In the first article about the line from Cuneo to the sea we covered the length from Cuneo to Vernante. The article can be found here. [9]
The Line South from Vernante to Limone
A schematic drawing showing the main locations on the line from Vernante to Limone. [17]
Banaudo et al write that “It was only in 1886, after the creation of the Rete Mediterranea, that the work on the fourth tranche from Vernante to Limone was awarded. It was 8,831 m long and had a gradient of 203 m, which was to be compensated for by a continuous ramp of up to 26 mm/m. This value would not be exceeded at any other point on the line. On this section, the rail remained constantly on a ledge on the steep slope on the right bank of the Vermenagna, where it was anchored by eleven bridges and viaducts totaling sixty-three masonry arches, as well as nine tunnels with a combined length of 4,416 m, or just over half the route:” [1: p28]
the Tetti-Chiesa tunnel which is 122 m long;
the Elicoidale tunnel (the Vernante Spiral tunnel) is 1,502 m long;
the Rivoira viaduct has fourteen 15 m arches and one 23 m arch;
the Rivoira tunnel is 251 m long;
the Santa Lucia viaduct has three 12 m arches;
a short span masonry arch over a minor road;
the Santa Lucia-Noceto tunnel is 348 m long with two openings;
the Noceto viaduct has six 8 m arches;
the Marino viaduct has two 8 m arches and two 12.50 m arches;
the Marino tunnel is 202 m long;
the Mezzavia viaduct, three 11 m arches;
the Mezzavia tunnel is 444 m long;
the bridge over the Ceresole valley has two 10 m arches;
the Boglia tunnel is 1,086 m long;
the San Bernardo viaduct over the Sottana valley has two 6 m arches and three 10 m arches;
the Cresta-Molino tunnel is 335 m long;
the Boschiera viaduct has twelve 10 m arches;
the Rocciaia tunnel is 126 m long;
the Rocciaia bridge is a single arch;
the first Rocciaia viaduct has four 8 m arches;
the second Rocciaia viaduct has eight 8 m arches.
We start this next length of the journey at Vernante Railway Station and head Southeast.
A plan of Vernante Railway Station. [10]Vernante Railway Station: the route to Limone leaves at the bottom-right of this image. [Google Streetview, July 2025]The view Southeast from the station car park, after demolition of the old goods shed. The main station building features at the centre of the image. [Google Streetview, June 2025]The main station building at Vernante seen from the West. [Google Streetview, June 2025]
Photographs showing the station building and the goods shed prior to its demolition can be seen here. [58] “Inaugurated in 1889, the station served as the terminus for the Cuneo-Ventimiglia line for nearly two years, until it was extended to Limone Piemonte. The passenger building features classic Italian architecture, with two levels. It is square, medium-sized, and well-maintained. Its distinctive feature is the two murals depicting scenes from the Pinocchio fairy tale, adorning its façade. The lower level houses the waiting room and self-service ticket machine, while the upper level is closed.” [58]
A photograph from the cab of a Cuneo-bound train arriving at Vernante. The passenger building is on the left with the goods shed beyond. [8]
The view from Via Frederi Mistral which passes over the tunnel mouth at the Southeast end of Vernante Railway Station. [Google Streetview, June 2025]
The very short tunnel (Tette-Chiesa, 122 metres in length) at the Southeast end of Vernante Railway Station. [Google Maps, July 2025]
The southern portal of the Tette-Chiesa Tunnel seen from a Cuneo-bound train. Immediately beyond the far portal trains would have to stop to manually engage a point for the running line or the train would end up on the safety siding provided for runaways on the steep downward gradient. [8]
The large retaining wall on the left of this image supports the railway as it runs immediately adjacent to the E74/SS20 but at a higher level. [Google Streetview, June 2025]The height of the retaining wall decreases as the E74/SS20 gains height. [Google Streetview, June 2025]
Banaudo et al comment: “Leaving Vernante, the track describes a complete spiral loop at Rivoira, which allows it to rise about fifty metres over a circular length of two kilometres. This loop includes the 1,502 m long ‘Elicoidale’ tunnel, which was completed on 30th December 1889, and the imposing viaduct over the Salet torrent. With its fifteen arches, from the top of which the rail dominates the lower level of the loop by 45 m, this structure can be considered by its proportions as the most imposing of the whole of line. [25] It is built entirely of cut stone, with the exception of the intrados of the arches which are of brick, and its seven central arches are reinforced at their base by a series of arcades forming an additional level, following a technique very popular in the 19th century.” [1: p30] The lower arcades are seen clearly in the 1929 postcard below.
This photograph is taken from the road at the point that the E74/SS20 begins to turn away from the lower railway (which can be glimpsed through the undergrowth) the viaduct high above both the road and the railway comes into view. This view looks North from the E74/SS20. A spiral tunnel allows the railway to gain height at this location. [Google Streetview, June 2025]This satellite image shows the portals of the Spiral tunnel to the East of Vernante. The line leaves Vernante Station and passes through a short tunnel before running alongside the E74 ‘Corso Torino’ to another tunnel mouth to the West of the side road. The line then climbs as it circles under that road twice and reappears high above the first length of line towards the top-left of this image. The height gained then means that the line needs to pass over a high viaduct before once again entering a tunnel (the Rivoira Tunnel) and then, at the bottom-right of the image, crossing another side valley on a bridge. [Google Maps, July 2025]OpenStreetMap shows the same location and illustrates the spiral tunnel quite well. [44]The lower portal of the spiral tunnel with the high viaduct (Rivoira Viaduct) visible to the left. [11]
The portal of the spiral tunnel at the top-left of the satellite image above, seen from a Cuneo,-bound train. Trains heading for Tende and beyond gained height while turning through 360 from the tunnel portal shown in the image immediately above. [8]
The Southeast portal of the short tunnel at the bottom-right of the satellite image above. This is the Rivoira Tunnel. [8]
The Santa Lucia viaduct just to the Southwest of Rivoira Tunnel. [8]
Between the Rivoira Tunnel and the Santa Lucia & Noceto Tunnel, the line crosses a minor road serving a few small hamlets. [Google Streetview,
The Santa Lucia & Noceto Tunnel runs diagonally across this extract from Google’s satellite imagery. [Google Maps, July 2025]
The Southeast Portal of the Santa Lucia & Noceto Tunnel seen from the cab of a Cuneo-bound train. [8]
The Noceto Viaduct to the Southeast of the Santa Lucia & Noceto Tunnel spans a local stream. [8]
This bridge is a short distance further Southeast. [8]
The Marino Viaduct further to the Southeast. All these views look towards Vernante and are taken from the cab o a Cuneo-bound train. [8]
The Southeast portal of the Marino Tunnel. [8]
Another viaduct over a short side valley to the Southeast of the Marino Tunnel, this is known as the Mezzavia Viaduct. [8]
The East portal of the Mezzavia Tunnel. [8]
Immediately to the East of the Mezzavia Tunnel the line bridges a stream before entering the Boglia Tunnel. The bridge spans the Ceresole valley. [8]
The view of the line looking West from Frazione Ceresole, above the West portal of the Boglia Tunnel. [Google Streetview, June 2025]
The Boglia Tunnel carries the line around a significant curve. This is the South-southwest portal of the tunnel from the cab of a train which has recently left Limone. Trains from Cuneo enter the tunnel traveling East and leave in a south-southwesterly direction. Just beyond the South-southwest portal the line bridges another side road serving a number of hamlets. It is the San Bernardo viaduct over the Sottana valley. [8]
The bridge shown in the image immediately above is at the centre of this satellite image. The tunnel to the North-northeast is Boglia Tunnel, that to the South-southwest is Cresta Molino Tunnel. [Google Maps, July 2025]
Looking East along the Sottana Valley, it is difficult to believe that the San Bernardo Viaduct has two 6 m arches and three 10 m arches, it is so well camouflaged by vegetation. [Google Streetview, June 2025]Looking West along the road through the structure, it is possible to see three of the five arches. [Google Streetview, June 2025]
The Cresta Molino Tunnel curves throughout its length (see below). Towards the South portal, it has an open gallery facing out into the valley. [8]
The Cresta Molino Tunnel curves form a South-southwest bearing to just to the East of South along its length. The gallery shown above is at its southern end. [Google Maps, July 2025]
The South portal of the Cresta Molino Tunnel is the South end of the gallery. [8]
After a very short length of track open to the elements, the line enters another short tunnel, the Rocciaia Tunnel. This tunnel is also on a curve with the line leaving the tunnel heading Southeast. [Google Maps, July 2025]
The Southeast portal of the Rocciaia Tunnel. After this tunnel the line crosses a bridge and two viaduct on its way into the station at Limone. [8]
The length of the line from Rocciaia Tunnel to the station throat at Limone is shown on the satellite image below. The parapet railings associated with the Rocciaia Bridge can be seen on the image of the South portal of the tunnel above. There are then two viaducts, as shown on the satellite image below. They cast shadows onto the valley side to the east of the line.
The bridge mentioned above, seen Looking Northwest from the cab of a Cuneo-bound train. [8]The viaduct immediately to the North of Limone Railway Station, also seen looking Northwest. [8]
Limone Piemonte as shown on OpenStreetMap. Note the bridge at the South end of the station site and the tunnel that trains enter soon after crossing that bridge. [18]
The good shed at Limone Station with the passenger facilities beyond. This image is a still from a video taken from a train heading for Breil-sur-Roya. [31]
Limone Railway Station as it appears on Google’s satellite imagery. [Google Maps, July 2025]
Looking North from the end of Via Colonello Domenico Rosetto.The goods shed is close to the centre of this image. [Google Streetview, June 2025]Limone Railway Station building and forecourt. [Google Streetview, June 2025]
A few more photographs of Limone Railway Station can be found here, [22] here, [23] and here. [24]
Express services took 1 hour 30 minutes to travel from Cuneo to Limone, mixed goods and passenger trains were scheduled to take 2 hours. Services from Limone to Cuneo were scheduled for 1 hour 20 minutes and 1 hour 50 minutes respectively [1: p31]
Banaudo et al tell us that a single third class ticket between Cuneo and Limone cost 1.65 lire. The service was deemed to be a local service and as a result the RM allocated older stock to the line, “consisting mainly of single-axle coaches, side door stock, and brake vans acquired from other companies. Traction was provided by 030 [in the UK these would be 0-6-0] locomotives coupled to two- or three-axle tenders, from the RM 3201 to 3550 series (future 215 FS Class),” [1: p31] out-stationed to the Cuneo shed by the Turin Shed. These locos had a range of different manufacturers in Italy, France, Belgium, Great Britain, Austria and Germany. [1: p31]
The construction costs for the length of line from Cuneo to Limone “did not exceed 10 million lire, a remarkable figure given the difficulty of the work and the number of engineering structures completed over nine years: nineteen bridges and viaducts, fourteen tunnels, and a large number of culverts, aqueducts, road overpasses and underpasses, and level crossings. The buildings of the seven stations are of classical design, conforming to the standard plans with hipped roofs used in Italy, as are the twenty-four ‘caselli’, roadside houses, distributed along the line near the level crossings and the main underpasses to house the track maintenance workers and their families. The bridges and viaducts, with the exception of two brick structures, are made of stone masonry with brick arch vaults and metal angle railings. The single track tunnels are lined with brick vaults and dressed stone portals, except where the solidity of the ground allows the exposed natural rock to be preserved.” [1: p32]
Banaudo et al note that “the first years of operation were not easy, … snow and falling rocks sometimes hampered train traffic. On 2nd October 1898, following torrential rains in the high valleys of Piedmont, the Gesso overflowed and the bridge between Boves and Borgo-San-Dalmazzo was destroyed. By December, the installation of a temporary wooden bridge by contractor Salvatore Vignolo of Genova-Sampierdarena allowed service to be restored. A permanent structure would be rebuilt the following year in the form of a single-span 74-metre steel truss bridge.” [1: p32]
Limone to Vievola: Crossing the Col de Tende
The next length/tranche running South from Limone was 10.5 kilometres long and extended the line from Limone to Vievola(in the valley of the River Roya).
Looking into Limone Railway Station from the tunnel mouth South of the Station. A short two-span bridge
At the South end of the Limone Station site the railway bridged Piazza Risorgimento/Viale Valleggia at the East end of Piazza Risorgimento and the River San Giovanni (Valleggia Torrent) on two adjoining bridges. [Google Maps, July 2025]The two bridges carrying the railway over both the road and the river. [Google Streetview, August 2011]
Omitting mention of the section of the bridge over the road, Banaudo et al tell us that, leaving Limone Station, “the line crosses the San Giovanni valley … on a 13-metre masonry single-arch bridge, then enters the 423-metre-long Limone Tunnel which passes under the San Secondo hill. A 26 mm/m gradient leads to the tunnel under the ‘Colle do Tenda’ … where the gradient eases to 2 mm/m as far as the highest point on the line, 1040 [metres above sea level, in the tunnel]. From this point a 14mm/m gradient extends to the South portal of the tunnel … at 990 [metres above sea level]. At the Southern end of the tunnel, … a single-span 19.90 m steel truss bridge crosses the Roya River. … A short 25 mm/m slope then leads to Vievola Station.” [1: p34]
The railway is protected by two galleries at the South end of Limone Tunnel. The first effectively extends Limone Tunnel southwards. This is the South portal seen from a train approaching Limone Railway Station. [8]
Also seen from the South from the cab of the same train, this is the South portal of the Short second gallery. The gallery entrance to the tunnel above can be seen only a very short distance beyond this gallery to the North. [8]
A level-crossing on the line just to the South of the galleries illustrated above and also seen from a Limone-bound train. [8]
The northern approach to the tunnel under the Col de Tende as it appears on Google’s satellite imagery. Sadly, the tunnel mouth, in the top-left quadrant of this image, is in shade. [Google Earth 3D, July 2025]
Open Streetmap shows the line heading South into the tunnel. [32]
This image shows the North Portal of the tunnel under the Col de Tende. It is taken from the cab of a train heading for Breil-sur-Roya in the late 20th century. [31]
Interestingly, the two tunnels on this length of the line are large enough to accommodate two tracks – this facilitates ventilation but also allows room for expansion should traffic levels later require it. [1: p34]
Another schematic drawing which this time shows the main locations on the line from Limone to Vievola. [17]
While all the previous construction tranches ended up in populated locations, Vievola was just a place name in the commune of Tende with a few farms and a chapel dedicated to the Visitation of the Madonna scattered in a small green area at the confluence of the Roya and the Dente rivers. Nowhere was available to house workers on the railway. So before works began at the southern end of the tunnel under the Col de Tende, the contractor had to construct a temporary village.
