welded track was originally developed in America in the late 1930's, notably on the Delaware & Hudson Railway in 1937 where it was known as the 'Velvet Track'. In America they tried welding lengths of line up to half a mile long then transporting this on a fleet of wagons to the site, they found the long rails curved easily to follow even quite severe reverse (S shaped) curves. In Britain some long rail sections were moved on long rakes of redundant long wheelbase twin bolster wagons (similar to those offered by Peco), purpose built rail carrying wagons have always been used by the railways, where long lengths of welded track are used these are carried on specially designed bogie wagons. The photos below show such a train comprised of JZA wagons passing through Manchester on a rather dull winters day in 2005.
Fig___ Photograph of a welded rail train
I understand that lengths of up to 230 yards can be supplied from the steelworks these days (thanks to continuous casting methods) but in N the rails would not bend so this is not really an option.
Most welded rail in Britain has been welded on-site using a special train which lifts the track, runs it over a series of wagons where it is welded and ground smooth, then re-lays it behind. Once welded and laid on the sleepers the rail is heated and the securing clips banged into place, as the rail cools it tries to shrink but it is held by the clips to the heavy sleepers. This pre-tensioning of the track eliminates, or at least greatly reduces, the danger of the rails expanding too much in hot weather and buckling. In the mid to late 1990's a new technology called 'flash butt welding' was introduced, this employs a machine able to run on road or rail which can rapidly weld plain rail ends together very quickly. This technology allows shorter rail lengths to be welded and hence allows for track with fewer joints even close to points and crossings.
Fig___ Photograph of a joint used with welded rail
Even with the pre-tensioning allowance has to be made for expansion in hot weather and contraction in the cold. Welded rails therefore have special 'expansion joints'. As the lengths of rail involved are long the cumulative expansion is large so at the joint the ends of welded rail are cut to wedge shape and a special chair is fitted to hold the ends together.
The joints are only about eighteen inches long and quite difficult to spot, the give-away is that two lengths of old bullhead steel rail are bolted across the four sleepers bridging the joint, note these sleepers are wooden even if the remainder of the track uses concrete sleepers. The old rail is only about two-thirds the size of the running rail, but even so it might foul the coupling pin on an N Gauge model. The best option would be the finest rectangular I section Plastruct Fineline (reference BFS-2) and sand the bottom sides almost flat before gluing the two lengths in place. Welded rail has no joints in the old sense of the term and hence eliminates the clickety-clack noises of the older tracks.
The Liverpool & Manchester Railway of 1830 settled on fish-bellied rails some fifteen foot (4.5m) long and each supported in seven cast iron chairs. They experimented with both wooden sleepers and stone blocks but with the development of effective wood preservatives they found the wooden type preferable and this became the norm. Porous wood is used as this absorbs more preservative (the average sleeper soaked up about three gallons of creosote) and wooden sleepers can last up to twenty-five years in service. Where there was a danger of fire, such as on the great timber 'fan' viaducts built by Brunel, the wood was not soaked in creosote but instead treated with a more expensive mercury based liquid.
The LNWR tried iron sleepers but this was not a success and various companies including the LNWR and GWR tried steel sleepers, but again these did not catch on in the UK. Everyone settled on Baltic Redwood timber sleepers, originally nine feet (2.7m) long, reduced to about eight foot six (2.59m) with the shortages caused by World War One. With the advent of the Second World War, and after that the Iron Curtain, Baltic timber became unobtainable, Canadian Douglas Fir was tried but this proved less than ideal and interest focused on concrete sleepers.
The earliest reference to concrete sleepers I have found was on the distinctly different Weston, Clevedon & Portishead Light Railway who used them from about 1909. The sleepers were unconventional in design, rectangular concrete blocks linked by transverse iron rods similar in appearance to the old stone block sleepers. These proved something of a success on the WC&PLR but other companies had little success with concrete sleepers of more conventional design as they tended to crack in service.
