Richard T

I was not satisfied with the triangle behind the outer loop track: it was too obviously formed of flat layers. I therefore added about half a kilogramme of lightly-coloured plaster to soften the contours. I also covered the final surfaces between the platforms with a thin coat of plaster. In this step I want to get the ground covered and coloured, not in its final state necessarily but “natural”-looking. This is the remodelled triangle; hard to see the contours here. The passenger platform. The black plinth is for the shelter, and is a 6' × 12' area formed of 5 mm foam board. The platform will be covered in fine gravel. Another view of the platform. I still have to form the stone wall behind the platform, which will be made from foam board. The area between the goods siding platform and the passenger platform. I am still not sure how best to complete the rear of this area; the gap between the ground level and the base of the backscene (my fault) needs to be concealed, however here the ground would be open to the rear. The stonemason siding; I have coloured the new groundwork and also the platform sides, all of which makes this look far more solid and homogenous, even without grass, weeds, bushes and buildings. The goods siding and platform; whilst there is still a lot to do, this is also now looking fairly solid and less like foam board... Preparing for grassing. To avoid the paint showing through I decided to cover the surface first with a layer of earth base flock. I sprayed further scenic glue onto the earth-coloured flock and applied 10 and 12 mm static grass with a Flockit applicator. Contrast between the grassed area in the foreground and the bare earth behind. Here I have applied the earth flock base onto the painted plaster, right up to the inaccessible far corner; the dark green is deliberate, to bring out wetter earth. I've grassed the strip in front of the retaining wall. This picture shows the grass and flock comparison at the rear. Close-up of the Allt Creag na Còsaig, formed using Woodlands Scenics water. I have now added grass to the right-hand bank, and following this picture added a final additional millimetre of water. The passenger platform, covered in buff-coloured stone chips. At this point I have not yet hoovered up the loose chips. Having now completed the basic ground covering I decided it was time to remove the foils protecting the backscene from stray spray. Although the scenery is far from complete, nonetheless I could not resist a series of overview shots of Clachbeg. The site of the mine in the foreground, while in the background the branch line descends through the cutting to the station. The blue colour of the stream is simply the fresh resin water; it will dry clear overnight. The stream and the entrance turnout behind a sward of grass. The station loop with the stonemason siding and platform behind it. The stonemason shed will fill the corner of the layout and disguise the grass bank behind it. To the right of the shed there will be a stone wall (tipped forward here). The goods siding and platform, as well as the turntable. The platform edge still has to be completed. A stone-built goods shed will fill this corner of the layout. View from the corner of the layout. Engineer’s view from the turntable spur. View of the station from beside the turntable. The stone wall beyond the passenger shelter is taking shape; I still have to decide how to fill the gap between the passenger shelter and the goods platform, on the right. Shunter’s view of the station from beside the goods siding. Engineer’s view from the mine spur up towards the station entrance switchstand. Engineer’s view descending the branch line into the station. Engineer’s view on the outer track of the station loop, approaching the passenger siding. Engineer’s view from the stonemason spur looking down to the goods siding. The slightly curved edge of the passenger platform is apparent in this perspective. Engineer’s view at the crossover to the (misaligned!) turntable. Again, the curved edge of the passenger platform is clear. The rails still have a slight sheen to them, even after a first wash of diluted black India ink. With this the essential ground cover is largely complete. The area beside the goods platform is still to be completed, as is the space between the stonemason platform and the passenger platform. Clearly there is still a lot of vegetation to come, but this is nonetheless a milestone: all basic ground cover complete, and finally the railway looks like it is part of a landscape.

Finally the groundwork and platforms are in place, and I can proceed with ballasting the track. The ballast is a mixture of excellent scale granite chips and fish tank gravel; the latter is rather colourful, but it will all be coloured dark grey so I hope that the two will blend. I inadvertently mixed a handful of larger fish tank gravel into the mix, so there are some 3"–4" pieces here and there. Altogether I have 8 kg of gravel to hand... The method I am using is similar to smaller scales: spread the gravel between the sleepers, tidy with a brush (and tweezers) and then spray with a diluted PVA mix, as already tested in the cutting corner. This is the start of the new ballast; coloured and glued ballast to the left demonstrates the difference. I carefully poured gravel between the rails and along the outside edge. This is a peculiarly satisfying task which cannot be rushed. A good start: loosely-brushed ballast along the station loop, the mine spur turnout and the stonemason siding. The ballast here is granite to the left, changing to fish tank gravel towards the right; even without colouring it is looking pretty good. Another view of the loop and stonemason siding. The siding track will be overgrown, but I want to fill earth into the gravel and then grass it over, rather than omit the gravel altogether. The fish-tank gravel has rather more light-coloured stone than I would have liked: I would have been better off staying with the granite chips. Nonetheless, it is commonplace in the Highlands to use river stone where possible. I decided to mix some smaller stone into the station ballast, prior to later grassing some of the track. Here I have started mixing some Woodlands Scenics medium and coarse gravel along the track: dark brown, medium brown and cinders. I started with the goods siding; the turntable pit stands out here as rather bare. The stonemason siding is buried deeper than the rest of the station in fine gravel. The coloured gravels contrast strongly with the white river stone; this will be toned down later. I also added fine gravel to the mine spur; the areas to the sides of the track will be grassed. The dry ballast complete along the station. The dry ballast complete at the station entrance. The track in front of the still-incomplete passenger platform. I am quite pleased with the mix of finer gravel to represent older ballast which has sunk into the earth a little. Detail of the goods shed end of the station. I used medium cinders and brown gravel to cover the wooden base of the turntable pit; weeds will come later. I sprayed the ballast with Woodlands Scenics glue, to which I had added black and burnt sienna ink, to tone down the stone and also to give the sleepers a more “creosoted” look. The effect is instant and pleasing. Before spraying the glue onto the ballast I oiled the tie plates of the turnouts, as well as the planed faces of the points themselves, and the stretcher bars, to prevent the glue from fixing them in place. After applying the glue I carefully cleaned up the moving parts using cotton buds; the mechanisms do not appear to have suffered, although a couple are a little stiff. Once everything is completely dry I shall clean the affected areas using a fibreglass brush to restore easy movement. The station loop once the ballast has been toned down and glued. The glue is still not completely dry, hence the sheen on the rails; if this remains I shall finish with a spray of matt medium. The station entrance after glueing and toning the ballast. It’s coming together.

Clachbeg station boasts three platforms in all: a 16' stone platform on the stonemason siding, a 27' low gravel platform for passengers, and a 31' stone platform beside the goods siding. The stonemason and goods sidings are 18" high (from the rail tops) and encroach 3" into the loading gauge clearway; the passenger siding is a simple gravel affair at rail height, bounded by sleepers on edge. I decided to lay these out before ballasting, as they form part of the ground cover and ballast border. I had a very few leftover pieces of 25 mm foam board, which proved ideal for the higher platforms: two layers built up from the foam board base conveniently ended up 18" above the rail tops. The scraps would result in a few gaps which I would plaster over. First step was to cut pieces from the scrap to make up the platforms: the simple foam cutter has proven its worth many times over. The goods platform test-fit; the rear right-hand corner is rounded to accommodate the backscene, and the shed will be an interesting project as it too will follow a similar shape. The stonemason siding, loose fit. This is awkward to reach in the far corner of the layout, although the siding itself is easy to reach. Finally the unsightly gap between the bottom of the backscene and the baseboard surface is beginning to be closed (a mistake I made early on...), and I thought I would tackle the gap at the end of the station whilst I was at it... This was very satisfying: freehand drawing with a blunt, soft pencil produced this wall in about ten minutes, in the same style as the retaining wall by the entrance turnout. It neatly finishes the station and should bridge the transition to the backscene. It stands slightly proud of the backscene itself and is affixed both with Polyfilla and white glue. The two buffer stops were fun to make. On the right is a simple bent-rail affair with a crossmember welded in place; the track beyond the turntable will be buried in grass. The wooden buffer stop at the end of the goods siding is based upon prototype photographs and is bolted together with 1" bolts (M2 threaded rod and nuts). Whilst in the mood for detailing I also added shunting marker posts from scraps of rail—these show the closest point to the turnout where wagons may stand without hindering the adjoining track. The tips of these rail ends will later be painted white. The sleepers laid out for the passenger platform. A scrap of 3 mm ply forms the base: not really necessary but it provided an even base for the sleeper edging. A couple of the sleepers had to be cut to interlock with the “real” sleepers at the ends of the curved platform front. The platform is designed to serve up to three passenger coaches along the outer loop of the station, and will feature a small, wooden shelter. Making a start with filling the passenger platform. The other end of the passenger platform. The stonemason platform with smooth plaster edging. I shall scribe stonework in situ, and also fill the gap to the rear. Another stone wall is planned to the right, to fill that unsightly gap. The goods platform plastered in and ready for scribing and surfacing.