After initial surveys were completed late in 1889, tunneling under the Col de Tende began at both ends. Banaudo et al explain that the 8.1 kilometre tunnel passed through various different strata: “Jurassic, Triassic and Cretaceous limestone, Permian quartz, Liassic marly schists and Eocene sandstone. The work progressed normally until September 1893, when the works reached a dislocated gneiss bed interspersed with clayey layers made fluid by the infiltration of water from the Roya, whose bed passes three times above the axis of the tunnel. Soon, mud floods invaded the approach tunnel with each attempt to advance over the course of ten months. The working face advanced only a dozen meters, while some forty flows of various materials obstructed the tunnel, sometimes over a length of 40 metres, while the vault suffered as much as 1.7 metres subsidence in places.” [1: p32][33]
The works from the South were suspended in July 1894 about 1.6 km from the tunnel mouth. Attempts were made to divert ground water from the route of the tunnel with little success and a further collapse occurred in October 1894. [33]
Meanwhile, work progressed from the North until at about 2.7 km from the tunnel mouth ground water started entering the tunnel at a rate of 60,000 litres/minute. The bed of the River Royal above the tunnel began to collapse. The contractor admitted defeat and refused to continue work on the line. [1: p34][33]
After a few months delay and with the work now being undertaken by the state a renewed effort was made to take the work-faces forward. The solution was to bore the tunnel using compressed air drills inside a metal shield and with water being removed by a parallel collector channel. It took 470 days to progress the works beyond the difficult strata. Banaudo et al say that once work was 43 metres beyond the critical zone, the contract was handed back to the original contractor on 31st March 1896. The total delay was 34 months at a cost of 300,000 lire! [1: p34][33]
On 15th February 1898 at 1pm, the team working from the North end of the tunnel broke through the remaining rock to meet the team working from the South.Remaining contract works would mean that opening of the line between Limone and Vievola would not take place until 1st October 1900. [33][34: p116][1: p35]
When trains left the confines of the 8 kilometre tunnel their crews were probably grateful for the fresh air. I cannot imagine what it must have been like for the crews of steam engines on the line. Electrification could not come soon enough. “The tunnel was equipped with a two-wire contact line when the electrification of Cuneo Gesso – San Dalmazzo di Tenda line in three-phase alternating current 3.6 kV – 16⅔ Hz took place with electric traction starting from 15th May 1931.” [33][35: p171-172]
South of the tunnel, the railway crosses the River Roya before entering Vievola Railway Station.
This satellite image shows the line leaving the tunnel (at the very top of the image) and crossing La Roya (towards the bottom of the image). [Google Maps, July 2025]
It is not possible to see the tunnel mouth in this panoramic photograph taken from the E74 (D6204), nor is it possible to see the railway bridge over La Roya. The railway can be seen, as can the buildings close to the tunnel mouth on the East side of the line. The railway bridge over the river is behind the trees in blossom one a line from the camera to the red-roofed buildings. [Google Streetview, April 2008]As the E74 (D6204) descends along the valley of the Rya, the railway bridges it, adjacent to a road (off to the right of the picture) which serves Vievola Railway Station. [Google Streetview, October 2008]
The completion of the fifth contract still required the development of Vievola station. It was to be built on a large platform created using spoil from the tunnel works on a vast embankment formed from the tunnel spoil, with an underpass provided for the then SS20 (now E74/D6204) and shown above.
The approach to Vievola Railway Station from the South, as seen from the cab of a Northbound train. [8]
Banaudo et al tell us that, at the station, “The two platform tracks for passenger service were supplemented by two sidings and a dead-end track running alongside the goods shed and the military platform. At the western end of this section, a small wooden shed, an 8.50 m temporary turntable, a water tower, and two hydraulic cranes allowed locomotives to use this temporary terminus as they would at any terminus. In the same area, a wooden buffet building was built, which a shrewd manager, no doubt hoping to take advantage of the cosmopolitan movement of connecting passengers, dubbed a ‘restaurant’ in French.” [1: p40]
Vievola was a railway terminal for traffic to and from Piedmont and a hub for road connections onwards to Nice and Liguria. Banaudo et al point us to a magazine published in 1899, which mentions a trial of a steam-powered road vehicle which it was hoped would provide a service to Nice and the coast until such time as a railway was built. [1: p40][37] The service was a trial organised by the House of Ascenso et Cie, and ran from Vievola to Ventimiglia. The journey, lasted a total of six hours, including a 43-kilometre climb. The vehicles used were Scotte trains. The car wagon carries a 27-horsepower engine and seated 14 passengers; it also towed a second 24-seater wagon. [1: p40][38]
“Due to their slowness, the difficulties of driving cars on the narrow roads of the time and the damage caused to the cobbled and cylindered roads,” [1: p40] the ‘Trains Scotte’ were not a success, they probably did not circulate for more than a few months or weeks. ….
The next length of the line can be found here. [46]
RM 3201-3519 (FS 215)Locomotives
Banaudo et al tell us that throughout the 19th century and on into the 20th century passenger stock and freight wagons were unchanged. Improved 0-6-0 tender locomotives came available as they were delivered by the Breda and Mavag companies, these were more powerful and faster locomotives than the RM Nos. 3201 to 3519 (which became group 215.001 to 215.398 at the FS). They were given RM Nos. 3801-3868 (which became the FS 310 series).
Banaudo et al also comment that “genuine mountain locomotives made occasional appearances: these were 040s [ in UK annotation 0-8-0s] with a three-axle separate tender, series RM 4201 to 4487 (future series 420 FS), built from 1873 to 1905 based on an Austrian model by a dozen Italian, Belgian, German and Austro-Hungarian firms. These machines, reserved primarily for the main lines of the Alps and the Apennines, occasionally intervened on the Col de Tende line, during bridge tests for example. At this time, Cuneo still had no allocation of machines and those going up to Limone and Vievola were attached to the Torino depot and the Moretta shed, on the Cuneo Airasca line.” [1: p41]
An FS Class 420 locomotive. [41]
“In the early 1870s, the SFAI needed a locomotive suitable for heavy work on the most important mountain lines, such as the Giovi railway and the Turin-Modane railway, for which the 0-6-0 locomotives were becoming increasingly inadequate. The Ufficio d’Arte di Torino chose a 0-8-0 locomotive of the Wiener Neustädter Lokomotivfabrik (then known as “Sigl”), very similar to the Südbahn Class 35 a that it already produced.” [41][42: p190][43: p31]]
“The Class 420 was a typical long-boiler, inside-frame 0-8-0 locomotive of the era, that showed its Austrian derivation with its two-shutters smokebox door, and its outside Stephenson valve gear. The locomotives built before 1884 had the distinction of having curved foot plating over the wheels, while later units had straight foot plating and small splashers. Some of the locomotives were given a replacement boiler before 1914, but their performance remained mostly unchanged.” [41][43: p31]
“The first 60 locomotives were built by Sigl (from which they derived the nickname with which they were known for their whole career) for the SFAI. Production continued until 1890, from both foreign (such as Maffei) and Italian firms (such as Ansaldo and Breda), for a total of 189 locomotives; all these were divided in 1885 between the Rete Adriatica and the Rete Mediterranea. Building of further locomotives for the RM resumed in 1897, and continued until 1905, bringing the total of the Class to 293.” [41][42: p190-192]
References
Jose Banaudo, Michel Braun and Gerard de Santos; Les Trains du Col de Tende Volume 1: 1858-1928; FACS Patrimoine Ferroviaire, Les Editions du Cabri, 2018.
Jose Banaudo, Michel Braun and Gerard de Santos; Les Trains du Col de Tende Volume 2: 1929-1974; FACS Patrimoine Ferroviaire, Les Editions du Cabri, 2018.
Jose Banaudo, Michel Braun and Gerard de Santos; Les Trains du Col de Tende Volume 3: 1975-1986; FACS Patrimoine Ferroviaire, Les Editions du Cabri, 2018.
Structures on the french side of the border would, when built, compete with the dimensions of the Rivoira Viaduct. The Eboulis Viaduct is 270 metres long and the bridge at Saorge is 60 metres high. However, the combination of these two dimensions (length and height) makes Rivoira Viaduct the most imposing on the line.
Franco Collidà; 1845-1979: the Cuneo-Nice line year by year; in Rassegna – Quarterly magazine of the Cassa di Risparmio di Cuneo , No. 7, September 1979; p12-18.
Franco Collidà, Max Gallo & Aldo A. Mola; “Cuneo-Nizza: History of a Railway; , Cassa di Risparmio di Cuneo, Cuneo (CN), July 1982.
Industrialist Joanny Scotte, originally from Epernay in the Marne department, began his business in the mid-1880s producing steam-powered cars. From 1897, he offered road trains consisting of a tractor or a steam-powered car, pulling one or more trailers designed for the transport of passengers or goods. These vehicles travelled on roads using solid tyres. They never really went beyond the experimental stage due to their slowness, the difficulties of driving the vehicles on the narrow roads of the time and the damage caused to the cobbled and cylindered roads. [1: p40] Scotte road train services were reported in the last decade of the 19th century in the Île-de-France region (Fontainebleau, Pont-de-Neuilly, Courbevoie), in the Aube region (Arcis-sur-Aube – Brienne-le-Château), in the Manche region (Pont-l’Abbé-Picauville – Chef-du-Pont), in the Drôme region (Valence – Crest), and for military use. Scotte partnered with the Lyon-based car manufacturers Buire and Audibert-Lavirotte to produce some of its vehicles. [1: p41]
Giovanni Cornolò; Locomotive a vapore; in TuttoTreno (in Italian), May 2014.
P. M. Kalla-Bishop; Italian state railways steam locomotives: together with low-voltage direct current and three-phase motive power; Tourret, Abingdon, 1986.
The railway from Nice PLM Station to Tende was completed in 1928. It was long in the gestation and in construction. The story stretches back more than a century and a half. ‘Le Chemin de fer du Col de Tende’ is historically a significant local and international line. Its inverted Y-shaped layout and its crossing of international borders means that it is known by a number of different names:
in Nice it is known as the Nice – Coni Line;
generally in Italy it is officially Ferrovia Cuneo Ventimiglia
in the Piedmont city of Cuneo’s economic/political circles, sitting at the top of the inverted ‘Y’, it is often referred to as the Cuneo – Nizza line in recognition of good relations with the community of Nice.
Its story is a saga of significant technical achievement: gaining 1000 metres in height ; having a dozen tunnels longer than 1 kilometre (including those of the Col de Tende (8098 m), the Col de Braus (5939 m) and the Mont Grazian tunnel (3882 m), which are among the longest structures on the French and Italian networks); having four complete helical loops, several S-shaped loops and a multitude of bridges and viaducts (some of which, such as those of Scarassouï or Bévéra, are architecturally significant railway structures. Of a total route of 143.5 km, 6.5 km are on bridges or viaducts and over 60 km are in tunnels. This means that close to 42% of the journey along the line(s) is on or within structures.
The line warrants a comprehensive detailed treatment and Jose Banaudo, Michel Braun and Gerard de Santos have provided just such a work. The 3 volumes of their work cover three distinct periods in the life of the line:
Volume 1: 1858 until the completion of construction in 1928; [1]
Volume 2: 1929 through to 1974 [2]
Volume 3: 1975 to 1986. [3]
The line’s construction spanned over 40 years and as a result a variety of different structural techniques were used. The first length built in Italy in the 19th century has some substantial stone and brick structures. Later work on the length from Nice to Fontane which was built between the two world wars employs much lighter design techniques. Then even later, after sections of the line were destroyed in the second world war, prestressed concrete construction techniques were used in the rebuilding of the line. [1]
The history of the area through which the line has been built has been tumultuous. This meant that the process of developing the line was tortuous. It took more than 75 years for the line(s) to be completed and then after a few short years of operation, the lines usage was disturbed by the machinations of dictatorships and then the second world war literally destroyed the region. Post war recovery was slow but nowhere more so than the length of the line between Ventimiglia and Breil-sur-Roya which was not fully reopened until 35 years after the end of the second world war. [1]
The reopening of the line after the second world war was vital for the economic development of Piedmont, the Riviera dei Fiori, and the Côte d’Azur – between which there was no efficient road connection and where the difficult terrain favored rail access. [1]
The immediate area offered tremendous tourism potential, both the train itself and the region it served. Ski resorts became accessible, particularly Limone, excursion trains came from all over Europe. But, after just a few decades of development the approach of the 21st century saw increased bureaucracy, financial disputes between the increasing number of partners, contradictory regulations and increased journey times. The result was that the line’s value and existence was called into question and that too sparked further conflict. “Paradoxically, European unification, which should have fully promoted this symbolic communication route, marginalized it!” [1: p5]
In 2014, my wife and I stayed in the village of Saorge in the valley of La Roya for the first time. We had travelled by train from Nice to Tende in an earlier year. In 2014, we had a hire car and on one occasion we followed the old road to the Col de Tende. In subsequent years it was not possible to drive up the old road as works on the much more modern tunnel seemed to have blocked access to the old road. On a more recent visit, we stayed in Saorge a year after serious flooding had destroyed much infrastructure in the valley. Travel towards the Col de Tende from Tende was not possible.
Early attempts to create a route from Cuneo toTende
In 2014, we drove up a road which was constructed by le duc Charles-Emmanuel 1er de Savoie (Duke Charles Emmanuel 1st of Savoy). It seems that he constructed a road over the pass between 1592 and 1616. Of this road, Banaudo et al say that, “the northern road [up to the pass] has about twenty hairpin bends, while access from the south requires an extraordinary … sixty hairpin bends.” [1: p9]
Our hire car was a very small vehicle, but nonetheless needed some careful manoeuvring at each hairpin bend. Once at the top, we were able to walk quite a distance between the different forts that stood on the ridge.
Banuado et al, tell us that since that route was constructed, a series of attempts were made to tunnel from lower points on the pass. Attempts from the North were made: in 1612 (achieved just 75m of tunnel before being halted); in 1781 which was abandoned 3 years later (164m of tunnel was achieved). [1]
In 1784, a carriage managed to traverse the pass for the first time.
Banaudo et al. Tell us that “the public works engineer Deglioli submitted an initial report on 3rd June 1852, supported by the diplomat Francesco Sauli (1807-1893), on the extension of the Marseille-Var railway, then planned in France, to Nice, Ventimiglia, the Roya Valley, and Piedmont, namely Cuneo or Mondovì.” [1: p11]
In 1854, the first train of the Società della Ferrovia Torino Cuneo arrived in Cuneo from Turin (via Trofarello, Savigliano, and Fossano). The first terminus was built in the Cuneo suburb of “Madonna-dell’Olmo, on the left bank of the Stura below the city. Ten months later, the time required for the completion of the viaduct over the Stura, Cavour and the Minister of Public Works, Pietro Paleocapa (1788-1869), presided over the inauguration of the new Cuneo platform/station on 5th August 1855, established in a temporary location at Basse-di-San-Sebastiano. The permanent station would not be built until 1870 on the plateau preceding the confluence of the Stura and Gesso rivers.” [1: p11]
In 1856, “Victor Emmanuel II, King of Sardinia, Cyprus and Jerusalem, Duke of Savoy and Aosta, Prince of Piedmont, Count of Nice and Tende, visited [Nice and] personally promised [a] railway to the people of Nice and distributed a lithograph depicting him, ostentatiously bearing a map bearing the dedication ‘Ferrovia da Cuneo a Nizza. Ai Fedeli Nizzardi’. … The Minister of Public Works commissioned a Roman military engineer, Filippo Cerrotti (1819-1892), to conduct a more in-depth study. On 29th May 1856, Cerrotti submitted a preliminary design for a standard-gauge line from Cuneo, ascending the Gesso and Vermegnana valleys, crossing the Col de Tende through a 6.5 km tunnel accessible by inclined planes powered by hydraulic funiculars, to emerge in the Roya River, which it followed to Airole. From there, two tunnels successively would take it through the Bévéra Valley and then into the Latte Valley, through which it reached the coast, which it then followed to Menton, Monaco, and Nice.” [1: p11]
The Nicois authorities accepted the proposed scheme in September 1856, their counterparts in Cuneo quickly endorsed the plans in principle but asked that an alternative route via the Col des Fenestres and the Vésubie, be explored and that a modification to the initial proposal should be explored, specifically a locomotive-powered line without the use of inclined planes. The municipality of Nice then commissioned another survey of alternative routes by Louis Petit-Nispel, but proposals were rejected by the Ministry of Public Works on 4th March 1858. [1: p11, p14]
Nothing happened, so the Nice authorities sent a petition to the Sardinian parliament (16th July 1858) but the request got lost in the midst of political machinations which surrounded the cession of Savoy and the County of Nice to France which was eventually confirmed on 22nd April 1860.