The first general introduction of concrete sleepers on main lines took place in the late 1930's. These were of conventional design, direct replacements for the timber type, and were fitted to carry chairs and bull-head rail. Concrete sleepers weigh more than wooden types but last fifty years instead of the average twenty-five for the wooden type, and with mechanical handling (introduced due to wartime manpower shortages) the weight was less of a problem. The problems of the concrete cracking in use have been resolved and concrete sleepers are now standard for plain track but to date wooden sleepers are still used for point work. Modern practice favours the use of hardwood sleepers as these last much longer than the soft wood types. Steel sleepers, imported from the USA, are in place on some limited sections of track in Britain but they are extensively used in Europe, mainly in Switzerland. Steel sleepers are not solid blocks, they are formed rather like an inverted pie plate with a dished under side, this saves weight and cost and helps the sleeper bite into the ballast. They are not ideal however as they lack weight, which is an issue when using pre-tensioned welded rail (discussed below), so they are not suitable for main lines. I have only seen steel sleepers once, stacked by the track side awaiting use. These were 'rust' coloured and had the track fixing clips built-in.
Currently (later 1980s) the railways are replacing about a million sleepers every year, over half of which are concrete. Old wooden railway sleepers had a ready market, used for everything from building heavy fencing to building up flower beds in peoples gardens. Creosote is distilled from coal tar and contains something called benzopyrene, in 2003 new research suggested that even the trace amounts of this material remaining in old sleepers represented a health hazard. In 2003 the European Union sponsored legislation banning the sale of old wooden sleepers. Network Rail is therefore faced with a problem of what to do with them, one option being considered being to build a special furnace that can use chipped sleepers as fuel (examples of which are already operating on the continent).
Railways require a stable base or track bed, after rain the earth softens and the track becomes uneven. The wagonways of the North East were mainly built to carry coal and ore to the ports, the ships arrived filled with a 'ballast' of broken stone and gravel and this was taken and used by the railways to form a path on which the track was laid. This allows rainwater to drain away giving a more stable surface and the term 'ballast' has remained in use ever since to describe the broken stone path along which the railway runs.
In the 1870's there were several derailments caused by rails breaking, the railway companies blamed the quality of the steel used for the rails and suggested it had been weakened by exceptionally cold weather. The British scientist Joule decided to experiment to test this idea. He used nails and needles for the experiments, this was the first time someone conducted such an experiment where the materials used were scaled down and there was much debate about the validity of scaling things down in this way. He found that the steel was actually strengthened by cold weather, when most of the accidents had occurred, and that the problem lay in the design of the track bed. This lead to further experimentation by the railway companies and the development of a standard method for ballasting track.
Fig___ Bullhead track and cross section of track bed
Most lines used crushed stone ballast, limestone and granite were both common. These materials start off white but soon develop a coating of brown or rust coloured brake dust and black sooty deposits from the steam locomotives. Since the introduction of diesel power there has been a steady deposit of oil along the track between the rails darkening the ballast to almost black in stations where trains tend to loiter. From observations on my local line it has taken three years to turn the fresh laid white ballast to a patchy brown.
Other ballast materials were also used, depending upon what was available locally, in Wales broken slate was common in the North and the dark waste from lead mining was common in the South.
Some ballast had high levels of toxic impurities (such as the lead mine waste) which inhibited weed growth, where this was not the case it was (and remains) necessary to treat the track with weed-killer to prevent the plants clogging up the ballast and reducing the drainage. Where track was laid in goods yards, marshalling yards and industrial premises where speeds were slow and uneven track was not such a threat a simple bed of ashes and cinders was laid. The track in these areas was not on a raised earth bank and this is clearly visible when you look at a yard area beside a main line. If you are using cork strip under your track do not use this in yard areas and paint the ballast and surrounding ground black.
On quiet branch lines ash ballast was also sometimes used for the running lines, although this would be laid on a conventional earth bank to assist drainage. Ash ballast tends to allow the track to sink deeper in than broken stone, often it crept up to the sides of the rail and its finer appearance is relevant to the modeller. Fine sand serves for conventional ballast, for ash something rather finer is required, I have had success with Chinchilla sand (from the pet shop). I use this for conventional broken stone ballast as well but it can be smoothed down with a wet finger after the glue has been added, producing a much flatter surface resembling ash and cinders. Ash and cinder ballast was (certainly in sidings remembered from my youth) almost black, presumably when freshly laid the ash would have added shades of lighter greys. When gluing the finer material down using the usual syringe of diluted PVA glue I have found it tends to contract and crack when drying so I add a thin coating of undiluted glue over the top and sprinkle on additional material in the affected areas.