Having completed the building of the trackwork, and glued all the templates to the baseboard, it was time to bed the templates in to the scenery. There are several things to do: painting of the rails, building up of the ground layer around the track, and finally ballasting the track. Clachbeg is in the far north west of the Scottish Highlands, so rails would have very quickly acquired a rusty patina, and the traffic was hardly intense enough to coat the rails with the sort of grime which we are used to from modern railways. I had tried painting the rails using a small brush, but was not happy with the result. The larger rails were somewhat grimy (I suppose I could have cleaned them before building the track; bit late now...) and the lighter rails were shiny. I also tried rail weathering solutions, but was far from convinced: they merely resulted in matt and blotchy rails; perhaps they are better suited to outdoor rails (or freshly-cleaned ones). Finally I decided to simply spray-paint the rails, then lay the ballast, then colour ballast and sleepers (and rails) with a dirty wash of scenic glue when I secure the ballast. That ought to tone down the rust and blend the colours. I used small cans of HMG “Dark Rust” from Howes Models; not sure how much I needed, I ordered four cans, and in the end used two. Having settled on a plan, next thing was to find time to tackle it... The entrance turnout. The rail to the left has been painted by hand with a paint which is supposed to rust naturally; after a week it remained grey, but I was unhappy with the quality of the brushwork rather than impatient with the natural rust. This is the tedious part: I used 2"-wide masking tape along the outsides of the rails, and ½"-wide tape between the rails to mask the sleepers and—in the corner where I have already started the scenery—the grass and existing ballast. There will be some overspray, but I am not too concerned about that, as long as I can avoid intense colour or hard lines. Taping the turnouts is quite fiddly; this was the first, and I rushed it, so shall have to return with a paintbrush to touch up some places. Masking the station loop. I had tried painting the rails of the mine spur using a rust-coloured paint pen by Woodlands Scenics: easy enough (if rather pricey), but I was not happy with the outcome: the shiny Code 148 rail shone through the paint even after a second coat was applied, and the pen was not suited to coating the complications of the fishplates and their bolt-heads and nuts properly. The light rail of the mine spur, and the branch line behind. I was a little worried about overspray, but the spray cans proved pretty accurate. The mine spur. A first impression: I was pleased with this. This was my first attempt, and it went fairly well. There was some overspray onto the sleepers where the tape had gone around spike heads, but I don't mind some rust stain on the sleepers. I should have protected the wall better, but the overspray there will be easily sorted. The bridge and culvert. Some overspray caught the timbering; I took a brush and painted some paint thinner onto the sleepers, where it was drawn into the grain and rather effectively drew out the rust colour and seeped it into the wood. I’ll tidy this further once I’ve applied the final wash of dirt. At the moment the rust is rather bright against the scenery colours. I needed a break from the fumes of the spray paint, so I decided to build up some of the terrain. I mixed up Artex powder with water and remains of tubes of watercolour paints, and black ink. The colours don’t make much sense, but I wanted anything earthy rather than white, so that should anything chip it won’t look unsightly. I applied about 500g of plaster at a time, using a small sculptor’s trowel, and then brushed it smooth with a water-laden paintbrush. I am building up the ground around the tracks to about the tops of the sleepers; I’ll fill up with ballast to the same level, and then blend earth and ballast in some places. I want to recreate a reasonably well-maintained look to the “main line” track, whereas the sidings will be less well-ballasted; grass will complete the effect I am looking for later. The turntable benefits from being buried in the landscape. At this point I have completed the ground between the front of the layout and the tracks, as well as between the tracks. Before working behind the rear track I shall form the platforms from foam board, then bed them in with plaster all around them. The buildings will later sit on the platform bases. The station loop with the stonemason spur in the background. The foam board will be cut to fit the platforms. This rail is still unpainted in the foreground: a part of me yearns to keep it that way... Having cleared the air from the first painting session it was time to finish the job. An hour of masking sleepers, then five minutes of spraying. Here we see the turntable turnout. The crossover turnouts. I sprayed each rail from the side, both front and back, as well as longitudinally at the turnouts to colour the rail inside the check rails as well as the sides of the switch rails. Overspray is not an issue here, as this will all be coloured as the scenery progresses. The turntable pit and buffer stops. I protected the backscene here with adhesive foil meant for protecting carpets when decorating: a good idea, and I am sure it will work, but applying it single-handed was very tricky; I envisaged becoming wrapped in it and unable to free myself... The smudgy marks are from my fingers as I unrolled the foil and stuck it against the backscene. Stonemason spur. With this the rail rusting exercise was by and large completed, and despite keeping the garage door wide open all the time I was light-headed from the spray paint smell. Time to take a break! The same scene once the paint had dried. It only took a few minutes to peel away the masking tape... The track does look much more realistic now... I spent some time with Ballistol oil cleaning paint from the tie-plates and ensuring the switch rails move smoothly. I also touched up any shiny metal which remained. I’ll tidy the sleepers once the ground cover and ballasting are finished. With that, the rails are done, so I can proceed with ground cover and ballasting.

My chosen combination, 7/8"n2 (7/8":1' or 1:13.7 scale model of 2' gauge), not only has heft and presence, but allows for an accurate representation of 2' gauge using commercially available wheel sets for 45 mm gauge (in common with Gauge 1, IIm, G and Fn3), and scale rail sizes to represent very light rails of 12 lbs/yd and 25 lbs/yd. [Strictly the gauge is 1 ¾", or 44.45 mm, however in practice wheel sets and mechanisms are interchangeable with 45 mm gauge.] Occasionally 7/8"n2 is referred to as “M”—the only notable occurrence of the 2' gauge in North America was in Maine, hence the “M”; this has not caught on outside the USA. What did catch on was the US-style nomenclature of inches-to-the-foot. Modellers of 600 mm gauge prototypes on the Continent sometimes use 7/8"n2, while others model at 1:13.6 or 1:13.3 scale for better accuracy using the same 45 mm gauge; the difference is imperceptible to the naked eye. I did also consider two other combinations, at opposite ends of the spectrum which interests me: ½"n30 (½":1' or 1:24 scale models of 2'6" gauge prototypes on 32 mm gauge), which would be pleasing and allow modelling of a number of credible prototypes; 1"n15 (1":1' or 1:12 scale models of 15" gauge prototypes on 32 mm gauge) to model a classic minimum-gauge estate railway. Both of these doll’s house scales allow use of commercially available wheel sets and power units, as well as commercially available models and figures. The attached table is useful in untangling the plethora of scales and gauges; I have attempted to name all those I have encountered, excluding model engineering gauges larger than 1:12.

Great addition, thanks! Updating the list.