“During his first visit to the new border department in September 1860, the French Emperor promised the people of Nice a rapid connection to Marseille and the rest of the country via the Paris-Lyon-Mediterranean Railway Company (PLM) line, whose construction was then well advanced beyond Toulon.” [1: p14]
Nice got its connection to Marseille by 18th October 1864, but hopes for a Nice to Cuneo link were overshadowed by the desire to have a direct link between Marseille and Turin via Sisteron, Gap, Briançon, the Col de l’Echelle, and Bardonecchia – a plan was eventually shelved (even though it was favoured by the French government and the PLM company) as a result of the deal-making associated with the Saint-Gothard line.
In the mid-1860s the Piedmontese railway network became part of the Società per le Ferrovie dell’Alta Italia (SFAI). Its focus became developing internal infrastructure in Italy, with the exception of a very large project … a 13.7 km (8.5 mile) long tunnel, carrying the Turin-Modane railway line under Mont Cenis, linking Bardonecchia in Italy to Modane in France under the Fréjus. [1: p17][8]
Despite this, economic and political groups in Cuneo remained committed to having a rail link and in 1868 proposed a joint commission of French and Italian engineers. The following year, “the provincial authorities granted a loan of 500,000 lire to the Lombard engineer Tommaso Agudio (1827-1893), who sought to develop the possibilities offered by funicular traction. He, in collaboration with the engineer Arnaud, recommended the construction of a narrow-gauge railway alongside the SS 20 national road, along its entire route from Cuneo to Ventimiglia. This hypothesis suggested curves with a radius of less than 50 m and gradients of 45 mm/m. The Tende Pass was to be crossed by the planned road tunnel, with two access ramps sloping at 87.5 mm/m, on which traction would be provided by a hydraulically counterweighted cable.” [1: p17]
His project was approved by the Italian parliament in 1862 but no progress was made on the French side of the border. The project failed and Tommaso Agudio moved on to other things, “experimenting with his cable traction system in 1874 in Lanslebourg, then by applying it in 1884 to the railway linking the Turin suburb of Sassi to the famous Basilica of Superga.” [1: p17]
With little progress being made on a rail link, road links became paramount, a commission chaired by the civil engineering inspector Sebastiano Grandis (1817-1892) renewed interest in 1870 in a road tunnel under the Col de Tende which Grandis imagined would obviate the need for a railway.
“Following the fall of the Empire, France and Italy were finally connected by rail, first through the Fréjus Tunnel, opened between Modane and Bardonecchia on 17th September 1871, and then through the Menton and Ventimiglia on the coast on 23rd February 1872. At the same time, traffic between Piedmont and the former County of Nice was growing at an encouraging pace: the Fontan customs post recorded an annual transit of 22,000 tons of goods and 76,447 head of cattle. Under these rather favorable conditions, Nice’s business community sought to revive discussions with a view to attracting to their port a share of the benefits of the upcoming opening of the Saint-Gothard line, whose traffic, they feared, would exclusively benefit Genoa via the Via Giovi, or Marseille in the event of the construction of the Col de l’Echelle route. In April 1871, a group of industrialists and politicians from the region, including the mayor of Nice, Auguste Raynaud (1829-1896) and his counterpart from Toulon, Vincent Allègre (1835-1899), founded a Syndicate for the Nice Cuneo Line with the support of the Alpes-Maritimes Chamber of Commerce. On 7th November, the municipal council sent a personal letter to Adolphe Thiers, the new President of the French Republic, to express the desire of the people of Nice to see this project, which had been on hold for some twenty years, realized. On 29th November, the syndicate appointed a study commission headed by engineer Joseph Durandy (1834-1912), … to establish contacts with interested Italian parties and determine the advantages and disadvantages of each proposed route.” [1: p19]
“In March 1872, the engineer Henry Lefèvre (1825-1877), a public works contractor and member of parliament for the Alpes-Maritimes, published an ambitious programme comprising two railway lines, Nice – Digne and Nice – Cuneo. They would run as a common trunk up the Var valley to the confluence of the Vésubie; from there, the branch towards Piedmont would follow this river to its source, crossing the Pagari pass under a 7000 m tunnel drilled at an altitude of 1300 m, to then reach Cuneo via the Gesso valley. The gradients would not exceed 35 mm/m, which would however require several reversals from Venanson, as well as the use of articulated Fairlie locomotives.” [1: p19][9]
Lefèvre’s project was based on poor maps and went through areas with a high risk of avalanches and heavy snowfall. Durandy suggested that a longer tunnel (almost 15km long) could be employed, Delestrac suggested following the undulations/contours on the left bank of the Vésubie as much as possible to reduce the number of engineering structures and limit the gradients to 25 mm/m.” [1: p19] Both these suggestions significantly increased the costs of Lefèvre’s 120 km project.
Other projects were proposed:
In 1872, Séraphin Piccon proposed a “103 km long narrow-gauge route, crossing the Col de Tende through a 5100 m tunnel at a height of 1150 m. Descending the valley of la Roya to Piena, reaching the Bévéra basin and Sospel through a 1300 m tunnel under the Col de Vèscavo, then heading up the Merlanson valley to pass under Mont Méras through a new tunnel leading to Peille, and thence to Nice through down the valley of the Paillon. Access to the Col de Tende would be via two inclined planes with inclinations of 40 to 85 mm/m totaling a length of 6100 m, while a 60 mm/m gradient over 4700 m would allow the line to gain altitude north of Peille.” [10] On these steep gradients, traction would be assisted by a rack or an auxiliary central rail (the Fell System). [11][1: p20]
Also in 1872, Baron A. Cachiardy de Montfleury of Breil submitted a renewed proposal to the Conseil General, based on the Narrow-Gauge route between Cuneo and Ventimiglia funicular sections developed by engineers Agudio and Arnaud. [12][1: p20]
Then in April 1873, Baron Marius de Vautheleret. presented a proposal for a narrow-gauge Cuneo-Ventimiglia line using the planned Col de Tende road tunnel, passing through Briga, then through a 13,000 m tunnel under the Marta peak and then along the Nervia valley to its mouth near Ventimiglia. This route aimed to simplify administrative procedures by bypassing French territory, even if it meant creating a costly underground tunnel to connect the Roya to the Nervia river valleys. Gradients would not exceed 35 mm/m except for 22 km on either side of the Col de Tende, where gradients of 38 to 40 mm/m would require the adoption of a rack or hydraulic funicular. [13][14][1: p20]
These last two projects were discarded, partly because they were narrow gauge and required steep gradients, neither of which would suit the anticipated important international traffic and partly because they only linked two Italian cities while passing through French territory and not serving Nice. Both the protagonists continued to push their case until the end of the 19th century.
The first project proposal by Piccon was also deemed incompatible with heavy traffic flows but in its favour was the intent to link the railway to Nice. The “Durandy Commission preferred this option, subject to significant technical adjustments, such as adopting the standard gauge and replacing the inclined planes with longer base tunnels. On this route, the syndicate hoped for annual freight traffic of 90,000 tons despite a higher cost per kilometre than the routes via the Tinée or the Careï, as well as a revival of passenger traffic.” [1: p20]
The PLM had little enthusiasm for the proposed line as their experience of lines in the Alps encountered technical difficulties and had profitability problems
In 1878, the Minister of Public Works, Charles de Freycinet (1828-1923), asked regional authorities to consider possible lines to become part of a network of secondary lines across the country. The Prefect of the Alpes-Maritimes submitted the line ‘from Nice to the Italian border’, running from Nice to Turin via the Paillon Valley, the Col de Nice, L’Escarène, the Col de Braus, Sospel, the Col de Brouis, Breil, the Roya Valley, and the Col de Tende. This route was registered No. 142 in the network in the law of 17th July 1879, where it appeared alongside the Nice – Digne via Saint-André and Nice – Draguignan via Grasse lines. [1: p21]
While the Cuneo-Nice line was a low priority for the national government in Italy, but Piedmont and Liguria did not give up, encouraged by the interest on the French side of the border. A number of different schemes were considered (from Baron de Vautheleret, Giacomo Pisani and Domenico Santelli).
Renewed interest at a national level led, in April 1876, the ‘conseil superieur des Travaux Publics’ approved the principle of a Cuneo – Ventimiglia railway, following the Roya along its entire course, including crossing French territory. The estimated cost for the 86 km on Italian soil was 38 million lire.
Two years later, while France was preparing its “Freycinet plan”, Italy had its ‘loi Baccarini’ (law 5002) which was passed in parliament on 25th July 1879 and included for a secondary line ‘from Cuneo to the sea’, “leaving all options open South of the Col de Tende so as not to prematurely offend any interests.” [1: p23]
By the end of July 1879, the process seemed well underway but no one allowed for the political machinations that would follow.
“The first disappointments emerged in France in 1880 during the budget debates, where the President of the Chamber of Deputies, Léon Gambetta (1838-1882), postponed the vote on construction funding. On 22nd July, the General Council of Bridges and Roads rejected an initial project, which included 30 mm/m gradients and 300 m radius curves, as too costly. In November 1881, the Ministry of War was even more categorical, formally opposing the extension of the railway beyond Sospel, and demanding that it serve the village of Lucéram from L’Escarène, the supply base for the defensive sector of L’Authion, Turini and Peïra-Cava. In this case, the line would have to adopt even more severe characteristics: 40 mm/m gradients, 150 m radius curves, switchbacks to cross the Col de Nice and helical loops to reach Lucéram…” [1: p24]
“In 1882, an important step towards opening up the Haute Roya region was taken with the commissioning of the Col de Tende road tunnel. … This structure, remarkable for its time, was designed for the movement of carts, horses, pedestrians and. cannons, because the defense of the Tenda and Briga area was a major concern for the Italian general staff! The journey now avoided the countless hairpin bends of the pass and the risk of snowstorms and avalanches.” [1: p24]
The Col de Tende Road Tunnel and the border between France and Italy. [17]
But while economic and emotional ties remained strong between Cuneo and Nice, they were weakening between Rome and Paris due to political, commercial, and colonial rivalries that would poison relations … for about fifteen years. The attitude of the city of Marseille was also difficult. The business community in Marseille was hostile to a new rail link between Nice and Italy. Fearing the expansion of the port of Nice at their expense. They lobbied against any possible expansion of the port of Nice, even to the extent of thwarting standard-gauge lines from Nice to Digne and Draguignan, ensuring that the lines were built to metre-gauge (with less transport capacity and obligatory double-handling of loads). [1: p24]
Locally, in Nice, some pushed for the line to be metre-gauge, thinking that might iron out the technical difficulties and strategic objections. [1: p24] Faced by the administrative impasse which stalled the project in France , the French Ministry of Public Works decided to close its Nice design office on 1st September 1887. Italy, however, worked unilaterally with the intention of opening up the Haute Roya without prejudging the continuation of the route towards France. [1: p24]
From 1882 until 1900 it was the Italians that took the initiative. A delegation from Cuneo secured 29.5 million lire from the Italian Minister of Public Works. The first length of the scheme received local approval on 25th March 1882. Work on site started in April 1882 on the length of the line from Cuneo to Vernante.
The present railway station in Cuneo dates from the late 1930s the older station is known as Cuneo Gesso Statzione. At the time of the building of the Line from Cuneo towards Nice and Ventimiglia, Cuneo’s railway station sat alongside the Gesso River across the town from the present station.
The original Cuneo Railway Station from which the line to Nice and Ventimiglia left in a southerly direction. This image was taken in 1903. It was shared on the Facebook Nel dipartimento della Stura – Cuneo – pagina. [19]This second photograph of Cuneo’s original railway station which was on the banks of the River Gesso shows both the station building and the bridge which carried the railway over the river. This image was taken in 1905. It was shared on the Facebook Nel dipartimento della Stura – Cuneo – pagina on 16th November 2017. [20]Although dated 6th October 1979 this postcard image originated in the early years of the 20th century. It shows the Cuneo Gesso Station as it was at the turn of the 20th century. The postcard was made to commemorate the reopening of the international railway line that connects the city of Cuneo with the city of Nice. This image was shared on the Facebook Ferrovia Internazionale Cuneo-Ventimiglia-Nizza page on 11th December 2017. [21]
The railway initially arrived from Turin, via Fossano. It came as far as Madonna dell’Olmo opposite Cuneo across the Sturia River on 16th October 1854 where a small building was built to serve as a temporary station. On 5th August 1855 the inaugural train from Cuneo left for Turin. In the same year the municipality built a bridge over the Sturia (at its own expense). After the construction of the bridge over the Stura, a second temporary station was built on an embankment in the San Sebastiano plain (where Giuseppe Garibaldi had arrived to visit his “Alpine Hunters” in 1856). Only in 1870 was a significant edifice completed which became Cuneo’s railway station. It was alongside the Gesso River and it was again built entirely at the town’s expense. [19]
The complete opening of the Cuneo-Ventimiglia line, which took place on 30th October 1928, caused significant logistical problems for both travellers and rolling stock at Cuneo station. The old depot, dating back to 1864, soon became insufficient to house the locomotives of the new line, [23: p41] a hastily built locomotive depot was provided (because of delays creating the new line and new railway station, and in the construction of the large mixed-use viaduct over the Stura di Demonte. [24][25]
The Locomotive Depot at Cuneo Gesso Station which was used until the new depot close to Cuneo Altipiano Railway Station was opened. The site was repurposed – it became a sawmill. This plan comes from From the December 1929 Technical Magazine of Italiane Ferrovie., It was shared on the Ferrovia Internazionale Cuneo-Ventimiglia-Nizza Facebook Group on 13th February 2024 by Francesco Ciarlini Koerner. [62]
The new depot was placed beyond the embankment of the road to Mondovì. A double track arched bridge took the tracks under the road. [26][27] On 7th November 1937[24] the new Cuneo Altipiano station was opened, located to the west of the city centre and connected to the new locomotive depot built on the right side of the Stura River. [24][25]
Cuneo Gesso quickly lost importance, remaining active only as a stopping point for the lines to Mondovì and Boves , the latter closed to traffic in 1960. [23: p55-57][25]
Near the station was the terminus of the Cuneo-Dronero, [28] Cuneo-Saluzzo [29] and Cuneo-Boves [30] tramways, active for different years between 1879 and 1948 [25][31: p120]. The Cuneo Boves line opened in 1903 and closed in 1935.
Ex Stazione Ferrovia Di Cuneo Gesso as it appears on Google Maps satellite imagery. he river is the Gesso Torrent and a modern concrete bridge now spans the river. The line heading South from the station originally served a temporary Locomotive depot but now serves the sawmill that replaced the depot. [Google Maps, July 2025]The old station buildings seen from the Southwest. The building is in use as a cafe/bar. Tracks remain in place beyond the building. [Google Streetview, May 2025]The bridge which now carries the railway over the River Gesso. [Google Streetview, 2022]
Construction of the new line started in 1882, it left the station to the South curving sharply to the left to cross the Gesso River on a 3-arch brick viaduct (each span was 24.8 metres) shared with the line from Cuneo to Mondovi which was under construction at the same time. [1: p25]
The line to Mondovi remains today, but no passenger trains use the line any longer. The line we are following from Cuneo to Vernante, left the line to Mondovi heading Southwest and passing through the villages of Boves and Fontanelle-di-Boves. Provision for freight and passengers was made at Boves, just for passengers at Fontanelle-di-Boves.
Preparing for this article, I found a document from 1904 which included the plans and profiles of the line on the Ferrovia Internazionale Cuneo-Ventimiglia-Nizza Facebook Group. It was shared as a series of photographs by Davide Franchini on 2nd March 2022.