One variation on standard track is 'inset track', where the track is set into the surface of road or yard area. The first use of this kind of track was in 1852 when Frenchman Emile Loubert used it for tram rails in the street.
Inset track became the norm for trams running in city streets but it was not common on railways. A few areas in goods yards might have inset track, and where light railway or industrial siding lines passed along public roads the track was set into the road surface, but their main use was in industrial locations and docks. In dock areas inset track was common but it costs more to lay and is more difficult to maintain so it was avoided where possible. As an example in the Trafford Park industrial estate in South Manchester the rail lines were all conventionally ballasted track beside the roadways, only being inset where they crossed roads.
Inset track could be laid using either pairs of standard rails (with one laid as an inner 'check rail') or specially rolled 'tramway' rail as shown in the drawing below. Note the 'tram' rail is actually a single rail with a groove in the top for the flange of the wheel. This flat bottomed rail is secured directly to the sleepers with steel spikes. Tram rails were not suitable for railway rolling stock as the groove was too shallow (the flange on tram wheels is much smaller than on a railway wheel). Specially produced inset rail suitable for railway wagons was used in some locations but as the railways had a plentiful supply of standard bull-head rail requiring only double versions of standard chairs the twin rail arrangement was common.
Fig___ Inset track
The Pandrol clip
tightens under vibration, which prevents the problem of trains
dragging the track along when the brakes are applied. Pandrol clips
save a lot of time and effort, the clips can be hammered in by a man
with a big hammer or by a rail-mounted machine, they are now in use
in over sixty countries. Pandrol clips appeared in the UK in 1959 and
they are now the standard rail fixing throughout Europe and in many
other parts of the world. The latest (late 1990's) designs of Pandrol clip use a
Where the lines handled normal railway traffic the surrounding road or yard surface was built-up to the height of the rails and the gap between the inner rails would then be filled using stone setts which could be removed for track maintenance as required. The surrounding area was generally concreted after about 1920, and after about 1930 a top coating of tarmac was often added. From the later 1930's, particularly on dockside lines, the ground was excavated, the track was laid, then concrete was put down to build up the level. The disadvantage of concreting in the track in this way was that maintenance became a lot more difficult and in goods yards it was not uncommon to have the road surfacing built up to either side of the track but with the sleepers exposed between the rails. This open kind of track is much easier to model in N, especially where the lines are curved, and short inserts can be provided where road traffic might be expected to cross the line. The illustration below, based on a photo taken in the 1930s, shows some inset track in a goods yard, note that the point switch rails are cut and hinged, the excavated area shows how long the hinged sections were. The point lever sits in a stone block, so it will not be damaged by passing road vehicles or trip up horses, in use the lever is lifted slightly, dragged round through 180 degrees, then folded back over and into the slot. The metal drum in the rear left is a 'capstan', discussed further in the section on turntables points and slips. By the 1970s the in-fill between the rails was no longer there (the stone road surface outside the rails was still in place) and more conventional points were in use in this yard although the capstans remained in place until closure in 1975.
Fig___ Inset track in a goods yard, note the design of the point blades
Inner 'check rails' were also fitted to normal track where curves were tight as they helped prevent derailments. They were first used as far back as 1729 on a horse operated wagonway in County Durham and they were used on the steam hauled railways wherever the curves were less than about six hundred foot (183m) radius. That is equivalent to over four foot radius in N, or 1.2m, which gives some idea of how tight model railway curves tend to be. Check rails are still seen today on tight curves, note that the check rails are only laid with the inner rail on the curve.
Abroad many railways settled on flat bottomed rails (Vignoles section rail) spiked directly to wooden sleepers, this meant that the heavy cast iron chairs were not required which saved on transport costs. In the event however it was found that it was necessary to add a steel base plate to the wooden sleepers to carry the rail.
In 1949 British Railways changed to using flat-bottomed Vignoles section rails. The rails were spiked to wooden sleepers or bolted down to concrete sleepers with special clamps but in 1957 a Norwegian engineer invented the Pandrol clip. This uses a simple flat plate with cast-in loops, the Vignoles section (flat bottomed) rail is laid in the chair and a twist of metal, the clip itself, is hammered into the loops to hold the rail in place.
Fig___ Flat bottomed rail and Pandrol Clips