So to the final step in the tracklaying for Clachbeg—I can distinctly remember not all that long ago when I did not think I would ever reach this point! I have been lucky enough to have had some time off, and have been determined to get to this point before returning to work; laying the trackwork is an extended undertaking, whereas many other tasks are more discreet in nature. The station turntable serves to turn incoming locos before they embark upon the six-mile return trip to Mains, and also to turn the passenger brake such that the brakesman’s compartment is to the rear of the train. It measures 10'4" across the bridge, ample for the short wheelbase rolling stock it will be used for. The model is a kit from Kitwood Hill Models designed for G scale and the kit is designed for US practice, however is close enough to the original to serve well at Clachbeg. The kit is made of plywood, but the finished model will represent a plate steel bridge set in a concrete pit with boarded sides. This shows the first stages of the kit completed: the baseplate with the ring rail soldered to PCB sleeperettes and the framing of the pit wall connecting to the outer rim. All laser-cut and an excellent snug fit; no filing or fettling required. I sprayed the base quickly to get rid of the fresh timber appearance, and the transformation was very effective. This is a single coat of Chaos Black. The turning plate upon which the bridge will sit is unpainted. The bridge assembly is straightforward and satisfying, the parts fitting square and firmly. I glued it together with cyanoacrylate in places and with white glue in places. Here the wheel frames are being glued to the completed bridge frame (which I had stained before assembly). The supplied motor and gearbox. This unit requires quite a deep hole in the baseboard. The motor runs very smoothly and the gearing results in a slow movement, however I found that the turntable plate had to be aligned very precisely for the motor to be able to turn it... a quick test with a weighted bridge confirmed that the motor is not robust enough to cope with this scale. Given the location of the turntable right at the front of the layout this is not a problem, so I decided simply to leave the motor aside and to operate the turntable by hand, as the prototype was operated. Measuring the baseboard before cutting the pit. One cannot have enough rulers... The outer rim of the turntable encroaches on the loading gauge of the goods siding, however the 8' way between track centres allows for the turning of 9'-long rolling stock even if the goods siding is occupied by a full-width wagon. The loco used from 1933 onwards, a Plymouth 3-ton switcher, measures 8'7" long. Of course, in case of a longer longer loco, the goods siding would simply have to be cleared first... Cutting out the pit. The white glue used between the foam board sheets proved remarkably tenacious; on the other hand the strata of the boards provided a flat and level base. Thanks more to serendipity than planning, the tops of the sleepers lined up with the rail surface of the turntable bridge without any further ado. Another view which shows how this final piece of the trackwork literally fills the last available corner of the layout. Checking the length of the approach rails. I have painted the turntable bridge with a grey rust-effect paint to represent plate metal; the paint should slowly rust. By contrast the timbers will remain stained wood. The rails are my usual 25lb/yd, fitted to the plastic chairs supplied with the kit. The chairs (actually representations of clips bolted to tieplates) are designed for Code 250 “flat-bottomed” rail whose foot is in fact no wider than its head; with minimal opening of the aperture I was able to slide the Code 205 rail into the “chairs”. They raised the rail minimally, which I was able to compensate for by sanding the foam board base gently with coarse sandpaper, before pressing the turntable well back into the pit. Glueing the approach template and turntable in place, the weights ensuring rail top alignment. Initial inspection by the Works Engineer. The turntable rails have been painted with a rust colour, but I am not happy with it, it looks too flat and solid. The approach rails have yet to be spiked and fished. After some research I fabricated a simple buffer stop made of three lengths of rail. In practice the stub of rail beyond the turntable was largely buried and never used. The approach rail fishplates lie ready for installation. Another view of the turntable with its buffer stop. The stub rails need to be nudged closer to the turntable to close that gap, even though no wheels should ever cross it. Final approval by the Works Engineer! He appears not to have noticed the fishplates still lying awaiting installation (so had I until I saw the photograph). Apart from two other buffer stops, the trackwork of Clachbeg is now complete (although I shall have to paint the rails and weather the sleepers and ballast it all, of course...).

This template is the largest of them all, and includes the two nearly identical, straight, right-handed turnouts which close the station loop and lead to the goods siding and to the headshunt respectively. Unlike the other turnouts these have straight 1:3 crossing vees, which should make construction rather more straightforward. The template includes one panel of inner loop track, as well as one-and-a-half panels of 9'-long 12 lb/yd tracks, being the start of the goods siding. The platform-side turnout—facing the stonemason turnout—will have a switchstand with indicator located inside the loop slaved to a locking switchstand at the layout edge, just as the stonemason turnout does. Unlike the other turnouts these have timbering perpendicular to their straight roads. The maximum sleeper length is 8', just enough to accommodate this formation. The template hardly fits onto the workbench, it is so big. The sleepers for the goods siding are already glued in place at the rear, raised ¾" (1.5 mm) on plastic strips so that the light rail will align with the heavier rail of the turnout. The scribbles and bays along the front of the template are where I had thought of mounting a central lever frame and point rodding. A comparison of the “real” template with the original from Templot reveals that I have relocated the rail joints, to avoid having them conflict with the check rails; this was the case with all five turnouts. The progress of building the turnouts is much the same as for the preceding ones: cut sleepers to length (about ten) distress sleeper tops with a wire brush stain sleepers cut and bend rails to fit template: main stock rail—toe end main stock rail—nose end diverging stock rail—toe end diverging stock rail—nose end main crossing rail diverging crossing rail main wing/closure rail diverging wing/closure rail main switch rail diverging switch rail main check rail diverging check rail [*]file or mill: crossing vee rails switch rails toe-end stock rails [*]drill switch rails for stretcher bar [*]drill all rail ends for fishplates [*]fabricate fishplates for loose heels and for stock rails (four pairs) [*]fabricate tie plates (six) [*]check all rails for fit [*]solder crossing vee together [*]fabricate stretcher bar [*]assemble stretcher bar to switch rails [*]loosely fish switch rails to closure rails [*]glue sleepers to template [*]spike crossing vee [*]spike nose-end stock rails [*]spike wing/closure rails including tie plates [*]spike toe-end stock rails [*]fish stock rails together [*]spike check rails [*]fit switchstand with rodding [*]fabricate fishplates for joints to other templates (six pairs) [*]glue template to baseboard [*]weight template down until glue has dried [*]fish rails to other templates’ rails [*]tin all fishplate bolts [*]trim all fishplate bolts This shows the tracklaying progress thus far; the big blue gap is where this template comes from. Sleepers cut; lots of 8"-wide timber here. Sleepers stained Milling a crossing vee. I am feeling much more confident after a little practice and really prefer this to filing by hand. I scribed the rail carefully from the template to show what was to be removed, and set up the angle by eye: easier than I had thought. The result of the milling: a beautiful fit which solders together easily. The only difference between these two turnouts is that one has the crossing nose on the straight closure rail, the other has it on the curved closure rail. Milling a switch rail. The two turnouts have the same geometry, so I made all four 5' switch rails in one sitting. The tricky bit is setting up the machine vice at the required angle (using an engineer's square and brass spacers to set the angle before clamping the vice to the milling table). The setup is crude but surprisingly accurate and the result is ideal. Checking all rails. These straight-crossing turnouts were particularly straightforward to prepare and build: only two curved rail lengths each. The plain track in the foreground is the start of the inner loop, while the lightweight rail at the rear is the start of the goods siding; its sleepers are elevated above the baseboard to compensate for the difference in rail heights. The template had been cut on the basis of 15' rails, so when I realised that the sidings were to be formed of 9' lengths of the lighter rail I ended up with having two rails straddling templates. Fishplates for the template. I make these in batches as and when required. The sleepers are all glued in place, and here I am ensuring that the long ones dry flat against the template (none have warped but at this point I was not taking chances). This time I remembered to assemble the wing rail–closure rail–switch rail–stretcher bar combination before anything else; the reason for this is access to the loose-heel fishplates and the stretcher bar nuts underneath the rails, which is not possible once the rails are in place between stock rails. Same for the other turnout. This is a good time to double-check that the stock rail planing still matches the switch rail planing, as the loose heel adds a scale ½" or so to the length. The completed crossover, spiked and fished, seen from the reverse side this time: the great advantage of building the trackwork at the bench. After glueing the template onto the baseboard and fishing the rails to the other templates I finally had a complete loop—indeed, apart from the station turntable, the trackwork is complete. Here the view from the goods siding to the stonemason siding, with the station loop and the branch line curving to the left and starting to climb past Creag na Còsaig. One length of light rail of the goods siding has still to be spiked. The two slaved loop switchstands can be seen here (well, one behind the other) as well as the two locking stands at the layout edge. View from the turntable turnout. From this perspective the 25 lb/yd (code 205) rails look quite massive (certainly when compared to the 12 lb/yd (code 143) siding rails), however they are not only prototypical but somewhat smaller than the code 250 (40 lb/yd) or even code 332 (70 lb/yd) rails frequently seen in trackwork in this scale. The loop switchstands. The indicator disks clear the loading gauge by an inch or so (scale). The locking stands in the foreground are connected by lengths of florist’s wire running in brass tube to the indicator stands. Change of rail weight. The rail tops are “out” by ½" or so despite the asymmetric fishplates, however the effect on a wagon rolling over the joint is barely perceptible. The completed passing loop at Clachbeg. Close-up of the crossover itself; the straight line extends from just beyond the tip of each frog and of course avoids any nasty s-curve. Not really part of the crossover or loop, but as the trackwork is nearly complete I took a few views from various points. This is looking up the mine spur to the station; to the left the branch line gradient descends to the guard rails of the mine turnout. View down the branch line towards the station, just as the station loop begins. Finally, the view from the stonemason siding looking towards the goods siding on the left. The only remaining trackwork will be the station turntable in the far right corner. The cramped location of the two loop switchstand indicators is clear in this shot—as is the slight and unintended kink in the goods siding.