The 1904 document cover. [47]The first plan shows the bridge crossing the River Gesso with the line heading for Nice and Ventimiglia bearing away from the line to Moldovi. [47]The line heading South. [47]The route of the old railway from Cuneo Gesso to Borgo-San-Dalmazzo, (c) Ale Sasso and licenced for reuse under a Creative Commons Licence, (CC BY-SA 4.0). [32]
As far as I can tell, the line to Boves has been built over. It seems to have followed the route of Via del Borgo Gesso South from the river bridge, then Via Bisalta, then Highway SP21 to Boves where the line curved back towards the River Gesso. Boves station was on a relatively sharp curve in the line. [33]
Boves Railway Station building. [35]A similar view of Boves Railway Station in the 21st century. [Google Streetview, June 2025]The altered station building as seen from the Southeast. [Google Streetview, 2012]The goods shed/warehouse seen from the East. [Google Streetview, 2012]The goods shed at Boves, seen from the West on the SP21. The original station building can be seen on the left of this image. [Google Streetview, June 2025]The location of Boves Railway Station in the 21st century. Via Gastalato (SP21) runs along the old railway line. The main station building has a silver coloured roof and sits at the centre of this satellite image. The goods warehouse costs to the West of the main station building and has a red roof. [Google Maps, July 2025]
Boves station had a passing loop and two sidings. The passenger building, converted into residential housing several years ago, was adjacent to a goods warehouse, now used as a provincial warehouse. [35]
Boyes Railway Station plan. [47]The line beyond Boves Railway Station ran through Fontanelle di Boves and then crossed the River Gesso again. [47]
The hamlet of Fontanelle di Boves was just a short distance beyond Boves Railway Station. It had its own passenger station which opened in 1942 after the line from here back to Cuneo was replaced by a new line on the other side of the River Gesso which ran into the new station at Cuneo. Just a short distance further down the line was the viaduct which took the line back over the River Gesso. Originally, this was a masonry structure of three 24.8 metre arched spans. [1: p25] The viaduct was overwhelmed and destroyed by a flood of the Gesso on the afternoon of 2nd October 1898. It was then replaced with the current 74 m metal truss girder bridge. [34]
This photograph shows the immediate aftermath of the destruction of the bridge between Fontanelle-di-Boves and Borgo San-Dalmazzo. It was shared on the Ferrovia Internazionale Cuneo-Ventimiglia-Nizza Facebook Group by Mario Zauli on 29th February 2024. As well as appearing on the Facebook Group, Banaudo et al include the picture in their book. They comment: “On 2nd October 1898, the Southern Alps suffered violent floods that swept away the three-arched masonry viaduct over the Gesso between Boves and Borgo San-Dalmazzo, built in 1883. It was rebuilt as a metal truss bridge, but initially trains used a temporary structure on wooden beams. In December 1898, this was tested by the passage of locomotive No. 4333 of type 040, series 4201 to 4493 of the Rete Mediterranea. (Photo Giacinto Garaffi – Diego Garel collection).” [37][1: p26]
The bridge is known as Ponte di Sant’Andrea, a second truss was positioned alongside the railway bridge and together the two bridges now carry the SP21.
After crossing the River Gesso and at about 12 km from Cuneo the line arrived at Borgo-San-Dalmazzo.
This schematic map shows the two rail routes. The solid line shows the original alignment that we have just been following. The dotted line shows the route built at the end of the 1930s. The two lines met to the West of Pont Sant’Andre. The 1937-built station is on the banks of the Stura River on the West side of Cuneo and on the dotted line. [34]The bridge (Ponte di Sant’Andrea) is flagged in the bottom-right, the newer line from Cuneo enters this image middle-top and runs down to the bottom-left. The older line curved round from the SP21 and its route is marked by the curved field boundary. [Google Maps, July 2025]
Returning to the 1937-built Cuneo Railway Station, the line from that station leaves Cuneo in a South-southwest direction. It is easiest to see the route of the line on a sequence of extracts from global mapping provided by OpenStreetMap. …
Cuneo’s Railway Station in the 21st century. [OpenStreetMap, July 2025][38]
A twilight view of Cuneo railway station taken from the cab of a multiple unit entering the station from the Southwest. [45]
The line runs alongside the locomotive depot to the South of the passenger facilities at Cuneo Railway Station and then enters a tunnel which turns South under the city. [39]The tunnel mouth to the South of Cuneo Railway Station can be made out at the centre-top of this image. [Google Earth 3D, July 2025]This time looking North, the Southern portal of the tunnel to the South of Cuneo Railway Station can be made out below the roundabout at the centre-top of this image. [Google Earth 3D, July 2025]
A rain-spattered cab view from the South, taken in the late evening, of the Southern portal of the tunnel which sits to the South of Cuneo Railway Station. [45]
After leaving the tunnel, the line began to curve round to the Southwest passing under Via Fontanelle and then under the roundabout at the junction of Via Mellana and Viale Federico Mistral. [40]
Looking North in the evening light under a footbridge close to Via Giuseppe Scagliosi through the cab widow of a multiple unit on the line. [45]
The view North from the bridge carrying Via Fontanelle across the line. [Google Streetview, 2019]Looking South from the bridge carrying Via Fontanelle over the line. The bridge in the distance sits underneath a roundabout at the junction between Via Mellana and Viale Federico Mistral. [Google Streetview, 2019]
A three arch bridge carries Via Fontanelle over the railway, seen again in the evening light from the South through the rail-spattered cab widow of a multiple unit. [45]
A short tunnel carries the roundabout at the meeting of Via Mellana and Viale Federico Mistral over the railway, seen again from the South through the rail-spattered cab widow of a multiple unit. [45]
Vegetation around the roundabout means that it it not possible to see into the cutting from the road.
The line continues in a Southwesterly direction running alongside Viale Federico Mistral. [41]
A brick-ringed arch bridge carries the railway over a side road off Viale Federico Mistral. This view is from the Southeast. The structure is at the top-right of the map extract immediately above. [Google Streetview, June 2025]
A very similar arch bridge carries the railway over a further side road off Viale Federico Mistral. The bridge is located in the bottom-left quadrant of the map extract above. [Google Streetview, June 2025]
Now on a more Southwesterly course the line passes under a footbridge, obscured on the map extract by the words Tetto Bidetti in the top-right corne of the extract.
Silhouetted in the evening light, this bridge crosses the line carrying a footpath over the railway. The image, again comes from the cab of a multiple unit heading for Cuneo. [45]
Close to Cascina Tallone, the line crosses Lungo Gesso by means of another brick ringed arch. This view looks under the railway from the Southeast. [Google Streetview, May 2022]
Near Cascina David another brick-arched bridge pierces the railway embankment where Via David passes beneath the railway. Again this view is from the Southeast on Via Sant’Andre. [Google Streetview, May 2022]
Near Cascina Landra another brick-arched bridge pierces the railway embankment. Again this view is from the Southeast on Via Sant’Andre. Thestructure appeasr bottom-left on the map extract above and top-right on the extract below. [Google Streetview, May 2022]
And close to where the line of the older route meets the newer route the line is heading South-southwest and turns towards the Southwest. [43]Now in Borgo San-Dalmazzo we have reached the point where the older line curved in from the East having crossed the River Gesso. [44]
Via Sant’Andrea passes over the line. This view looks Northeast towards Cuneo. [Google Streetview, May 2022]
Also taken from the bridge carrying Via Sant’Andrea over the railway, this view looks across the road SP21 towards Borgo San-Dalmazzo. [Goog;e Streetview, May 2022]
The view Southwest from the bridge carrying the SP21 over the railway. The route of the older line is marked by the field boundary visible to the left of the line. [Google Streetview, June 2025]
The older line curved round to the Southwest and followed a straight course towards Borgo-San-Dalmazzo Railway Station. The newer line has taken its place on the approach to the Station from the Northeast.
San-Dalmazzo is a very old trading town located at the crossroads of three valleys: the Stura, the Gesso and the Vermenagna. The station had three platforms, a goods yard, a 5.50 m turntable and a large overflow yard that could be used for the embarkation and disembarkation of military units deployed in the area. “When the railway arrived in Borgo-San-Dalmazzo, this small town had already had a rail service for several years. In fact, private entrepreneurs Ercole Belloli and Carlo Chiapello opened a 1.445 m gauge horse-drawn tramway between Cuneo and Borgo in 1877, passing through the San-Rocco-Castagnaretta district on the left bank of the Gesso. Horse-drawn traction was replaced by steam locomotives on this modest 8-km line in 1878.” [1: p27][48]
The Cuneo-Borgo San-Dalmazzo-Demonte tramway linked the cities of Cuneo, Borgo San Dalmazzo and Demonte from 1877 to 1948. In the late 1870s, following the success of similar initiatives in the Turin area, the construction of tramways was pursued in the province of Cuneo. [48] As we have already noted, this was just one of a number of such tramways in the area.
The Cuneo Borgo-San-Dalmazzo tramway was extended in 1914 to Demonte (26.4 km) and converted on this occasion to a 1.10 m gauge to facilitate the exchange of goods with the Compagnia Generale dei Tramways Piemontesi (CGTP) which operated the Cuneo Boves line (8.3 km) from 1903. The Boves steam tramway disappeared in 1935 and that of Borgo and Demonte in 1948. [1: p28] The story of these tramways seems worth investigating, but their histories are a matter for a different article!
The station had an ignominious place in history. During the Second World War two convoys of Jewish deportees departed from the Borgo San Dalmazzo railway station bound for Auschwitz , coming from the adjacent Borgo San Dalmazzo concentration camp. The first convoy, on 21st November 1943, completed its journey via Nizza Drancy with 329 people on board. Only 19 survived. The second convoy, on 15th February 1944, with 29 people on board, headed instead for the Fossoli transit camp where it was combined with transport no. 8 bound for Germany. Only 2 survived. [49][50]
The Deportation Memorial , with a row of cattle wagons similar to those used then (the wagons are from 1953) commemorates the names of the deportees, their age and nationality and their family relationships. [50][51]
Burgo San-Dalmazzo to Robilante: The second construction contract covered the length from Borgo San-Dalmazzo to Robilante. Work began in late 1883. From Burgo San-Dalmazzo the line leaves the plain and begins its ascent up the Vermenagna Valley, heading towards the Tende Pass. The route, was designed to accommodate heavy traffic, so the line does “not include any curves with a radius less than 300 m, with two exceptions: one at the southern end of Cuneo station and one at the exit from Borgo station, where the route curves sharply to the left in a 257-meter curve to reach the left bank of the Gesso River. There, a 21 m three-arched masonry viaduct, shared by the railway and the SS20 road, crosses this Alpine torrent for the third and final time.” [1: p27]
This satellite image shows the sharp curve from the Railway Station at Borgo San-Dalmazzo to the viaduct across the River Gesso. [Google Maps, July 2025]
As the railway curves round towards the river its embankments are pierced twice to allow local roads to pass beneath the line.
This is the first structure. [Google Streetview, June 2025]The second structure, closer to the River Gesso. [Google Streetview, June 2025]
The southern approach to Borgo San-Dalmazzo Railway Station, seen from the cab of a multiple unit. The line to the right of the image is a siding which terminates close to the River Gesso. [45]
The 3-span viaduct across the River Gesso carries both the railway and the SS20. [Google Streetview, June 2025]Looking South along the SS20 as it crosses the Gesso. The railway cantenary is on the left with the tracks hidden behind the dividing fence. [Google Streetview, June 2025]
A view from the South showing the road on the left. This is a view from the cab of the multiple unit again. [45]
Once over the river the road and railway remain at a high level with an access road to the SS20 passing under both the railway and the road. [Google Streetview, June 2025]
After crossing the river the line ran on through Roccavione. …
Roccavione Station is a simple station with two public platforms and one track serving a military platform. Another level crossing sits beyond the South end of the station site.
Looking back from the level-crossing at Via Piano Sottano towards Roccavione Railway Station. [Google Streetview, June 2025]
A similar view looking North into Roccavione Railway Station from the cab of the multiple unit. The station has no passing loop. [45]
The view Southwest across the level-crossing. [Google Streetview, June 2025]Looking South-southeast as the line continues up the Vermenagna valley. [Google Streetview, June 2025]
The line follows an easy gradient between the SP259 (which used to be the SS20) and the left bank of the River Vermenagna to Robilante Railway Station. [1: p27]
The line runs Northwest to Southeast across this extract from Google Maps satellite imagery. It runs close to the SP259 between Roccavione and Robilante. [Google Maps, July 2025]A link road under the railway and under the SP259 beyond. It provides access from Via Piano Sottano to the SP259. [Google Streetview, June 2025]What in the UK we might choose to call an underpass or cattle-creep under the railway. Apologies for the slight distortion of the image which comes from the way in which Google’s algorithm merges the 360° camera photographs. [Google Streetview, June 2025]
Robilante Railway Station had three platform tracks, a small goods yard, a water feed, a 8.50m turntable and an engine shed. Beyond the station track gradients increased significantly and provision needed to be made for banking engines in steam days. [1: p27]
Robilante Railway Station. [Google Maps, July 2025]The station building sat on the Southwest side of the line, This view looks through the station to the Southeast, (c) Gum Gum. [Google Maps: July 2023]Robilante Station building and forecourt seen from the Northeast on Via Roma. [Google Streetview, June 2025]This view looks Northwest through the station towards Cuneo, (c) Mattia Vigano. [Google Maps: April 2019]
A similar view to that immediately above but taken from the driver’s cab on a multiple unit. In the distance in this image the old goods shed can be seen to the left of the line. The shed is no longer present in the more modern image above. [45]
A station plan for the station at Robilante. The line is oriented Northwest to Southeast. The turntable is located at the Northwest end of the yard. The engine shed is opposite the passenger building. The goods shed was Northwest of the passenger facilities and is shown here with a single siding passing through the building. The bridge, shown in images below crosses the station throat at the Southeast end of the station site. Not shown on this early plan are five sidings added for clinker wagons from the Buzzi Unicem cement plant nearby. [47][53]
Robilante Goods Shed seem from the cab of a multiple unit. As noted above, the shed has now been demolished. [45]
This image taken from the Southeast of the station from the cab of an approaching Cuneo service gives a broader view of the station site. [45]
A broader view of Robilante Station taken from a road at the Southeast corner of the station site. [Google Streetview, June 2025]The road overbridge at the Southeast end of the station site. [Google Streetview, June 2025]The view Northwest from the road bridge which carries Via Luigi Emina over the line. [Google Streetview, June 2025]The view Southeast from the same road bridge. [Google Streetview, June 2025]The bridge which carries Via Luigi Emina over the line, seen from the Southeast. [Google Streetview, June 2025]
The second phase of the construction work on the line terminated in Robilante. “The preliminary design for the third phase from Robilante to Vernante was submitted to the Ministry of Public Works on 11th January 1884, and work began the following summer. On this 6,419-meter-long section, the railway crosses the mountain with gradients of 25 mm/m.” [1: p27]
This extract from Google Maps satellite imagery shows the length of the line from Robilante to VernanteVia Ferrovieri runs immediately adjacent to the railway for some distance, passing under a road over bridge along with the railway. This view looks ahead up the Vermenaga valley. [Google Streetview, June 2025]The same bridge seen from the Southeast. [Google Streetview, June 2025]
The two images immediately above were taken at the end of a road serving a small industrial area. The first looks Northeast, the second, Southeast. [Google Streetview, September 2023]
After passing under the SS20, the line runs alongside the road for a kilometre or so.
Trains can be seen passing immediately adjacent to the road. [Google Streetview, June 2025]
A short distance further South a side road from the SS20, Via Tetto Pettavino, bridges the line. The two photographs below were taken from the bridge.
Looking North towards Robilante. [Google Streetview, June 2025]Looking ahead along the line towards the viaduct over the River Vermenagna. [Google Streetview, June 2025]The railway crossed the Vermenagna River by means of a viaduct of 5 arched spans – three of 21 metres flanked at each end by an 8 metre span. [Google Maps, July 2025][1: p27]
A photograph of the viaduct over the Vermenagna surrounded by trees can be found here on Flickr. [54]
Banaudo et al tell us that seven further significant structures were included in the contract which covered the line as far as Vernante [1: p27] all of which sit within approximately 3 kilometres along the line:
the Rio Vermanera masonry viaduct, with three 8-metre arches;
the Ponte Nuovo Tunnel, 425 metres long;
the Brunet Tunnel, 161 metres long;
the Corte-Soprano Tunnel, 95 metres long;
the San Giovanni masonry viaduct, with six arches measuring 7.90 m, three measuring 13.75 m, and one measuring 6 m;
the San Giovanni Tunnel, 138 metres long; and
the Costa Tunnel, 147 metres long. [1: p27]
The first of these – the Rio Vermanera Viaduct is pictured below.
The Rio Vermanera masonry viaduct, seen from the West, one span of which crosses the Strada Vermanera, another spans the Vermaners stream. [Google Streetview, June 2025]
The same viaduct seen from the East. [Google Streetview, June 2025]
Strada Vermanera provides road access to a number of small hamlets to the East of the railway line. [Google Maps, July 2025]
The Ponte Nuovo Tunnel: this extract from OpenStreetMap shows the tunnel curving significantly. It ran from just to the South of the Rio Vermanera Viaduct to open out immediately adjacent to the SS20/E74 but at a higher level. [55]
Immediately beyond the southern portal of the Ponte Nuovo Tunnel, a masonry retaining wall supports the railway above the SS20/E74.
Looking back towards the South portal of the Ponte Nuevo Tunnel the parapet railings of the retaining wall can be seen on the left of this image. [45]
The southern portal of the Ponte Nuovo Tunnel is at the far end of this retaining wall. Immediately at the Southeast end of the retaining wall is the short Brunet Tunnel (161 metres long) [Google Streetview, June 2025]
The Brunet Tunnel is shown dotted on this extract from OpenStreetMap. [56]
The South Portal of the Brunet Tunnel. [45]
The next tunnel is only 200 metres or so along the line, the Corte-Soprano Tunnel is even shorter at only 95 metres in length. [57]
The South Portal of the Corte-Soprano Tunnel. [45]
Just to the Southeast of the tunnel portal is the next structure, the San Giovanni Viaduct. masonry viaduct, with six arches measuring 7.90 metres, three measuring 13.75 metres, and one measuring 6 metres. [Google Maps, July 2025]
It is not feasible to get a photograph of the full length of the viaduct. The three images below give a good impression of its length and height.
Two further short tunnels, the San Giovanni Tunnel (138 metres long) and the Costa Tunnel (147 metres long) follow in the next few hundred metres.
The two tunnels are only separated by a short length of the line. [Google Maps, July 2025]
The South portal of the San Giovanni Tunnel. [45]
The South portal of the Costa Tunnel. [45]
The railway continues to climb higher on the eastern slope of the Vermenagna Valley and reaches Vernante, about 23 km from Cuneo.
Another of the plans and profiles that we encountered earlier in this article. This one shows the final approaches to Vernante Railway Station. Some of the structures described above can be seen on this plan. [47]
On the final approaches to Vernante Railway Station two further structures can be seen on the plan above. They carry the line over minor roads. The first spans Via La Tina, the second spans Vicolo Castello/Strada da Castello.
Looking East through the underpass which takes Via La Tina under the railway. [Google Streetview, June 2025]Looking East through the structure that carries the railway over Strada da Castello. [Google Streetview, June 2025]
Vernante Railway Station was the end of the third tranche of works on the railway. Vernante is “a busy centre of livestock breeding and craftsmanship where renowned knives are produced. Vernante station … has two platform faces with a passing loop, … [a goods shed] and platform for goods traffic, a 5.50 m turntable and a curious installation, unique on the line, the “binario di salvamento”. This is a counter-slope safety [line which leaves the main running line close to the station throat] on the Limone side. The switch is permanently positioned to provide access to the safety line, so that any vehicle drifting down the 26 mm/m gradient south of the station can enter it, be slowed down by the opposite gradient and then come to a stop. Each descending train must stop before the switch, so that it can be maneuvered on site to allow normal entry into the station. This simple but effective precautionary measure applies to other steep-gradient lines on the Italian network, in the Alps and the Apennines.” [1: p27]
A plan of Vernante Railway Station. [47]Vernante Railway Station. [Google Streetview, July 2025]The view Southeast from the station car park, after demolition of the old goods shed. The main station building features at the centre of the image. [Google Streetview, June 2025]The main station building at Vernante seen from the West. [Google Streetview, June 2025]
Photographs showing the station building and the goods shed prior to its demolition can be seen here. [58] “Inaugurated in 1889, the station served as the terminus for the Cuneo-Ventimiglia line for nearly two years, until it was extended to Limone Piemonte. The passenger building features classic Italian architecture, with two levels. It is square, medium-sized, and well-maintained. Its distinctive feature is the two murals depicting scenes from the Pinocchio fairy tale, adorning its façade. The lower level houses the waiting room and self-service ticket machine, while the upper level is closed.” [58]
A photograph from the cab of a Cuneo-bound train arriving at Vernante. The passenger building is on the left with the goods shed beyond. [45]
While construction work was underway on the first three tranches (Cuneo to Vernante), the Italian rail network was undergoing a major reorganization. The Law passed on 27th April 1885, placed control of the railways into the hands of “the new Società per le Strade Ferrate del Mediterraneo, more commonly known as Rete Mediterranea (RM), … including the route ‘from Cuneo to the sea’.” [1: p28]
In 1887, the time had come for the first trains! “The Cuneo-Robilante section was inaugurated on Saturday, 16th July 1887, and opened for service on Monday 18th. Less than two weeks later, Francesco Crispi became President of the Council of Ministers, and relations between Italy and France would soon be strengthened. Then came the beginning of the future Cuneo-Mondovi line, which opened on 2nd October 1887, as far as Roccadebaldi. The Roccadebaldi and Robilante lines thus formed a common section for 359 meters, starting from Cuneo [Gesso] station and crossing the Gesso River on the same viaduct. … Two years later, the Robilante-Vernante section was … opened on 1st September 1889.” [1: p28]
As footnotes to this article we note that:
Banaudo et al comment: “construction of the Ceva Ormea branch line began in the upper Tanaro Valley. With a terminus about 30 km from Vernante or 25 km from Tenda and Briga, this line would play an important role in the battle of interests that would unfold in the final years of the century to confirm a definitive route to the sea.” [1: p28]
They also give details of the locomotives used on the line in these very early years, by Rete Mediterranea (RM). The locomotives were 030s (in the UK 0-6-0s) with tenders and came from the roster of the Turin depot and loaned to the Cuneo-Gesso Locomotive depot. They belonged to just one series: “Nos. 3201 to 3519 RM, which became group 215.001 to 398 at the FS. [The series was built] between 1864 and 1892 based on a model derived from the French “Bourbonnais” locomotives of the PLM. These 450 hp engines were equipped with saturated steam, single expansion, and Stephenson internal distribution. The [later] Cuneo depot, established in 1907, still had five type 215 locomotives in 1922, mainly operating service trains.” [1: p86] It is also worth noting that some of the locos used on the line after 1899 came from a second series of locomotives (“Nos. 3801 to 3869 RM, later 3101 to 3169, then group 310.001 to 069 at the FS, built from 1894 to 1901 [1: p86]). While these locomotives were old enough to have served in the period from 1887 to 1891, they only arrived on the line during 1901. … I anticipate there being a separate article about motive power on the line in due course.
We finish this first part of the journey from Cuneo to the sea at Vernante. The next article about the line will begin at Vernante and head South towards Limone and Vievola. It can be found here. [61]
References
Jose Banaudo, Michel Braun and Gerard de Santos; Les Trains du Col de Tende Volume 1: 1858-1928; FACS Patrimoine Ferroviaire, Les Editions du Cabri, 2018.
Jose Banaudo, Michel Braun and Gerard de Santos; Les Trains du Col de Tende Volume 2: 1929-1974; FACS Patrimoine Ferroviaire, Les Editions du Cabri, 2018.
Jose Banaudo, Michel Braun and Gerard de Santos; Les Trains du Col de Tende Volume 3: 1975-1986; FACS Patrimoine Ferroviaire, Les Editions du Cabri, 2018.
“The locomotive developed by the Scottish engineer Robert Francis Fairlie (1831-1885) from 1869 on the Ffestiniog narrow gauge railway in Wales, had two boilers connected by a single central firebox. Each boiler supplies steam to a pair of cylinders driving an independent group of axles. This system was developed in France from 1888 by artillery captain Prosper Péchot (1849-1928) and engineer Charles Bourdon (1847-1933), creators of an articulated narrow gauge locomotive widely used by the French army.” [1: p21]
Séraphin Piccon; Etude Comparative de Deux Lignes de Chemin de Fer Entre Nice et Coni; 1872.
The Fell System which created “additional adhesion using a raised central rail, patented by British engineer John Barraclough Fell (1815-1902), was first applied in the Alps in 1868 on the railway running along the Mont Cenis route between St. Michel-de-Maurienne and Susa, pending the completion of the Fréjus Tunnel in 1871.” [1: p21]
A. Cachiardy de Montfleury; Chemin de Fer de Nice a Coni; Imprimerie Cauvan, Nice, 1872.
Marius de Vautheleret; Chemin de Fer Cuneo Ventimiglia – Nice Traversant le Col de Tende; Editions Giletta, Nice, 1874.
Marius de Vautheleret; Chemin de Fer Cuneo – Nice par Ventimiglia et le Col de Tende; Kugelmann, Paris, 1883; Trajet direct de Londres à Brindisi par le Col de Tende; Kugelmann, Paris, 1884; Ligne directe Londres – Brindisi par le Col de Tende; Retaux, Abbeville, 1890; Le Grand Saint-Bernard et le Col de Tende Ligne Ferrée Directe de Londres à Brindisi avec Jonction à la Méditerranée; Malvano & Mignon, Nice, 1897.
Stefano Garzaro & Nico Molino; La Ferrovia Di Tends Da Cuneoba Nizza, L’ultima Grande Traversata Alpina, Colleferro (RM); E.S.T. – Editrice di Storia dei Trasporti, Luglio, 1982, (Italian text)
Ferrovie dello Stato; Circolare Compartimentale del Compartimento di Torino 54/1937, (Italian text).
Franco Collidà, Max Gallo & Aldo A. Mola; Cuneo-Nizza: Storia di una ferrovia, Cuneo (CN); Cassa di Risparmio di Cuneo, Luglio, 1982, (Italian text).
The locomotive depot area, left vacant after the opening of the new Cuneo station, was later reused by a sawmill connected by a siding to the Gesso station. [25]
Manchester Mayfield 3rd September 1955. Longsight’s Stanier 2-6-0 42960 is ready to depart with a suburban service. Photo H C Casserley.
A short note about extensive alterations at Manchester London Road Station appeared in the December 1958 issue of The Railway Magazine. The major alterations were designed to accommodate the electrification of the line between Manchester and Crewe. [1]
The Railway Magazine reported that “The improvements include[d] the construction of three new platforms, the lengthening of the existing platforms, to accommodate 16-coach electric trains, and the widening of the concourse. The station [would] thus have 14 platforms, of which ten [would be devoted to main-line and local traffic on the former London & North Western line, and the remainder to trains on the Great Central route. When the alterations [were] completed, the adjoining terminus at Mayfield [would] cease to deal with passenger traffic. A new power signalbox [would] control the area extending to East Didsbury and Heaton Chapel, and will replace 13 manual boxes. Electric trains [would] not be an innovation at London Road, because the Altrincham line was electrified in 1931, and the Sheffield line in 1954.” [1]
The text in bold highlights the closure of Mayfield Station to passenger traffic. This article focuses on Mayfield Railway Station. ….
Mayfield Station had only ever been something of which I was vaguely aware despite having lived in the Manchester area for large parts of my life.
Manchester Mayfield Station was, “on the south side of Fairfield Street next to Manchester Piccadilly station, [Manchester London Road station, as it was in 1958]. Opened in 1910, Mayfield was constructed as a four-platform relief station adjacent to Piccadilly to alleviate overcrowding. In 1960, the station was closed to passengers and, in 1986, it was permanently closed to all services having seen further use as a parcels depot.” [2]
“Opened on 8th August 1910 by the London and North Western Railway, Manchester Mayfield was built alongside Manchester London Road station (later Piccadilly) to handle the increased number of trains and passengers following the opening of the Styal Line in 1909. [4][5: p7] The LNWR had considered constructing a new platform at London Road between the [Manchester, South Junction and Altrincham Railway’s] MSJAR’s platforms 1 and 2, which were renumbered 1 and 3 in anticipation, but this was abandoned in favour of the construction of Mayfield; the platforms nevertheless remained renumbered. [6: p167] Four platforms were provided and passengers could reach London Road via a high-level footbridge. [6: p167][7: p43] Mayfield suffered the effects of bombing during World War II, when it was hit by a parachute mine on 22nd December 1940.” [8: illustration 40] [2]
Manchester Mayfield Railway Station as it appears on the 25″ Ordnance Survey of 1914. [21]
Mayfield was a relief station, mainly used by extra trains and suburban services to the south of Manchester [6: p167] – places such as Cheadle Hulme, Buxton, Alderley Edge, Chelford and Stockport. [9: table 97] “In the London Midland timetable of September 1951, the Pines Express from Bournemouth West is shown as arriving at Mayfield at 4.30pm (16.30) on Mondays to Fridays. On Saturdays, this train used Piccadilly station, then known as London Road. [10: table 17] In the 1957-8 timetable, the Pines Express still arrived at Mayfield on Mondays to Fridays, now at the time of 4.45pm (16.45).” [11: table 21][2]
For a brief period during the electrification and modernisation of London Road station, Mayfield Station was the Manchester terminus for many diverted services. [12: p86-87] It was closed to passengers on 28th August 1960 with the completion of the electrification and modernisation works at Manchester London Road station. [13: p92]
“The site was converted into a parcels depot, which opened on 6th July 1970. [4] Royal Mail constructed a sorting office on the opposite side of the main line and connected it to Mayfield with an overhead conveyor bridge, which crossed the throat of Piccadilly station.” [2]
The depot closed in 1986, following the decision by Parcelforce, Royal Mail’s parcels division, to abandon rail transport in favour of road haulage. The tracks into Mayfield were removed in 1989, as part of the remodelling of the Piccadilly station layout. The parcels conveyor bridge was removed in 2003 with the Sorting Office being rebuilt as the Square One development, prestige offices used by Network Rail. [2]
The site of Mayfield station is the property of London and Continental Railways. [2] The interior of the station was used in Prime Suspect as a drug dealer’s haunt. [4] It was also used as a double for Sheffield railway station in The Last Train. The roadside building was gutted by a fire in 2005. [4]
There are, or have been, various plans for the use of the site of Mayfield station. These include:
Reopening as a station
“A study was carried out by Mott MacDonald in 2000, which looked at possibilities of increasing capacity at the Piccadilly station. One solution put forward would see the track quadrupled between Slade Lane Junction and Piccadilly, with a pair of through platforms in the Mayfield goods yard to the south of Piccadilly’s platforms 13 and 14 linked to additional running lines to Ashburys station. This proposal was supported by the Greater Manchester Passenger Transport Executive as it would increase usable train paths through Piccadilly by between 33% and 50%; the extra track would, however, require an expensive extension to the Piccadilly – Deansgate viaduct carrying the track from Slade Lane. The location of the proposed platforms was also criticised, as it would entail ‘a long walk for passengers wishing to interchange with other terminating rail services at Manchester Piccadilly or access the city centre’.” [2]
Other options would have the station used again as a terminus, providing a rail link to Manchester Airport or, alternatively, the lines might be extended through Mayfield and connected to the existing line to Manchester Oxford Road railway station. [2][4]
“Further proposals were put forward in 2009 by the Greater Manchester Integrated Transport Authority for reinstating Mayfield as an operational station, to alleviate capacity problems at Piccadilly Station. [14] However, as part of the Northern Hub railway development scheme across Northern England, Network Rail now plans to increase capacity on the existing Oxford Road-Piccadilly route by widening the viaducts and adding two additional platforms (15 and 16) to the south side of Piccadilly station. [15] There are no plans to re-open Mayfield station for public transport.” [2]
Commercial redevelopment –
In 2008, an alternative scheme involving Manchester Mayfield was put forward. This proposal would see the station as part of a new 30-acre (120,000 m2) city centre district immediately adjacent to Piccadilly Station. That project would have created more than 6,000,000 square feet (560,000 m2) of offices contained in office blocks up to 12 storeys high, and would be completed over a period of 15 years. The scheme was led by “Mayfield Manchester”, a joint venture company between Ringset, part of the Wrather Group, and Panamint; the company owns around 90% of the land around the station as of 2008, but do not own the station itself. In April 2008,Manchester Mayfield were said in talks with its owners of the station site, BRB Residuary. [2]
Other schemes were also under consideration:
Conversion into a Coach station by National Express to replace their Charlton Street facility [2]
Government Offices – in May 2009, the site was earmarked for a development which would have housed 5,000 civil servants. It would have required the demolition of Mayfield station. This did not go ahead at the time but the idea was revived in 2015 as one of a number options for the site. [16] one of those options was for a very significant redevelopment of the area around Piccadilly station and the Mayfield area, involving the demolition of both Mayfield station and Gateway House. [28][29] However the status of this is now unknown due to the cancellation of the HS2 Manchester leg. [2]
Entertainment Venue – in 2019, some of the site was converted into Depot Mayfield, a 10,000 capacity venue for culture located at Manchester’s historic former railway Mayfield as part of a £1 billion regeneration project. [17] It regularly hosts The Warehouse Project, a series of club nights. [2]
There is continued interest in the site as an urban regeneration area and it is proposed to replace the station with offices, residential developments and a significant urban green space.