The stonemason siding is just 18' long, straight, and is made of 9' lengths of 12lb/yd rail (salvaged from the original mine tramway). It will have a wooden buffer stop, and be faced by the stone platform of the stonemason shed. It holds two short wagons. The stonemason turnout has the curving road as the main line and the straight road leads to the siding. The curve radius is 33', and the turnout is curved through the 1:3 crossing vee. As with all the CMER turnouts it has loose-heeled switch rails. The turnout is made of 25lb/yd rail. The turnout at Clachbeg was controlled using the usual switchstand with indicator disk. The turnout also forms part of the tiny, two-coach passenger platform, and therefore the turnout control was situated inside the passing loop rather than on the platform side. On the model this location could be hard to reach if high wagons occupied the inside loop, so the original switchstand will be controlled remotely from a repeater switchstand—without disk—at the layout’s edge. I started off with the easy part: the siding. Just six sleepers and four 9' rails, all on one template. The 12lb/yd rail (Code 143) seems impossibly delicate after working with the heavier rail section. The template of the stonemason turnout, some 25' long. Checking that the timbers have been cut correctly and that the switchstand timbers are long enough. As can be seen from the scribbling on the template I have “shoved” timbers to achieve the right gap for the switchstand and then to fill the gap that created. Aerial close-up of the crossing vee and wingrails, which went together without drama. I filed notches in the foot of each wingrail at the knuckle to facilitate a sharp bend. Checking the fit of the wingrails. Milling the stock rails. It all looks rather crude in this unforgiving close-up but ended up somewhat smoother. This was a first attempt with my Proxxon set-up; with a little care and patience it produces very acceptable, fine control, and the outcome is so much better than manual filing. Checking the fit of switch rail and stock rail. Last check of all the rails on the template before adding the sleepers. This is important as, once the sleepers have been laid, parallax makes checking rail fit rather hit-and-miss. The scorch marks tell the tale of soldering the crossing vee with a gas-powered soldering iron. I started by laying the crossing vee and then used gauges to work everything from that. The little engineer’s square is used to avoid parallax errors; usefully the crossing nose peeps over the sleeper by just a smidgin and is thus easy to locate. Making up the fishplates for the stock rails and to fish the closure rails to the switch rails. Each fishplate is scribed four times to mark out the drilling spacing. Drilling spacing crib sheet—by now the numbers are etched into my memory! Feathering the edges of a tieplate. These are snipped from 8 thou (0.2 mm) tinplate and measure about a scale 6" × 9". The six tieplates; I punch the holes with a bradawl, which dimples them before puncturing; I use hammer and anvil to flatten them again and then run a 1 mm drill bit through the holes to finish off. Following the crossing vee, the straight stock rails were located using gauges, and then the curved stock rails, and then the wingrails. I check the alignment of the rails through the knuckle with a short straightedge. The switch rails. I had not learned my lesson from last time, and faced two insurmountable problems here: how to bolt the loose-heel fishplates in the cramped space, and how to fit the stretcher bar to the switch rails... hmmm... In the end I pulled the spikes of the wing rails and the stock rails so I could connect the stretcher bar to the switch rails, then place the switch rail assembly in place with the stretcher bar underneath the stock rails, then fish the switch rails to the wing rails, then spike the wing rails down, then spike the stock rails back in place. Hopefully I’ve learnt the lesson now! Marking the holes for the switchstand bolts. The completed turnout including the switchstand and linkage to the stretcher bar. The locking handle has been removed from the switch stand, as this will be slaved to another stand (of which more later). Test-fit on the baseboard. The gap in the foam board is of no importance. This was also the moment to check that the siding template fitted as planned. The transition to the lighter rail of the siding. This time I remembered to compensate, using 1.5 mm plastic sheet under the siding template to lift it enough that the rail heads meet. The two roads and their joints; in this view the raising of the siding template is visible. Everything lined up perfectly. I was pleased to have reached this stage. Still not glued or fished. Another view of the templates, still not connected but nicely aligned. Glue for the turnout template. After placing the template very carefully (it initially swims on the glue and it is important to check rail alignment) I weighted it down firmly and left it for an hour. As I make progress in laying the track I shall have to store my scenic materials elsewhere... The stonemason siding. The dimensions of the platform and stonemason’s shed are marked onto the baseboard. The buffer will follow too. Another view of the siding. The far corner of the layout is hard to reach, and will be filled by the stonemason’s shed, nonetheless I shall have to add some scenery there. The vertical surface behind the buffer stop will be a stone retaining wall. The siding glued in place. The turnout control: in the foreground is the stand without indicator, connected to the indicator stand by a length of 0.5 mm florist’s wire in a brass tube. Aerial view of the control layout. The tube will disappear under ballast. The templates are cut to the loading gauge, so there is not much room available. I had considered using point rodding, as can be seen sketched onto the templates, but decided that this would be just too complex, as I would have had to include working compensators for expansion and contraction... maybe next time! Close-up of the foot of the indicator stand: the mechanics are as simple as can be. The foot of the switchstand at the layout edge. It is still unattached to the mounting blocks. Once I was satisfied that all the measurements were correct and the switch rails were being correctly thrown, I glued the tube into place. A wider view of the switchstand inside the loop and its control at the layout edge. The fitted turnout and siding. I still have to tin the fishplate bolt threads and then trim the bolts to length. A view up the branch line and down the mine spur; the tracklaying is really bringing the layout to life. A final view back up the branch line; this shows the deviating gradients nicely. Next up is the pair of crossover turnouts leading to the goods siding and the headshunt...

The template with the entrance and mine turnouts had been glued in place onto the baseboard but was still not fished together with the existing track, so this was a satisfying task to tackle. At the same time I decided to place the bridge in place definitively, and lay the loop track sections, so altogether I needed to fabricate 64 fishplates and then use 128 nuts and bolts to fish the various rails together. This is the joint from the branch line onto the bridge. This shows the bridge rails on the left, a six-foot connection and then the entrance turnout on the right. A close-up of the joint between the mine spur’s 12lb/yd rails on the left and the turnout’s 25lb/yd rails on the right: special fishplates will be needed to line these up. The problem is exacerbated here because I had not thought to elevate the last portion of the trackbed of the mine turnout to compensate for the ¾" (57 thou or 1.5 mm) difference in the rail heights. The connection to the mine had not lined up very well, so some miniature crowbar effort would be needed to align the rails. Several of the fishplates were prepared with elongated holes to accommodate differing gaps between the rails of the adjacent templates. When fishing together two rails on the same template this can all be adjusted on the workbench and the gaps cut to fit; despite careful measurement the same accuracy is not possible when fitting templates to the baseboard. The fishplates for the 12lb–25lb joints were made from 2.5 mm bar and channel respectively. I drilled two holes closer to one edge and then filed off 0.5 mm to fit the lighter rail. Fiddly but not nearly as time-consuming as I had feared. The joint to the mine spur, completed. Getting the rails aligned and the tiny bolts through their holes required a little dexterity. In this shot I have already trimmed off the tinned threads of the bolts. The rail tops do not align perfectly; later I filed the bump into a short ramp. Close-up of the transition fishplate. Close-up of the rail joint; I was pleased with this, although I shall try next time to shape the fishplate better to align the rail tops. The filed “ramps” on the heavier rail can be seen here: crude but effective. A wagon crossing this joint clacks but rolls smoothly enough. The other joints were straightforward, and I proceeded with three very easy sections of track: three curved rail lengths of the station loop, each built on its own template. This is a part of the outer loop, very straightforward and not too sharply curved. After glueing the template to the baseboard I weighted the track down to ensure vertical alignment with the neighbouring template: the thin ply boards under the weights span the rail joints. Despite my best efforts with the multiple layers of foam board the surface is not perfectly plane, as the underlying shelving inherited some of the unevenness of the garage floor. I am not at all unhappy with the result (a couple of minor undulations of a few millimetres) as it gives the track a realistic look; geometric perfection does not go well with light railway budgets! The outer loop rails fished together. The inner loop rails receiving the same levelling treatment. At this point the easy interlude is over: next up will be the three station turnouts: stonemason spur, and the crossover.