The new green space, ‘Mayfield Park’ opened in 2022. [18]
Manchester Mayfield Redevelopment and ‘Mayfield Park’. [19]
“Mayfield will facilitate transformational change at the eastern gateway of the city centre close to Piccadilly Station. The 20 acre site provides the opportunity to create a distinctive and unique city centre district. The vision for Mayfield is for a distinctive, world class development delivering significant new commercial space, and up to 1500 new homes alongside a mix of retail and leisure facilities all centred on a new 6.5 acre city centre park.” [19]
C.R. Clinker; LNWR Chronology 1900-1960; David and Charles, Newton Abbot, 1961.
Sydney Richards; Manchester and its Railways; in Railways: The Pictorial Railway Journal, Volume 8No. 91, Railway World Ltd., London, November 1947.
S. Hall; Rail Centres: Manchester; Ian Allan Publishing, 1995.
E. M. Johnson; Scenes from the Past: No. 3, Manchester Railway Termini; Foxline, 1987.
British Railways London Midland Region Passenger Services Timetable 16th September 1957 to 8th June 1958.
British Railways London Midland Region Passenger Services Timetable, September 10th 1951 until further notice.
British Railways London Midland Region Passenger Services Timetable 16th September 1957 to 8th June 1958.
Oswald S. Nock; Britain’s New Railway; Ian Allan Publishing, 1966.
C. R. Clinker; Clinker’s Register of Closed Passenger Stations and Goods Depots in England, Scotland and Wales 1830–1977; Avon-AngliA Publications & Services, Bristol, October 1978.
The December 1958 issue of The Railway Magazine featured three photographs of Beyer Garrett locomotives at work in East Africa. These were giants of the metre-gauge that grappled with long loads on steep inclines and at times sharply curved track radii. [1]
1. EARClass ’55’ Garratt No. 5504 at Diva River
Class ’55’ Garratt No. 5504 on the up mixed train at Dura River. [1: p849]
The KUR EC5 class was a class of 1,000 mm (3 ft 3 3⁄8 in) gauge 4-8-2+2-8-4 Garratt-type articulated steam locomotives built during the latter stages of World War II by Beyer, Peacock & Co. in Gorton, Manchester for the War Department of the United Kingdom. The two members of the class entered service on the Kenya-Uganda Railway (KUR) in 1945. They were part of a batch of 20 locomotives, the rest of which were sent to either India or Burma. [2: p64]
The following year, 1946, four locomotives from that batch were acquired by the Tanganyika Railway (TR) from Burma. They entered service on the TR as the TR GB class. [2: p64]
In 1949, upon the merger of the KUR and the TR to form the East African Railways (EAR), the EC5 and GB classes were combined as the EAR 55 class. In 1952, the EAR acquired five more of the War Department batch of 20 from Burma, where they had been Burma Railways class GD; these five locomotives were then added to the EAR 55 class, bringing the total number of that class to 11 units. [2: p64]
This locomotive was Works No. 7151, War Department No. 74235, War Department India No. 423. It was one of the two that went to Burma Railways (their No. 852) from where it was purchased by Tanganyika Railways in 1946 and became their No. 751. It came to the EAR in 1949 and received the No. 5504. [3]
Sister locomotives in Class 55 can be seen here [7] and here. [8]
Dura River was the last station on the Western Extension before the end of the line at Kasese, Uganda. The River flowed North to South towards Lake George and was crossed by the railway at the Eastern edge of the Queen Elizabeth National Park. Mapping and satellite imagery in the area are not highly detailed – the following images are the best I can provide. …
The EAR 58 class was a class of 1,000 mm (3 ft 3 3⁄8 in) gauge, 4-8-4+4-8-4 Garratt-type articulated steam locomotives built by Beyer, Peacock & Co. in Manchester, England, in 1949. [9]
The eighteen members of the class were ordered by the Kenya-Uganda Railway (KUR) immediately after World War II, and were a slightly modified, oil-burning version of the KUR’s existing coal-fired EC3 class. By the time the new locomotives were built and entered service, the KUR had been succeeded by the East African Railways (EAR), which designated the coal-fired EC3s as its 57 class, and the new, oil-burning EC3s as its 58 class. [2: p66][9]
No. 5804 was built in 1949 (Works No. 7293) and originally given the KUR No. 92. Its sister locomotive No. 5808 (Works No. 7297, given KUR No. 96 but never carried that number) was the first to enter service with the EAR. [9]
EAR ‘Class 58’ Locomotive No. 5803 (a sister to 5804) is seen here at Changamwe, Kenya, with the Mombasa–Kampala mail train, circa 1950-51. [9]
Other locomotives in the class can be seen here, [11] here, [12] and here. [13]
Kikuyu Station is 20 kilometres or so from Nairobi, during construction of the railway, railway officers established a temporary base in Kikuyu while they supervised work on the laying of the track down at the rift valley escarpment.
Daily mixed train, headed by class ’60’ Beyer-Garratt locomotive No. 6021, Sir William Gowers,” about to leave Kasese, terminus of the East African Railways & Harbours Western Extension in Uganda. [1: p849]
The EAR 60 class, also known as the Governor class, was a 1,000 mm (3 ft 3 3⁄8 in) gauge 4-8-2+2-8-4 Garratt-type articulated steam locomotives built for the East African Railways as a development of the EAR’s existing 56 class. [2: p77]
The 29 members of the 60 class were ordered by the EAR from Beyer, Peacock & Co. The first 12 of them were built by sub-contractors Société Franco-Belge in Raismes (Valenciennes), France, and the rest were built by Beyer, Peacock in Gorton. The class entered service in 1953-54. [2: p77]
Initially, all members of the class carried the name of a Governor (or equivalent) of Kenya, Tanganyika or Uganda, but later all of the Governor nameplates were removed. [2: p77]
No. 6021 was built by Beyer Peacock (Works No. 7663). It was not one of the class built by sub-contractors Société Franco-Belge. It was given the name ‘Sir William Gowers’ when first put into service, losing the name along with other members of the class in the 1960s after independence. …
Other members of the class can be seen here, [17] here, [18] and here. [19]
Kasese Station only became part of the rail network in Uganda in 1956. The construction costs of the whole line from Kampala were very greatly affected by the difficult nature of the country in the final forty miles before Kasese. Severe problems were presented by the descent of the escarpment, which involves a spiral at one point, while from the foot there is an 18-mile crossing of papyrus swamp through which a causeway had to be built, entailing a vast amount of labour. The extension to Kasese was built primarily to serve the Kilembe copper mines. Construction of the line from Kampala to Kasese took approximately five years. [21]
Roger Hutchinson tells the story of Lord Leverhulme’s interest in Lewis, Harris and (to a much lesser extent) Skye, in the years following the First World War. Lord Leverhulme purchased Lewis and later Harris with a view to developing the islands.
Mingled in with the story are Lord Leverhulme’s plans for transport infrastructure and particularly railways. Hutchinson first introduces railway plans in comments attributed to Thomas Mawson who wrote about Lord Leverhulme’s plans in the Manchester Guardian. Late in 1918, Mawson wrote: “before any … industry can be a success [on Lewis] it will be necessary to provide better transit facilities by sea and land. Safe harbours are the first essential of all economic developments. Engineers are accordingly at work making preliminary surveys for railways which will connect the principle harbours of the island with Stornoway, their natural base. We may soon have a railway on the east coast, connecting Port of Ness, another railway connecting with Callernish on the west, and possibly a third to Loch Seaforth, connecting the Isle of Harris directly with Stornoway. As supplementary to these the main roads are likely to be further improved and motor routes created as feeders to the railways. … A natural corollary to the introduction of railways and harbour facilities will be an increase in the number and size of the steamers trading with the mainland. A trawling fleet, too, is bound to appear as an arm to the fishing industry. Ice factories, cold storage, and canneries for the curing and treatment of fish for export are already planned, and the work of construction will soon begin.” [1: p91]
Hutchinson goes on to comment: “The idea of giving Lewis a couple of branch railway lines was neither original nor utterly ridiculous. Just twenty years previously, in 1897 and 1898, two separate private enterprises, the Highland Railway Company and the Highland Light Railway Company, had surveyed and proposed small-gauge lines between Stornoway and Tarbert in Harris and between Stornoway and Breasclete and Carloway on the west coast. The £500,000 schemes had collapsed when the amount of capital required to be raised by private subscription – £290,000 – was not forthcoming.” [1: p93-94]
Also, “Lewis did have railways, around the turn of the 19th/20th century. There was a railway from the quarry at Bennadrove to Stornoway. Posts related to this track can still be found in the Castle Grounds, opposite the Caberfeidh Hotel.” [5] …
It seems that under Lord Leverhulme’s tenure, “a trackbed was laid near Garrabost in Point, but a railway was never built. The same fate befell the track, linking Carloway to Stornoway along what is now the Pentland Road.” [5]
Very little other evidence exists of his proposed railways, and it is entirely possible that the remains referred to in the last paragraph could relate to much earlier railway proposals. [8]
Lord Leverhulme “planned to develop several smaller fishing harbours around the island’s coast that would be linked by new [his proposed] railways and roads to Stornoway, which would be transformed into a huge fish-processing centre. There was also to be a cannery, an ice-making factory, and a plant to make glue, animal feed and fertiliser from the offal.” [2]
“A chemical industry would also be developed to process the plentiful seaweed around the island; peat would be used in large scale power stations; and unproductive land would be transformed into forests, or fruit or dairy farms. Lewis would grow to become an island of up to 200,000 people. … He had prominent architects and town planners produce a vision of a future Stornoway in 1920. There was to be a town hall and art gallery, a bridge to connect the town to the Castle grounds, long avenues and a railway station, with a war-memorial on South Beach. None of these were constructed although he did give the town a gas supply, and he also intended to use electricity to light the streets.” [2]
For all Lord Leverhulme’s grandiose plans, he was unable to stay the course. His plans “failed in Lewis partly from trying to force the people into too rapid and too fundamental change; he was used to a totally different lifestyle and he tried to define progress on his own terms.” [2]
He faced determined resistance from the local population whose overwhelming desire was for croft land and the freedom to choose what work to undertake. As a result, they did not take to the idea of industrial jobs centred in Stornoway and, in fact, regarded that kind of work, even though salaried, as effective slavery.
“Despite the growing tensions, Leverhulme spent over £1 million in Lewis. Two model housing schemes were built in Stornoway, one on Matheson Road and another on Anderson Road, to house Leverhulme’s managers and employees. New roads were built in Lewis: a concrete bridge, now known as the Bridge to Nowhere, was constructed in Tolsta as part of a scheme to create a coastal road linking Tolsta to Ness at the north tip of the island. The road was never completed.” [3]
Lord Leverhulme’s failure to understand the basic, even visceral, connection between the people and the land was significant.
His proposals centred on his perception of a significant fishery in the waters around Lewis which would sustain industrialised fishing and canning. In reality the stocks were not as great as he believed and the postwar demand for canned fish deteriorated (partly because of barriers to trade with Russia imposed by the UK after the Russian revolution) and as more and more fresh fish from other sources became available. He saw the sale value of fish reduced by 90%.
Lord Leverhulme believed that his offer of good housing and allotments close to Stornoway would ultimately be more attractive than a hand-to-mouth crofting lifestyle. He could not have been more wrong. “What the crofters most needed was casual work to supplement their subsistence farming; what was proposed was regular employment in an industrial process. They did not want to be dependent on any landlord, even a millionaire philanthropist, for their livelihood, and most preferred to take control of their own destiny.” [3]
He stubbornly refused crofters access to good farmland in favour of his desire to see the island self-sufficient in milk, which could anyway be cheaply be imported from Aberdeenshire. As a consequence many men of Lewis raided those farm lands, began building and setting up crofts.
Only a matter of a few short years after the conclusion of the war, Lord Leverhulme had decided that his project was over.
A young doctor, Halliday Sutherland arrived in Stornoway in 1923 “a half-built factory on which work had been abandoned, a derelict small-gauge railway, and thousands of pounds’ worth of machinery rusting on the shore.’ Anxious to uncover the reasons for such a depressing scene, Sutherland approached what was presented in his later transcription as a bellicose old man working a croft in the Back district. The man had no desire, Sutherland said, ‘to answer a whistle at six in the morning and work for wages in Lever’s factory. No damn fear. Poor as I am, I’m master here, and could order you off this croft. … Why did some of us raid his pasture-land? A dairy farm for the island it was to be. I’ve another name for that a monopoly in milk. No damn fear. We are poorer now than we were. Why? Because the line-fishing in the spring has failed. Why? Because of these damned trawlers that spoil the spawn, and half of them are Lever’s English trawlers. He makes us poor, and then wants us to work for him.'” [1: p169][4]
Had Lord Leverhulme’s grandiose plans for Lewis and Harris resulted in lasting changes, there would probably have been some significant changes on Skye. Not the least of these changes may have been the provision of some form of railway from Kyle of Lochalsh onto Skye and through Broadford and Portree to Dunvegan inthe North of the Isle of Skye.