After completing the station entrance turnout, next up is the extremely sharp turnout to the mine spur. The mine spur actually pre-dates the station: it is the last remaining section of the ½-mile standalone track constructed of portable rail panels when the mine was first opened, along which a Highland pony named Blackie drew a single skip to carry the high-grade ore to an old iron furnace for smelting. That endeavour was unsuccessful and most of the track was taken up and reused, but the last few panels were left in situ and connected to the new branch line at Clachbeg. (Re-cycled 12 lb-yd rails can be found on the stonemason and goods spurs as well as on the tiny turntable at Clachbeg.) The minimum radius for the Clachbeg branch is half a chain, however the 1:3 curved left-hand mine spur turnout uses that for the outer road, and the inner, diverging road is laid to a radius of just 12'3", which was the ruling radius of the portable track system. The mine spur has an axle-load limit of 1½ tonnes and only lightweight internal-combusion locomotives may traverse it with a single skip; it sees a couple of traffic movements a week. The turnout is on the same template as the station entrance turnout, and is constructed in the same manner. The main challenge is the tightness of the curves and thus the dramatic angles involved. This shows the marking-out of the inner stock rail. The full set of rails laid out on the template. The inner loop is at 33' radius, the turnout at 12'3" radius... The main road is straight through the crossing vee, and the switch rails are straight; the diverging road is curved through the crossing vee. By eye, against the template, everything looks ok to within half a millimetre. Close-up of the central ironwork. The crossing vee is not very pretty but works fine, and will be blackened and rusted. A new approach to the inner fishplates using brass channel: like the original this holds the square-headed bolts from turning, making the fitting of the nuts easier. Beside the inside fishplate is an outside one for comparison. The inside fishplate measures 30 mm and is made of 2.5 × 1 mm channel with an inside width of 1.7 mm. Slightly blurry closeup of the fitted channel fishplate: one bolt turned a little but held. Pleased with this! Double-checking with lots of track gauges: even in this scale things can get cramped. The completed turnout. The tight radius of the diverging road made this very tricky, as the flanges of wheels rub inside and out against the rails and check rails. A number of tests with my little wagon and some Sierra Valley wheel sets led to much fettling of the diverging road check rail as well as the inside of the wingrail. Mounting the switchstand was quick and easy; its throw is adjustable, so I was able to adapt it quickly to the throw (I had drilled the holes in the stretcher bar a scale inch too close together). View along the main road of the turnout with the station entrance turnout in the background. Detail of the crossing vee and the diverging check rail and wingrail. To prevent the flanges from binding I resorted to using my Proxxon drill with a sanding disk to sand the faces of the check rail and wing rail; I then burnished the sanded areas with a wire brush to achieve a silky smoothness again. This tiny adjustment made all the difference, and the wheels glide through the sharp curve now. I also filed and burnished the nose of the crossing vee. Although not a turnout, the first piece of the outer loop is on the same template, and after the two turnouts this was a doddle to make. This picture shows the two rail lengths after bending to size. As the rail bender cannot bend the last couple of centimetres of rail, I bend the whole rail length and then trim off both ends to fit. Drilling the 1.5 mm holes in the rail ends. I mark the distance from the rail end using a micrometer, which is enough to scratch a mark against the web of the rail to show where the hole should be; all the rest is by eye. A view through my magnifying lamp showing the fiddly fishing together of the rails. The hexagonal nut I am about to fit sits inside the brass nut spinner held between thumb and forefinger, while I press the bolt head to prevent it from dropping out when I hold the nut against it. With practice this is quick, and satisfying. I tighten the nuts very gently, such that once locked (by tinning the bolt thread with solder prior to shortening the bolts) the fishplates remain a slightly loose fit against the rails. The completed template; it's about a metre wide. A test fit of the template onto the baseboard led to some minor fettling of the edges to get the rails to line up. This template also sees the start of the rising gradient to the branch line and the falling gradient to the mine, so it will have to be heavily weighted when glued in place to ensure that it flexes as desired. The 3 mm MDF is ideal for this. Checking the branch line end of the template before glueing in place. These rails will require fishing together in situ, which will be rather more fiddly than at the workbench. The connection to the mine spur includes a change from 25lb/yd rail to 12lb/yd rail (Code 205 to Code 148), which will need special fishplates. The rail ends will need to be tweaked a little to achieve the fit, but once fished together this will not be noticeable. With hindsight—hmmm—I should have left the mine spur template loose until it was connected to the next one, however this is no big issue. As long as each template is accurate to within a millimetre then all is well. The template with the entrance turnout and the mine turnout, glued in place on the baseboard. This shows the effect of the diverging gradients, adding a sense of vertical dimension. A trackside view showing the mine spur dropping away to the left and the branch line climbing in the distance. I still have to fish the rails together to join this template to the mine spur and the branch line. Finally, a view up the branch line showing the start of the ascent in the middle of the crossover between the turnouts. I hope that Works Engineer John Hurt is content with progress so far...

I cut the sleeper timbers, distressed the top surfaces with a wire brush, and checked that I had them all (I cut the sleepers for the entire template while I was at it). I laid the sleepers out on the template to check all were present and correct. The long switchstand timbers have been moved away from the template locations to fit the width of the switchstand foot. I could not resist laying out the rails at this stage to get an impression of the finished turnout. Also visible is the switchstand—an example of the influence of North American practice on the CMER. Occasional deep snow meant that these stands with their highly-visible indicators were much appreciated by train crews. I stained the timbers using Ronseal Dark Jacobean wood stain; later this will be darkened further with a wash of diluted india ink. I turned my attention to the switch rails—more precisely, to how they would be switched. Working in this scale means I can bolt through the rail foot using the same M1 square-headed 6 mm bolts that I use for the fishplates. Another view of the bolts. Now to figure out how best to suspend the stretcher and kickplate between them and connect to the switchstand. Whilst the CMER generally did not bother with rail tie plates it did use them to support switch rails. These are 1/8" × 6" × 12" in size and serve purely to protect the sleeper and carry some grease under the switch rails; they are fabricated from 0.2 mm tinplate and pre-punched and drilled before mounting on the sleepers. Each joint requires two holes in the rail end; these are 1.5 mm diameter, allowing a loose fit with the fishplates. The originals would have been oval to allow for rail expansion and contraction. The stained sleepers were dry by now, so I glued them to the template and used a straightedge to verify that they are lying flat—which they are. Spiking starts with the outside or main line stock rail. I spiked each end first, aligning the rail over the template by eye (real risk of parallax here as these are full-depth 4" sleepers, 7.4 mm in the model). The tie plates are then slipped under the rail and spiked in place. I also spiked the first stock rail of the diverging line in place at this stage. I make up fishplates for each template as I go (it’s not the most thrilling job); here are the ones needed for this turnout. The inner fishplates are of thinner material (0.2 mm) than the outer ones to eliminate striking of flanges upon the bolt heads. These are drilled 1 mm in diameter and a reamer is twirled in each hole to remove burrs. Close-up of the fishplates fishing the main line stock rails together; the bolts will be trimmed to length later. Another view of the fishplates; next up is to spike the second stock rail in place. With the main line stock rail spiked in place it is time to test-fit the crossing vee... and of course it lines up exactly as it should. Track gauges to the ready! This is the part that makes me nervous, for this is where accuracy is paramount. The slender Cliff Barker G1 track gauges really came into their own here, as they have exactly the 1 3/8" crossing flangeway I need for the CMER (2.6 mm). This test fit suggests everything is going to work out as planned. Another view of the test fit. The diverging stock rail will be spiked only after the crossing vee and the main line closure/wing rail have been placed and thoroughly tested. Measure twice, and cut once... I checked and checked: wheels seemed to roll through the knuckle smoothly enough. The loose heel joint for the switch rails. An initial test fit showed that the thicker, outside fishplate was restricting movement, so I bent it slightly. I shall fish the switch rails with one bolt per fishplate only. I tin the completed joints with solder before trimming the excess length from the bolts. This shot shows the switch rail in its deflected position. Measuring thrice revealed that I needed to trim the switch blade by a millimetre and a bit... had to come off the pointy end, of course, as I had already drilled the holes at the joint; luckily this was easily accomplished. Looking good to progress to the next stage. I should have heeded that niggling sense of unease... Aha: having carefully spiked the main line closure/wing rail into place, complete with its attached switch rail, I realised that I would not be able to fit the stretcher bar, as I wouldn't be able to reach underneath it to attach the nut to the bolt through the rail foot... Ho hum! In the end I decided to cut the bolt attaching the switch rail to its fishplate and remove the switch rail. I posed the picture to illustrate the issue: the stretcher bar with its kick plate extensions will fit the two switch rails like this, however needs to be underneath the stock rails... Note the diverging closure rail with its fishplates, ready to be moved into place. Having connected up the stretcher bar I then had to unspike one stock rail sufficiently to fit the switch rail assembly in place... then I replaced the stock rail and continued spiking the remaining closure/wing rail, diverging stock rail and check rails in place—so focussed on not making any further mistakes that I forgot to take any pictures until it was done. All spiked in place and fished together. Another view of the same status. Fishplate bolts still need to be locked and trimmed, but this was the time to test different wheel sets and push my only wagon back and forth through the turnout. Very pleased to find that everything works as planned, with no binding or catching. No doubt the wide tyres help—but even very narrow ones roll through with only a slight dip between the knuckle and the vee. Next item is the turnout control. As per prototype this is a North-American–style switchstand with an indicator disc. This is connected to the stretcher bar using a simple length of florist’s wire with a loop in it. Bird’s-eye view of the turnout control, showing the 3" throw of the blades. A close-up of the stock-rail joints fished together, and the long flares of the check rails. A close-up of the spiking around the crossing; I used smaller spikes in the confined space between rails and larger spikes elsewhere—hardly visible after the fact but easier to place. Finally, a view of the turnout from the other end. The continuous curve of the diverging road is clear from this angle. I am pleased with this: less work than I had originally anticipated, and easier to put together thanks to pre-bending each rail before connecting anything to anything else. Clachbeg only boasts five turnouts—of which only two are alike—so this represents significant progress. Next will be the curved 1:3, turnout to the mine spur with a 12'-3" radius diverging line, which is on the same template, and then I can transfer the whole template to the baseboard and bed all the engineering into the landscape.