David Spaven & Julian Holland provide a map of proposed railway lines in Scotland. This map shows these proposed but unbuilt railways as dotted lines. Of particular relevance here are the lines on Lewis and on Skye. Although it should be noted that the routes marked predated Lord Leverhulm’s interest in Lewis and Harris by some considerable time. [7: p166 – extract from larger map]
Spaven and Holland’s map does not tell anything like the full story of the planned railways for Lewis and Harris. The map below, provided by Ian B. Jolly shows considerably more detail. It is included in an article in The Narrow Gauge magazine. [15]
Proposed railways on Lewis/Harris. [15: p10]
The Napier Commission’s report on crofting published in 1884 proposed the use of light railways on Lewis and Skye. As a result, The Hebridean Light Railway Co. was formed to promote 130 miles of railway in Skye and Lewis. This led to surveys being undertaken by Alex MacDonald, Engineer, of a possible railway linking Stornoway to Breasclete and Carloway. The survey report was dated 1st June 1893.
Ian Jolly reports that “the trackbed of this line was constructed and, for the first two or three miles out of Stornoway now forms the A858 road, while the rest of the route is an unclassified road to Breasclete.” [15: p12] The unclassified road is the ‘Pentland Road’ which has two arms, one to Carloway and one to Breasclete.
Jolly also notes that O’Dell and Walton, include in their book, ‘Highlands and Islands of Scotland’, a map showing projected railways for Lewis and Skye in 1897 and 1898. In fact, this is just a small part of a map covering the whole of Scotland and are at best schematic in nature. [15: p12][16: p206]
The Outer Hebrides and Skye, showing various schemes which did not come to fruition in the later years of the 19th century and mark with the year in which the schemes would have been constructed. The lines drawn are no more than indicative of the routes proposed. There is no indication of schemes proposed in the 20th century. [16: p206]
One further map is worth noting. This map is provided in a paper by John and Margaret Gold and shows Lord Leverhulme’s development plan for Lewis and Harris. …
Lord Leverhulme’s Development Plan: this gives a good idea of the scope of Lord Leverhulme’s imagination. His ideas were built on the assumption that the fishery around Lewis and Harris was likely to sustain yields over many years. His plans were well-developed. [45: p197]
Later, in April 1919, Jolly says, there were proposals (elsewhere reported as being considered during the first world war) put to a meeting at Staffin, in the north-east of Skye, when representations were made to the Secretary of State, Ministry of Transport and the Highland Reconstruction Committee for a system of light railways on the East side of Skye. There was a similar meeting at Uig, also on Skye, in September 1920 when representations were made to the Ministry of Transport for a light railway connecting Uig to Kyleakin via Portree. Both these meetings were reported in the local press at the time.. [15: p12]
Jolly mentions a comment by Lord Leverhulme which was reported the Highland News, 17th May 1919, that the new harbour at Stornoway should come before the light railway.
Nigel Nicolson, in Lord of the Isles, notes that in a relatively short time (circa. 1920) Lord Leverhulme was looking seriously at light railway schemes on Lewis and had marked out their courses. Leverhulme’s scheme would have had a terminus at Stornoway and three lines:
1) A line South through Balallan to Aline with later extension to Tarbert.
2) A line West, then North to Callanish and Carloway and return to Stornoway via Barvas; and
3) A Branch North from Barvas to serve townships near the Butt of Lewis and return down the east coast through Tolsta to Stornoway.
The total track mileage would have been about 100 miles. The gauge was to be 3ft using WDLR rails being sold as surplus. Lines to be steam worked but Leverhulme wanted electric working! [15: p12] [17: p110-111]
None of these schemes came to fruition.
Photographs and further information can be found in a copy of The Narrow Gauge which is available online. [15]
The Stornoway, Breasclete and Carloway Route
We noted above that construction work did commence on the lines surveyed in 1893, specifically that the trackbed was constructed “and, for the first two or three miles out of Stornoway now forms the A858 road, while the rest of the route is an unclassified road to Breasclete.” [15: p12] It should be noted that the unclassified road follows the planned railway to Carloway as well as to Breasclete and is known as the Pentland Road.
It is nigh impossible to establish the location of the Stornoway terminus from this limited information. Given that modern roads follow the formation of the planned railway route it is quite easy to follow the routes to Breasclete and Carloway. There is, however, a specific, relevant resource held at Stornoway Public Library. It comprises 4 sheets from the 1″ Ordnance Survey 1st Edition mapping from the 1850s, that have been stuck together, with the route itself annotated on top. A digital version of this map has been made available by the National Library of Scotland (NLS). [18] Please note that after navigating to the correct webpage, it will be necessary to scroll down to find the annotated map.
The proposed railway “was planned to connect Carloway and Breasclete on the west coast with Stornoway. Work began on the scheme, but ran into economic and legal problems. Although the railway was never constructed, the ‘Pentland Road’, largely followed the same route, and was built instead by 1912. The road was named after John Sinclair, better known as Lord Pentland who was the Secretary for Scotland between 1905 and 1912 and who helped to secure funding for the completion of the road.” [18]
The map is made up of “four original Ordnance Survey first edition six-inch to the mile maps from the 1850s that have been stuck together, with the route itself annotated on top. … Near Carloway, there is an additional Blue line shewing route originally surveyed changed to avoid damaging arable land.” [18]
Carloway and Breasclete were the western termini of the network with the line to Breasclete appearing to be a branch line.
The extracts from the annotated 1″ Ordnance Survey of the 1850s run in sequence from Stornoway to Carloway and then from the junction to Breasclete. They are the sepia coloured map extracts. Beneath each extract from the 1″ Ordnance Survey is the 2nd Edition 6″ Ordnance Survey from around the turn of the 20th century. These extracts precede the construction of the Pentland Road but show the route the road(s) will take as a dotted track.
Modern satellite imagery is then provided alongside some Streetview images to show the built roads which were completed in 1912 and which are still in use in the 21st century.
There is some doubt over the route of the line approaching and entering Stornoway. Two possibilities with supporting drawings start our look at the line. …
Stornoway to Carloway
Close to Stornoway the alignment of the planned, but never built, railway is not certain. The first possibility is shown immediately below. This takes the mapping provided by the NLS. … [18] The second alternative was discovered by ‘Tom’ in the National Archives at Kew and highlighted on his blog. [53][54]
First, the NLS supplied drawings from the Stornoway Public Library. …
The most easterly length of the proposed railway is shown turning South into Stornoway but no indication is given of the planned terminus. [18]A closer focus from the 6″ OS mapping shows a road following the line of the proposed railway. That road appears to predate the planned railway and it is possible that the line would have run within the road width or on the verge. [19]Willowglen Road first runs North-northwest, then turns through Northwest and West before leaving Stornoway in a West-southwest direction. It is not clear where the Stornoway terminus of the line was expected to be, perhaps to the West of the modern A857 on the portion of Willowglen Road which runs North-northwest from its junction with the A857? [Google Maps, July 2025]Looking North-northwest along Willowglen Road. Rather than this road being built on the line of the planned railway it is likely that the line would have run on the verge of what may well have been a narrower highway at the end of the 19th century. [Google Streetview, September 2024]Now heading West-northwest, the width of the modern Willowglen Road accommodates the planned railway route. [Google Streetview, September 2024]Further West with Willowglen Road now heading in a West-southwest direction. The older road may well have been narrower than the modern road and could have accommodated a railway on its verge. [Google Streetview, September 2024]
Second, the alternative alignment for the East end of the line which appears in documents at the National Archives. [53][54]
The Hebridean Light Railway Company, a blog by ‘Tom’ includes this photograph of a plan from the National Archive at Kew.This plan matches the plan provided by the NLS throughout the length of the line with the exception of the eastern end of the line. [54]
This image shows the eastern end of the line at Stornoway. The route takes a line to the North of what is now Willowglen Road, and to the North of what was Manor Farm, now the Cabarfeidh Hotel, then swinging in a wide arc round the East side of Stornoway before running across the South of the town. [54]
The superimposed red line is a diagrammatic representation of the route, but it does have some resonance with the later temporary railway built by Robert McAlpine & Co. which is covered towards the end of this article. It does however match with other papers in the bundle which ‘Tom’ discovered in the National Archives. …
A very low resolution photograph of the plan of the proposed railway around Stornoway. This is a match for the red line shown above. [54]
A closer view of the last portion of the proposed line on the South side of Stornoway. This compares well with the 6″ Ordnance Survey extract below. It shows that the plan was for the line to terminate at the West end of South Beach Quay. [54]
An extract from the 6″ Ordnance Survey of 1895, published 1899. [55]
The remainder of the route to the West of Stornoway. …..
The red line is the line of the 1893 survey. [18]There is a short section – the hypotenuse of a triangle formed by two roads – at the surveyed line which at the turn of the century was no more than a track along the line of the planned railway. There was then a section of road to the South of Mary Hill before the surveyed route separated from existing roads at the East end of Loch Airidh na Lic. [20]The same area as shown in the map extracts above. [Google Maps, July 2025]This photograph looks from Willowglen Road down the first length of the planned railway route which was independent of existing roads. The planned railway would have run ahead down the centre of the image. A lane can be seen to the right side of Willowglen Road which leads onto the old railway route as shown below. [Google Streetview, September 2024] The first length of road built over the line of the planned railway making use of the civil engineering work undertaken before the railway scheme was abandoned. [Google Streetview, September 2024]The A858 enters this photograph from the left and turns left to run directly ahead of the camera. From this point onwards the road which is now the A858 was built over the line of the railway which was not completed. [Google Streetview, September 2024]The view West along the A858 and therefore also along the line of the intended railway. [Google Streetview, September 2024]Further West, another West-facing view along the lines of the planned railway. [Google Streetview, September The route of the planned railway ran along the South shore of Loch Airidh na Luv. [18]The formation for the planned railway can be seen following the surveyed route. [21]The same area as it appears on Google Maps. [Google Maps, July 2025]Looking West along the A858 which is built on the line of the planned railway. Loch Airidh na Luv is on the right of the photograph. [Google Streetview, September 2024]Further West along the A858, also looking West. [Google Streetview, September 2024]The route surveyed continued West along the South side of Amhuinn a’ Ghlinn Mhoir. [18]The formation follows the surveyed route. [22]The same area as it appears on 21st century satellite imagery. [Google Maps, July 2025]Looking West at the third point from the right of the satellite image above. [Google Streetview, September 2024]Looking West at the third point from the right of the satellite image above. [Google Streetview, September 2024]The surveyed route then switches to the North shore of Loch Vatandip. [18]The track follows the surveyed route, bridging Allt Greidaig just East of Loch Vatandip. [23]The same area in the 21st century. The A858 turns away from the surveyed line of the railway and the Pentland Road begins. [Google Maps, July 2025]Looking West at the road junction the images above. The A858 bears away to the left, the Pentland Road continues ahead and bears to the right. [Google Streetview, September 2025]The loch on the left is Loch Vatandip. [Google Streetview, September 2024]Further West along the single track Pentland Road, looking West. [Google Streetview, September 2024]The surveyed route ran West-northwest above the North shore of the loch. [18]While the surveyed line is straight on the map extract above, the line of the track shows a minor deviation as it heads West-northwest between Loch Vatandip and Loch Mor a Chocair. [24]The same area in the 21st century, the same minor deviation in the alignment of the Pentland Road. [Google Maps, July 2025]The same minor deviation in the road alignment seen from the East. [Google Streetview, September 2024]The surveyed route continues on the same bearing. [18]The track on the formation of the proposed railway matches the survey, passing to the South of Loch Beig a Chocair and bridging two streams – Loch a Chocair and the Greta River (or the River Creed). It seems that work on the railway extended to the construction of bridges ready for the final addition of the rails. [25]The same length of road in the 21st century. [Google Maps, July 2025]The first of two bridges on this length of road, built for the railway that never arrived! [Google Streetview, September 2024]The second of those bridges, also seen from the East. [Google Streetview, September 2024]This is the first of four map extracts where the original survey route is shown in blue. The red line being that which was used. No reason for this alteration is provided. [18]Small culverts or pipes are not marked on the OS mapping but there must be one over Allt a’ Bhiorachan at the left of this map extract and possibly two other smaller culverts or pipes close to the centre of the extract. [26]Having checked each of the three locations where streams run under the surveyed route which is now a road, there is no visible structure, so there is probably no more than a drainage pipe at each location. [Google Maps, July 2025]The surveyed route has now turned slightly to run East-West, before turning West-northwest again to the North of Loch an Tobair. [18]It seems that the final alignment of the earthworks prepared for the railway was, over the first half of this extract, North of either of the marked survey lines. A further culvert/pipe must have been provided for the stream flowing South into Loch an Tobair. [27]The same area in the 21st century. [Google Maps, July 2025]No sign of a structure at the point where the feed to Loch an Tobair passes under the road so a drainage pipe must suffice. The wide open skies on Lewis are amazing! [Google Streetview, September 2024]Now back on an West-northwest alignment, the original survey line (blue) and that deemed to have actually been used (red) run in parallel. [18]The same length as it appears on the 6″ OS mapping at the turn of the 20th century. No bridges are marked at the crossing point of the two streams which suggests that smaller culverts or drainage pipes were used. [28]The same area in the 21st century. [Google Maps, July 2025]The first (most easterly) drainage ditch crosses the line at this location, a pipe of some sort must pass under the road. [Google Streetview September 2024]At the second (more westerly) location, standing water is visible to the South of the road, drainage from North to South must be by a pipe. [Google Streetview, September 2024]The two surveyed routes come together again North of the East end of Loch an Laoigh. [18]Three gravel pits are marked along this length of the formation. No bridges are marked so culverts must have been employed for the two watercourses. The track appears to run a little to the North of the surveyed alignment. [29]The two streams shown on the map extracts above. Both show water downstream of the road, one appears to have a corrugated plastic pipe under the road. [Google Maps, July 2025]The location of the more easterly watercourse seen looking West: a plastic pipe can be seen to the left of the road. [Google Streetview, September 2024]Drainage water can be seen to the left of the road in this West-facing view at the location of the more westerly watercourse. No drainage pipe is visible from the road. [The proposed junction with the line to Breasclete heading West-southwest and that to Carloway heading Northwest. [18]The linto Carloway heads Northwest and crosses Allt Mhic Ille Chetheir. [30]The junction: Breasclete is to the East and Carloway to the Northwest. [Google Maps, July 2025]The junction seen from the Southeast. [Google Streetview, September 2024]The road to Carloway: both arms of the road are called Pentland Road. [Google Streetview, September 2024]The proposed line ran to the West of a group of three lochans – Loch Mor a Ghrianain, Loch Beag a Ghrianain and Loch an Fheoir. [18]Continuing Northwest the planned line to Carloway crossed Allt nan Lochanan Traighte and ran passed a small quarry which was not marked on the 1″ mapping of the 1850s. [31]The modern road continues to follow the planned railway route. Google Maps, July 2025]The road travels on a causeway with drainage ditches on each shoulder. [Google Streetview, September 2024]The surveyed route curves around the top of Loch Laxavat (Lacsabhat) Ard. [18]A larger area than shown on the survey sheet above which shows clearly a relatively tight curve on the alignment of the railway formation to the Northeast of the Loch. [32]This satellite image matches the area shown on the extract from the 1″ Ordnance Survey of the 1850s. [Google Maps, July 2025]The road follows the planned railway route curving to the left to avoid higher ground. [Google Streetview, September 2024]Then curving to the right around a rock outcrop. [Google Streetview, September 2024]The tight curve mentioned in the notes about the extract from the 6″ OS mapping above appears towards the bottom-right of this extract from