OOps... somehow I did

Updated to add links for scenic materials and also for rolling stock. Surprising (and gratifying) how many small suppliers end up being involved!

Finally I am tackling a turnout, the first I have scratch built in two decades. Having spent a couple of days procrastinating and reading whatever I could find about turnout building I decided it was time to take the plunge and stop worrying about getting it wrong. The planned sequence of events is to bend and cut and shape the rails, then to cut and stain the sleepers, then to fix the sleepers to the template, then to spike the rails in place, remembering to arrange for the switch stand to throw the switch rails. This is the station entrance turnout, built with 25-lb/yd (Code 205) flat-bottomed rail; it is curved, the minimum radius is 34' (756 mm), and it will have loose-heeled switch rails. It is a curved crossing, meaning that the diverging road is curved through the crossing, while the main road has a straight section about 2' before and after the crossing vee (this caught me out when I bent the rails). The turnout timbering is 8" wide (14.8 mm), so plenty of room for spikes. The turnout is part of a larger formation which includes the mine turnout. This shows the entrance turnout. It consists of twelve rails: as it is about 25' long and the maximum rail length used on the CMER is 15', there are four stock rails; then there are two wing rails, two check rails, two running rails and two switch rails. It will require eight fishplates in itself and about 140 spikes. The stock rails are complete, as is the outer check rail. For the avoidance of later confusion I mark each rail with a letter (A to L in this case) as well as the joint numbers at each end. This view shows the angle in the diverging stock rail. I bent this simply by holding one end in the vice and bending the rail (held firmly against a short steel rule) by hand and by eye; two adjustments only were needed. There is more science on paper, but I think this will be accurate enough and as accurate as the original most likely was... The diverging stock rail is straight as far as the switch rail heels and then curves. The crossing vee soldered up. Filing was done largely with an eight-inch file and by eye, with frequent reference to the template. The Code 205 rail is substantial, but on the other hand firm enough that I was able to file robustly and quickly. While soldering I used pins in the baseboard to hold all in place—the holes are still visible. I then turned it over and—after bracing it with two scraps of brass strip which I later removed—soldered from the underside too to ensure a strong joint. Loose assembly of the crossing vee with wing rails, stock rails and check rails. The straight section in the outer (main) road stock rail as it passes the crossing is clear in this perspective. The guard rails have long flares rather than being filed—this is prototypical for the CMER. All of the fixed rails are complete in this view; the crossing vee running rails still need to be trimmed, and then the switch rails are next. First switch rail and its stock rail. I transferred the markings from the template for two sections to file on each rail: the foot only, and then both the foot and the head. In this view I have filed the rebate in the inner stock rail too, Close-up of the mating surfaces of the stock rail and its switch rail. I aimed to file the stock rail back until the file touched the web, whereas I filed the switch rail until the web almost disappeared. To hold the switch rail for filing I cut a suitable oblique groove in a piece of scrap wood and clamped the rail upside-down into the groove, then filed the rail as it protruded at a gentle angle from the edge of the wood. All of this was marked up by eye, but seems to be accurate enough. The result of the careful filing and a number of trial fittings. Both switch rails completed. Next up will be the timbers.

The culvert carries the mine spur across the Allt Creag na Còsaig and is as simple and low-tech as the railway could get away with. Two 4" × 8" beams cross the stream and the sleepers are simply laid upon them. As this spur sees traffic of a single wagon and lightweight loco once or twice a week, and axle loads are restricted to 1½ tons, this was deemed perfectly adequate by the Works Engineer. The mine spur is laid with 12lb/yd rail (Code 143), necessitating lightweight fishplates. I use 30 mm sections of 2 mm × 1 mm for the outside and 0.5 mm thickness for the inside plates; I gently file one side of the inner ones longitudinally to ensure a snug fit against the rail web, but even so in the 12'-radius curves the wheel flanges strike the bolt heads of the fishplate on the outside rail (they are slightly thicker than scale), so once in place I grind their tops down by 0.5 mm, which solves that problem. Here Doctor John holds an inside and outside fishplate to demonstrate the different thicknesses. As I would be building the culvert into the layout I built it directly onto the template with cutouts for the stringers. The sleepers which will rest on the stringers are not glued to the template; once glued to the stringers the template was cut away more-or-less where John is standing. The final 9' panel of track at the mine itself is on its own template; this section is level whereas the culvert is part of the 1-in-25 climb up to the station. As a result the fishplates connecting these two pieces could not be fitted at the workbench but had to be assembled in situ. By now I have mastered this, so it took a matter of minutes only. Note I have not followed the Templot timber spacing here: I could not figure out how to vary timber spacing for differing rail weights and decided it would be easy enough to do on the spot, and in this scale it is. The completed rail joints connecting the culvert to the mine spur end. The stringers are mounted into slots cut into the baseboard; their original abutments have disappeared over time. Note that I distressed these sleepers to varying degrees using a wire brush. They have been stained and will remain more-or-less this colour, as the sleepers on this spur have not been regularly creosoted—in fact, they have never really been maintained at all. I still have to weather them accordingly. The completed culvert. The sleepers are not bolted to the culvert. The fishplate bolts still have to be trimmed. The opposing 1-in-25 gradients of the mine spur and the branch line create a sense of vertical space on the layout.

Now that the sleepering timber has arrived I was able to cut the 6' long bridge sleepers and lay the rails. I would have loved to have been able to fit a larger bridge into this layout, but it is after all only a micro layout and anything larger would have been disproportionate. I am still in two minds about whether guard rails should be fitted; if they were fitted they would have to be made of two rail lengths each, in order to extend two to three feet beyond the abutments at each end. Maybe... Here the bridge is in the foreground, and the cutting template behind it. The Works Engineer is checking that the track gauges are correctly positioned. I did not glue the bridge timbers to the template, but held them with thin strips of double-sided adhesive tape, so that once the rails were spiked in place I could very carefully remove the whole unit from the template. This worked well. For each rail length I spike the first sleeper and then the last sleeper and then the middle sleeper; I then spike the intervening ones. The rail is bent to shape beforehand so it lies on the template with no more than a millimetre or two of overall deviation over its length. The various track gauges are designed for heavier rail with wider rail heads, so that my gauge wanders between 2' and 2'1½"; this is not tragic as the rolling stock will be fitted with (prototypically) wide-treaded wheels. Glueing the sleepers to the bridge, after several in situ checks to ensure the bridge track aligned with the abutment tracks. The second, fourth and sixth sleepers will further be bolted to the stringers and the others nailed in place. Test-fit of the bridge in situ. Note the sleeper bolts. The vertical alignment requires card shims under the cap beams.

This is the part I have really been looking forward to: laying individual prototypical rail lengths and bolting them together with fishplates... The CMER (as a reminder) was originally laid with a mixture of 12lb/yd and 25lb/yd rails, which are 2" and 2¾" high respectively. The main and branch lines were gradually all relaid with the 25lb/yd rail in 15' lengths, while the 9' lengths of the lighter rail were used in sidings. The original 36" sleeper spacing was retained in sidings, however as rails frequently distorted, subsequently during maintenance the main line and the Clachbeg extension were relaid with a seven sleepers per 15' rail spaced at 12½"–2'7"–2'7"–2'7"–2'7"–2'7"–12½" centres and on bridges eight sleepers spaced at 1'–2'2"–2'2"–2'2"–2'2"–2'2"–2'2"–1' centres; turnout sleepering is also spaced at around 24" centres. Scale lumber stacked up: 6" × 4" sleeper timber in the foreground and 8" × 4" turnout timber in the background. This is lovely basswood from the USA, from Kappler Scale Lumber. I started with the template of the track from the fiddle yard turntable through the cutting to the bridge. three full rail lengths plus a very short piece at the left end of the picture. I cut the requisite sleepers and roughly sanded off the edges at their ends. As these ones will not be seen I have not distressed them at all. I used the Fasttracks rail bender to bend the Code 205 rail to fit the Templot-generated template. I had to order this rail from the US as no-one in the UK stocked it; to save on shipping costs I arranged with Microengineering that they cut their stock 6' lengths into five 14 13/32" lengths, which works out as 16'5" scale: ideal for me as I would be cutting each length anyway. The short rails at the turntable end of the template and the joint. I left a ½" gap between rails (handily the thickness of a piece of card stock). The short rails are straight and about 3' long: they form the transition between the 33' radius curves of the fiddle yard tracks and the cutting track. In my impatience I decided to glue the sleepers in place at this stage... it would have been better to stain them first! Still, easily-enough remedied later. Drilling the four holes in a fishplate, using a very rudimentary jig and simple scribed measurements. Platelaying ironmongery: 500 bolts and 500 nuts as well as two nut spinners for the M1 nuts. Fishplates! For the Code 205 rail I use 30 mm long pieces of 2.5 mm × 1 mm brass strip from Eileen’s Emporium. This is a prototypical size and fits snugly into the web of the rail; I learnt that for curved rail these are quite stiff, and use 0.5 mm thick pieces on the inside of the rails, which also has the effect of ensuring that deep flanges do not strike the bolt heads as they are slightly more recessed. The bolts are little square-headed marvels from miha in Germany, 1mm wide and 6mm long, with hexagonal nuts. I drill the fishplates with a 1 mm drill and the rails with a 1.5 mm drill. Fitted... the first ones took nearly thirty minutes, as I grew used to the tiny nuts and bolts. I attempted unsuccessfully to use the pin chuck as a spanner. The fishplate here is as-is; I learnt to stroke the end edges against a needle file to remove burrs. The bolts are rather too long and will be shortened later with side-cutters—the length is useful when assembling the joint. The nuts are not tightened much (as with the prototype: they only align the rails), so I add a dab of superglue to each to hold them in place. I like the look of the fishplates inside the rails in this along-the-track view; once you start looking, this is often a feature of narrow-gauge track, in which the fishplates are relatively massive. The second length of rail attached to the initial stubs. The fiddle yard-side sleepers are unstained. This shows a Sierra Valley wheel set on the track; its flanges easily clear the medium spikes I use with this rail size (I use small spikes with the lighter rail). The proportions of the sleepers and rail are absolutely prototypical and are very pleasing. Close-up of the rail bender in action. Although rated up to Code 148 it performs perfectly well on Code 205 rail. It cannot bend the last ½" or so of each rail, so I mark this and measure from the mark, later cutting off the tail if the rail is to curve to its end. A completed rail joint showing the ½" expansion gap. Note too the slightly closer sleeper spacing at the joint, as well as the spiking pattern.

Time to take the plunge and to progress from bare baseboard to something more natural... After much reading I decided to opt for fairly mainstream landscaping techniques: a Flockit applicator, Flockit glue, Woodland Scenics Scenic Glue in a spray bottle, and a mixture of 6 mm and 12 mm static grass fibres and Woodlands Scenics long grasses, as well as a collection of scenic scatters, gouaches, India ink, some horsehair mat and some classic Iceland moss. First off I laid down lots of grass. The 12 mm XL fibres are about right for this scale. I learnt that they cling to everything, especially the cling film protecting the backscene! Covering areas of a few square inches at a time I covered the whole corner liberally. I used scenic glue in a spray, which was not smart, as glue covered the rock faces and quite a lot of fibres adhered unrealistically to them. However where it worked, it worked really well. I also experimented with short fibres followed by long, and vice versa, both producing good if differing results. I mixed autumnal and summer and lush greens. The cutting, including the ballasted track. For ballast I used Swift Sixteen material, which is the right size but was very light grey. I brushed it between the sleepers (so much easier than in smaller scales) and then decided that I wanted it darker. Not having a pipette to hand (poor planning) I decided to spray scenic glue onto it to fix it, and I used some India ink to colour the glue. The result was actually rather good: the sleepers acquired a pleasing creosoted appearance and the ballast was toned down. I wiped the rail tops to keep them smooth; I shall rust the sides and the fishplates later. I did all this after adding the grass—I had to mask the grass before spraying on the blackened glue, fearing the worst, but it worked out well. The colouring did not penetrate well beneath the rails—see far left of picture—so some touching up is required with a small brush. After 24 hours the ballast has set rock solid, as it were. The culvert (not creosoted: I shall have to weather it separately) showing the effect of the 6" high grass fibres. Next step was to add tussocky grass. Easy to do, and I aimed for 15"–24" tussocks, in damp ground. Note too a little experiment with dark green scatter material on top of the grass, fixed with scenic spray—I was starting to enjoy myself! The mine area grassed over (alas, when the buildings catch up this will have to be disturbed). I also added some heather on Creag na Cosaig (horsehair teased out and sprinkled with scatter material) and the basis of some gorse bushes (Iceland moss). Clearly the bridge is missing in these shots. The gorse with flowers added (Humbrol enamel paint). Water added to the Allt Creag na Còsaig. The unweathered bridge was test-fitted and this is what it looks like. The rails are very shiny! I am pleased with the effect so far; I still plan some more gorse bushes on the skyline and a couple of trees.

The cutting is where the track emerges from the fiddle yard (a cutting not a tunnel; the cutting is deep enough that the viewer cannot see around it to the entrance to the fiddle yard), and here the track passes between Creag na Còsaig to its rear and Creag Ór to its front, crossing the Allt Creag na Còsaig on a 15' bridge before reaching the station entrance turnout. The tiny gold mine is (fittingly) at the foot of Creag Ór and its spur crosses the stream on a 6' culvert. All of the above means that I need to complete the scenic treatment of the stream before fitting the bridge and culvert in place... so I decided to tackle the scenery of this self-contained corner of the layout at this stage—tied in with laying the completed track of the cutting and the mine spur. This way I'll be able to run trains to and from the fiddle yard as I progress with the further tracklaying. It is decades since I tried my hand at any sort of scenic work, and consequently I put this off for some time before screwing up the courage—thinking it all through proved worthwhile. This was my starting point: the bare track templates and coloured but otherwise bare baseboard. This also shows the tiny footprint of the mine—a separate challenge to be addressed later. The Works Engineer, Doctor John, surveys the site. Note the 12lb/yd rail of the mine spur vs the 25lb/yd rail of the branch line. The mine spur sleepers are spaced at 36" centres, whereas the branch line sleepers are at 31" centres, and the longer bridge sleepers are at 26" centres. The pins in the scenery indicate where I was thinking of placing bushes. The track templates were glued down and pressed firmly into place overnight, after careful alignment of the bridge with the tracks at each end: the bridge is still removable at this stage. For the culvert I decided to build it straight in, as the streamed underneath it is very accessible. First task was to build up the ground level. In the cutting I used plaster, coloured approximately using squirts of gouache in the mix, to cover the sides of the track template up to a level just below the sleeper tops, leaving place for ballast between and around the sleepers. For the mine spur I simply filled up to the sleeper tops, as any ballast here has long since sunk and merged with the ground. I also filled the area around the mine shaft (cunningly using a different colour in the plaster mix). I have still to develop the detailed plan of the shaft and derrick and shed, so this area will simply be grassy until such time as the buildings catch up. This shows the mine spur plastering carried on up the spur; I completed the ground cover almost to the end of the template. On the branch line to the rear there is a 6' length of track between the bridge and the turnout template: this short joining template has also been completed and glued down, to provide for aligning the bridge, and it will also be ballasted. My ad hoc colouring of the ground cover plaster proved rather too Saharan for the Highlands, as this picture reveals. Detail of the mine spur, which is sinking slowly into the ground. Those are 9' rail lengths. Detail of the short culvert on the mine spur: as simple as could be. Any abutments have long since been overgrown and sunk into the stream banks. A wash of Burnt Sienna mixed with a little black greatly improves the look. This is the top of the mine spur, with the entrance turnout template visible (as well as the bridge, temporarily placed there). Culvert area with improved ground colour. The corner of the layout with ground colour applied, ready for ballasting of the branch line. The upper part of the mine spur, which descends at 1 in 25 to the culvert, where it levels off. Doctor John is keeping an eye on progress.

I made good progress with the turntable, after various calculations to ensure that the top of the rails would match the spur leading to the layout itself. Finally I chose to lay a sheet of 18 mm MDF onto the turntable and then lay Code 143 rail directly to the MDF, to align with the top of the Code 205 rail on sleepers on a template... It worked, thank goodness! The first picture shows the template for the fiddle yard ends laid out for size. The cutouts in the MDF are for the stiffening girders. The Code 143 rail has been skipped directly to the MDF. The Peco rail in 3' lengths is shorter than the fiddle yard; the joints were simply aligned and soldered. I realised that the rail ends were going to be exposed to snagging and bending (some rather poor planning had resulted in their overhanging the MDF by a couple of millimetres...), so I made a virtue of the necessity and fabricated folded tinplate brackets to which the rail ends are soldered, which will protect them as well as keeping the gauge. Joys of not running power through the rails! This shows the indexing. The turntable base has a slight play, which is of course amplified over its length. Much turning this way and that was done to establish how the rails line up with, respectively, the branch line, and with the two short spurs I managed to squeeze in. I fitted the sliding part of the latch and used a Sharpie to colour the tip of the bolt, then used that to mark the edge of the MDF, and then I drilled the MDF at the appropriate places: nine holes required at each end of the board! The indexing is accurate enough to allow derailment-free crossings; there is still a little play in the mechanism, so I shall need to check quickly by eye and nudge a little, however that should not be a problem at the pace of operations of Clachbeg! The wagon is being used to test free running across the joints; it is a flat by miha, destined to be wagon 53, although that may have to change as the tank is too long for this 6'3" chassis... Also visible here is the end of one of the two very short spurs I squeezed in.

Amended to include a link to fascinating source material relating to Archie, or Vijay Padurachee.

Extract from The Minor Diaspora – Indians in odd places in the nineteenth century, R.K. Chetty, Chattapore University, 1926. When one considers the talent which goes unobserved in Indian’s teeming masses, it is fortunate indeed to find one such talent that was recognized. Vijay Padurachee was born in the early 1860s in an obscure village in Uttar Pradesh. It soon became apparent to his parents that their son was either grossly stupid or a boy of great precociousness. His parents brought him to the attention of the local headman, a man who had worked in the house of a Burra Sahib. Over the course of many years the headman had amassed a small library from those books lost by the sahib. The books were on display in the headman’s house, and helped to impart a reputation for learning and sagacity. With the indulgence of the headman, Vijay was encouraged to look at the books, amongst which were Defoe’s Roxanna, an illustrated copy of Maxwell’s Highland Sports and Pastimes, Magnall’s On the Principles of Steam Propulsion, and—most importantly—an English–Urdu dictionary. By the time the boy was six he was able to read everything in the headman’s library and had become proficient in English, the most important language in India. When Vijay was about ten his life changed for ever. A platoon of the 15th Corps of Engineers rode into the village one day, and their havildar called the village elders to a meeting. It became apparent immediately that there was a problem, the havildar was a Sikh and was not fluent in Urdu, and the headman, that local paragon of learning, was woefully lacking in English. After several minutes of fruitless haranguing someone had the bright idea of calling Vijay to try his English. The havildar was astonished by the boy’s command of the language, and through Vijay was able to explain that the Engineers were surveying the proposed line for the railway which was to pass just to the south of the village. When that was done, in a few months the army would need labourers, and were looking for able-bodied men. A couple of days later the havildar returned, this time accompanied by an English officer. The village crowded round; most of them had never seen an Englishman before, and this one was a small man with flaming red hair. Vijay was called for and had a conversation with the officer. The villagers gathered round listening to the incomprehensible language, proud that Vijay had such learning. Later, according to the headman, the officer had said “extraordinary” at least five times. Vijay not only became official interpreter during the railway construction, but developed a lifelong passion for railways. In his memoirs, the English officer, Major Wellington de Vere (1825–1900), recalled that first meeting with Vijay: “he was a small, skinny boy, like thousands like him all over the Province. His eyes, though, were alive and bright with sparkle and a thirst for knowledge.” De Vere made arrangements for Vijay to be educated in Lucknow at an English medium school for Indians. His passion for the railways was recognized by the Headmaster, John Fitzherbert (a friend of Lockwood Kipling), and Vijay was encouraged to join the Indian Railway Service as this was one of the few ways in which young Indians of no social standing could advance themselves. During these years Vijay must have picked up Hindi. After completing his schooling he was apprenticed to The Great Trigonometric Survey of India; here he had a thorough grounding in all aspects of the laborious and exacting task that was to take over a hundred years to complete. At the age of about twenty-five Vijay Padurachee joined the survey and engineering section of the Darjeeling Himalaya Railway and was fortunate enough to work under Herbert Rumsey at Kurseong. All went well for a time, but then the great good fortune that had played so prominent a role in the young Indian’s life deserted him. He was described at the time as a quiet, serious man who spoke English with the prolixity of someone who had learnt the language from novels of the eighteenth and early nineteenth centuries. There are no known likenesses of him, but he must have been good-looking, for Ishbel Rumsey fell madly in love with him. As soon as her parents became aware of what was probably no more than a one-sided teenage infatuation, Vijay was summarily transferred away from Kurseong back to Tinharia in what was, in effect, a demotion. His section manager there was Captain Treducq, second-in-charge was Lieutenant McTosh. Vijay was demoralised, realising that—through no fault of his own—his future in the railway service was blighted. McTosh had told Vijay about Clach Mhòr, the remote and obscure estate in the North West Highlands of Scotland. Not surprisingly, Vijay had read about the Highlands and—knowing that Scotland was a small country—assumed that Clach Mhòr was neighbour to Balmoral. The two men became fast friends, united by a passion for the two-foot gauge. When McTosh inherited the estate in the 1880s it was only natural for him to try to persuade his friend to come over to Scotland with him, and it was only natural for Vijay to accept. Vijay, with his dark good looks, was an exotic import to this land of fair-skinned and freckled people. He soon realised that Clach Mhòr was more remote from Balmoral than Darjeeling was from Calcutta, and so his hopes that he might meet the Queen were dashed. However, he soon had other social obligations in the lively, lissome form of Maree Morrison, a housemaid at the Lodge and daughter of one of the Kinlochbervie Morrisons. After a few months Maree, although still lively, was becoming noticeably less lissome. Vijay was a man of honour and said that he must make the journey to Kinlochbervie and confront her parents. Like most peasants from Central India, Vijay had a deep mistrust of the sea and preferred to make the overland journey of three days to Kinlochbervie rather than face an uncertain ocean crossing of just a few hours by fishing boat. He need not have worried—Maree’s mother just smiled indulgently and said that nobody ever counted, that was just the way things were done round here. Her father, Hector Morrison, looked “more sheepish than his flocks” as he shook the young man’s hand. Vijay found peace and happiness in Clach Mhòr, and seldom left the estate. The locals found the name Padurachee difficult to pronounce. As there were already dozens of “Paddy”s employed in the railway building, he became known as Archy. Archy and Maree had just one child, a son, Murdoch, born in 1890. And so they lived their quiet, contented lives, their focus always on Murdo. He was encouraged to learn Urdu and the folklore of India by his father and his mother, gentle Maree, taught him to love the countryside, the mountains and glens that made up his home. Maree died in 1908, followed within three months by her devoted husband. He had never shown much interest in religion, and the parish priest at the time shocked some parishioners by allowing Vijay to be buried in the north west of the cemetery next to Maree. “He was a good man,” said the reverend, “and are we all not God’s children?” Here you can still (1925) see the foreign name V.J.Padurarchy (sic) engraved on the stone. In spite of his parents’ best efforts, Murdo was a wild boy, preferring the hills and crags to his schoolwork, occasionally in trouble with the estate manager for stealing birds’ eggs, and for being the first person to have a race with the railway over the whole network and win. It was in climbing the unnamed windswept mountains round his home that Murdo found his passion, and in due course he became a guide for the wealthy tourists who were making their way up to this remote corner to enjoy a brief communion with nature. Later, Murdo survived the nightmare of the Western Front and returned home to become almost reclusive among the crags and glens, coming to realize that he might prefer his father’s stories of the heat baked plains to the wilds of the Highlands. In 1923 he applied to join the ill-fated Mallory and Irvine Himalayan expedition, and he was hired as an interpreter. There were many in the estate happy to see him go. And so Murdo Padurachee returned to the foothills of the Himalayas in 1924, almost thirty-five years after his father had left. Now he could match his father’s descriptions of the foothills of the Himalayas and relive the excitements and enthusiasms of previous generations as the expedition took the narrow-gauge railway as far as they could go. Murdo Padurachee died in an avalanche sometime towards the end of May, 1924, at the age of twenty-seven. There is no-one else of Indian ancestry left in Clach Mhòr, the Highland home of the Padurachees for almost forty years. My thanks are due to the Manager and staff of the Darjeeling Himalaya Railway, and the kindness shown last year on my visit to Clach Mhòr Estates. In both locations I rode on their wonderful two-foot railways.

Initial plans for the Stonemason’s shed. Paradoxically this will be a brick building (records show it was built of bricks from the Brora Brickworks; the bricks were probably taken on as ballast by a coastal steamer and then unloaded at Kinlochy), with a stone plinth and stone platform and a slate roof. The design is my own and will be refined further. At this stage I’m tempted to build this “properly”: using stones for the base and scale bricks as well as wood and glass for the windows, and full roof timbers and slates for the roof... Maybe that will be too ambitious, however I think I shall give it a try once I reach the stage of constructing the buildings.