the survey plans. [18]This 6″ OS extract takes the line to a point just to the West of the River Ohagro which feeds onto the North of Loch Laxavat Ard. It will be noted that there is a break in the embankments built for the proposed railway where a bridge would have been placed over the river. A short diversion provides access by means of a ford across the river. The ‘as built’ looks NE of embankments do not follow the survey to the East of the River Ohagro. [33]A similar area to that shown on the 6″ Ordnance Survey. [Google Maps, July 2025]The road can be seen undulating ahead, possibly foreshortening exaggerates this effect. The 6″Ordnance Survey shows that embankments were constructed at the end of the 19th century. [Google Streetview, September 2024]The bridge over the River Ohagro which feeds into Loch Laxavat Ard. This bridge was not constructed as part of the railway contact and had to be built as part of the construction of the Pentland Road early in the 1910s. [Google Streetview, September 2024]Another culvert takes the line over the Allt nan Cnocan Dubh. [18]This next extract from the 6″ Ordnance Survey takes the track beyond Conan Dubh to approximately the same point on the surveyed line as the 1″ extract above. [34]A very similar length of the road as shown in the map extracts above. [Google Maps, July 2025]Looking Northwest along the Pentland Road at the centre of the satellite image above. [Google Streetview, September 2024]Further West the road curves round a rick outcrop on the North side of Conan Dubh. [Google Streetview, September 2024]This next length of the survey takes the proposed line as far as Loch Thorrad. [18]No obvious provision is made for the proposed line to cross the Allt Loch Thorrad, so a culvert or drainage pipe must be presumed. [35]This satellite image takes us as far as Loch Thorrad (which can be seen on the North side of the road at the left side of the image. [Google Maps, July 2025]Loch Thorrad is to the right of the road as it curves a little to the Northwest. [Google Streetview, September 2024]From Allt Loch Thorrad onwards the line heads Northwest. [18]After crossing the Allt Loch Thorrad the earthworks got the planned railway stay to the Northeast of the Carloway River. One tributary to the Carloway is crossed as the proposed line headed Northwest. [36]The road now follows the valley of the River Carloway. [Google Maps, July 2025]This and the next image show the road following the planned railway route alongside the River Carloway. [Google Streetview, September 2024]The road picks its way between rock outcrops and the river. [Google Streetview, September 2024]The earthworks remain on the Northeast side of the Carloway River for most of this length. [18]The same length of the proposed railway. The Carloway River stays on the Southwest side of the river until the top-left of this extract, where the line crosses the River. One stream is culverted under the railway. [37]We are relatively close to Carloway: the road follows the Northwest bank of Carloway River before the river passes under the road near the top-left of this satellite image. [Google Maps, July 2025]This and the next image are two photographs showing the Carloway River meandering around close to the road. [Google Streetview, September 2024]Buildings at the edge of the village of Carloway can just now be picked out in the distance e. [Google Streetview, September 2024]Looking Northwest along the Pentland Road over the bridge carrying the road over the Carloway River which flows left to right under the bridge. [Google Streetview, September 2024]Two alternative alignments for the proposed railway appear again close to Carloway. The original surveyed route is shown by the blue line. The planned route was moved so as to avoid the better farmland. [18]Track which follows the formation of the planned railway crosses Gil Fasgro and runs immediately adjacent to the Carloway River. [38]The Pentland Road runs down towards Carloway following approximately the red line from the survey. Google Maps, July 2025]At the junction in the bottom-right of the satellite image the Pentland Road heads North following the river valley. [Google Streetview, September 2024]The Pentland Road runs alongside the Carloway River. [Google Streetview, September 2024]Closer to Carloway and still alongside the River. [Google Streetview, September 2024]On the right of the image the Heidagul River joins the Carloway River and from this point on the combined stream is known as the Heidagul River. Google Streetview, September 2024]The final length of the survey shows the revised alignment (in red) close to the river and crossing the Carloway River close to Carloway Bridge. The surveyed route extends as far as the pier a Borraston, Dunan Pier. [18]The last length of the Carolway line as recorded in the bundle from the National Archives that ‘Tom’ discovered and wrote about on his blog. [54]This extract from the 6″ Ordnance Survey of the turn of the 20th century covers a similar area as the 1″ map extract above. The pier can be seen bottom-left. The line of the planned railway is less clear from the 6″ OS in Carloway but becomes much clearer on Google Streetview images as it follows the North shore of the estuary. [39]The 21st century satellite imagery highlights.more clearly the route of the planned railway and what became the Pentland Road through to the Dunan Pier near Borraston. Of particular interest is the arrangement of structures close to Carloway Bridge. A bridge over the Heidagul River and a bridge which now carries road over road will both have been built as part of the aborted railway works. [Google Maps, July 2025]This view shows the two structures noted above. The masonry arch structure is Carloway Bridge which carries the modern A858. The bridge over the river in the foreground was built for the railway as was the bridge which carries the A858 over Pentland Road. [Google Streetview, September 2024]The bridge over the Heidagul River built for the planned railway, seen from Carloway Bridge. [Google Streetview, September 2024]A closer view from the East of the bridge built to carry the road over the railway that never was! [Google Streetview, September 2024]Looking back East towards the two bridges carrying the A858 in Carloway. [Google Streetview, September 2024]
The next seven images form a sequence showing the last length of the route to the pier at Borraston. Note the causeway in the third image which will have been built for the railway. …
The pier at Dunan near Borraston. [All seven images: Google Streetview, September 2024]Dunan Pier was the end of the line: shown here in an extract from the 6″ Ordnance Survey 2nd Edition from the end of the 19th century. [49]
The Junction to Breasclete
The surveyed route of the branch line to Breasclete curved round the North side of Loch an Tairbeart nan Cleiteichan and Loch an Tuim. [18]The track which appears on the 6″ OS mapping from the turn of the 20th century takes a single radius curve a little to the North of the surveyed alignment. [30]The single track road built on the earthworks of the abortive railway project curves round the North side of Loch an Tairbeart nan Cleiteichan and Loch an Tuim. It is a smooth curve as shown on the 6″ Ordnance Survey from the turn of the 20th century. [Google Maps, July 2025]The road to Breasclete is also called Pentland Road. It heads away to the left of th. [Google Streetview, September 2024]The surveyed route then turns to a Westerly alignment South of an unnamed lochan and across the North end of Loch na Ba Buidhe. [18]Another small quarry sits on the North side of the track. Presumably the small quarries at intervals along each of the planned railway routes were used to supply stone for embankments along the formation. [40]The same length of the road. [Google Maps, July 2025]The view West from the centre of the satellite image above. [Google Streetview, September 2024]The surveyed route ran across the North side of Loch na Ba Buidhe, Loch a Ghainmheich and Loch Avaster (Amhaster), turning to head West-southwest. [18]The same area as it appears on the 6″ Ordnance Survey from the turn of the 20th century. [41]The same area in the 21st century. [Google Maps, July 2025]Looking West from the centre of the satellite image above. Loch a Ghainmheich is on the left . [Google Streetview, September 2024]Continuing in a Southwesterly direction, the surveyed route ran on the North side of Loch na Beinne Bige. [18]A similar area as it appeared at the end of the 19th century. The track following the built formation for the railway follows the surveyed alignment closely but turns away from it to the left of this map extract. [42]The same area as shown on the 6″ Ordnance Survey extract above. [Google Maps, July 2025]Allt Glas flows under the road, presumably in a drainage pipe, twice the first of these locations is shown here. [Google Streetview, September 2024]Allt Glas flows under the road again although it appears to both pass under the road and to have found a path on the North side of the road. [Google Streetview, September 2024]Allt Bealach na Beinne also passes under the road, Allt Glas joins it on the left of this photograph. [Google Streetview, September 2024]Looking West-southwest along Pentland Road, Loch Na Beinne Bige is on the left. [Google Streetview, September 2024]The final length of the surveyed route for the planned railway. Two alignments are shown, the original (in blue), the revised (in red). [18]This extract comes form the documents held by the National Archives and photographed by ‘Tom’ for his blog as noted below. [54]The track which follows the prepared formation for the planned railway passes to the North of both of the surveyed routes as it runs through the village of Breasclete, regaining the red surveyed alignment to the West of the village and running through to the pier. [43]A very similar area to that shown on the 6″ OS map extract above. [Google Maps, July 2025]Facing West-southwest approaching the crossroads in Breasclete. [Google Streetview, September 2024]Continuing West-southwest along Pentland Road towards the pier at Breasclete. #[Google Streetview, September 2024]The second crossroads in Breasclete. [Google Streetview, September 2024]Approaching the pier at Breasclete. [Google Streetview, September 2024]
We have followed the two lines that almost got built on Lewis. sadly, lack of funding resulted in a project that was quite well advanced, being abandoned. As noted, the earthworks were later (1912) used to create the single track Pentland Road which appears in many of the modern images above. There remains some uncertainty over whether the early construction works were designed first for a railway or were just designed as an easily graded public road. [50] It is possible that some construction work for a road was undertaken but the National Archives hold plans for a railway dated to the same period, predating the construction of the Pentland Road which was not completed until 1912. [53]
The plans, which include proposals for railways on Skye and on Lewis were accessed by ‘Tom’ in preparing for a modelling project centred on these intended railways. This image comes from an early blog. [53] The images relating to Lewis come from a later blog. [54]
Lord Leverhulme’s Planned Railway Station, Stornoway
Lord Leverhulme was very interested in town planning, The National Library of Scotland has on its website, a town plan of Stornoway drafted by James Lomax-Simpson, Leverhulme’s godson and also his chief architect at Port Sunlight. The plan is entitled, ‘Port Sunlight plan of Stornoway, showing proposed lay-out’ and is dated 16th July 1919. It is 710 mm x 710 mm in size. The plan is included on the website, courtesy of The Stornoway Trust. [44]
“Simpson took charge of the Architectural Department of Lever Brothers from 1910 and he was made a director in 1917. In his role as Company Architect, he worked in over twenty-five different countries around the World, but he also carried out much work for Lever himself, including alterations and additions to Lews Castle. The plan also illustrates part of Leverhulme’s ambitious ideas for redeveloping Stornoway along garden city lines, with new suburbs, broad avenues, circuses, and open spaces. The new planned railways, that were part of the wider plans for the economic transformation of Lewis, curve in and down to the Harbour on the eastern side of the town. Existing roads are shown with dashed lines. In places, ‘Parlour Cottages’ were planned, which had been constructed at Port Sunlight, as larger ‘Arts and Crafts’ residences for working families with a parlour at ground-floor level. Although visionary and ambitious, some of the new planned streets would have demolished much of the original old town. Over time, the plans were subsequently altered, shown as annotations on top of the original plan. Some construction began along these lines in the 1920s, but economic difficulties and considerable opposition to Leverhulme’s plans by the islanders curtailed developments, and the schemes were largely abandoned by 1923.” [44]
Small extracts from the plan are included here. They show a proposed railway station close to the Harbour on the East side of the town. Each of the three extracts is paired with the ESRI satellite imagery provided by the NLS. …
The proposed station location with the main station building facing out onto a circus/roundabout close to the harbour. [44]A double track line was planned Northeast from the station. [44]The detail becomes more sparse further Northeast. [44]
These plans did not see the light of day!
Goat Island
Lord Leverhulme’s plans included the construction of a causeway to link Goat Island to the mainland and the provision of additional quays on the West side of the island. He expected to provide a light railway along the causeway to link his Cannery and associated industries to the quays. John & Margaret Gold provide a plan showing Leverhulme’s intentions for Stornoway and Goat Island. [45: p200]
John & Margaret Gold comment that in Leverhulme’s Plan: “An industrial area was located in the east of the town. Goat Island would act as home base for the MacLine Drifters and Trawler fleet and was joined to the mainland by a causeway. The ice plant and cannery were situated inland near the site of the existing fish-oil and guano works. A light railway would connect them with the quays. There were tweed mills, electricity generating plant, a laundry and a dairy to take the increased output from the east coast farms. Between the industrial area and the residential districts was the railway station serving both freight and passenger purposes.” [45: p200]
Lord Leverhulme’s Development Plan for Stornoway: the railway line noted in the paragraphs about Stornoway’s railway station can be seen to the right of centre. The light railway planned to serve Goat Island is shown in the bottom-right of this map. [45: p200]Goat Island in 1895 as it appeared on the 6″ Ordnance Survey of 1895, published in 1899. [46]
It would not be until after the Second World War, in 1947, that the causeway was built. It was 2,030 feet in length. Work undertaken that year also included the construction of an embankment to the south of the causeway; the construction of the Slipway and a jetty at Goat Island; the demolition of No. 3 Pier. The work was authorised by The Stornoway Harbour Order Confirmation Act, 1947. [48]
Goat Island and causeway as they appeared on the 6″ Ordnance Survey of 1958. [47]Goat Island and Causeway in the 21st century. [Google Earth, July 2025]The causeway to Goat Island. [Google Streetview, September 2024]Goat Island seen from the causeway. [Google Streetview, September 2024]
While the causeway was built, the railways were not!
The Branahuie Railway (3ft-gauge)
One line that did get built on Lewis in Lord Leverhulme’s time was a 3ft-gauge line built by “Sir Robert McAlpine and Co. for the Harris & Lewis Welfare Development Co. Ltd. (a company owned by Lord Leverhulme) – part of a £345,000 contract to build the canning factory, roads and houses. … [It] was in operation by 1920 when the first loco arrived – [that] was McAlpine’s Loco No 34, an 0-4-0ST built by Hudswell Clarke (Works No 1037) in 1913 and delivered to McAlpine’s Pontstycill Reservoir contract near Merthyr Tydfil in South Wales. It carried the plant number 778 when it arrived but that had changed to 606 by the time it left in 1923 moving to the Maentwrog reservoir contract in North Wales. It then worked on other contracts until it was sold for scrap to George Brothers in 1956. The second loco to work on the line was another 0-4-0ST built by Hudswell Clarke in 1901 (Works No 597). It was new to Newcastle & Gateshead Water Co Ltd at Whittledean reservoir carrying the name ‘PONT’. It was sold back to Hudswell Clarke who resold it to McAlpines in 1906 on their Culter reservoir contract. It arrived at Stornoway as Plant No 1780 in 1920 leaving on 25th May 1923 as Plant No 813. Last recorded as being for sale at McAlpine’s Ellesmere Port depot in 1929.” [8]
A first reference was made to the Branahuie line in the Highland News on 15th May 1919 when Sir Robert MacAlpine & Son wrote to the Council seeking permission to lay a light railway from Manor Farm to Goathill Road crossing public roads at three different places. Gates and fences were included in the scheme which received Council permission. In June 1919, MacAlpine applied for permission to lay a water main at Manor Farm to supply water to engines. This was agreed at charge of £5 per annum. [15: p12]
Jolly records these details: “The line was some five miles long and was used for the construction of Leverhulme’s Cannery, from where it ran northwards past Goathill and Manor Farm (Coulregrein), where there was a watering point, to the Town Council’s Dormitory Quarry beneath the War Memorial. This line ran around the then outskirts of the town and much has been built over. Another line ran south from the cannery to the locomotive shed (also now built over). From here another line ran eastwards across the fields to Sandwick, then for 2.5/3 miles beside the A866 to the beach at Branahuie. The evenly graded trackbed is very distinct alongside the undulating road on this section. A shallow cutting can also be seen on the northern line. … At least two steam locos were used on the contract by MacAlpines.” (15: p12]
Jolly provides this drawing of the route of the 3ft-gauge contractor’s railway. It is schematic in nature and not to scale. Manor Farm and Goat Hill Farm appear to the Northeast of Stornoway and of the line. The Cannery is marked, as is the Loco Shed. The line to Branahuie is also shown. I have not been able to find any greater detail as to the route of the line than the text description of the route above. [15: p9]
Jolly continues: “The cannery was completed in late 1921 or early 1922, and at the end of May 1922, the “Contract Journal carried an advertisement: ‘For sale-railway track and plant inc. two 3ft gauge locos Hudswell Clarke, … built 1901 and 1913, and 59 wagons 3ft gauge, 34 wagons 2ft gauge. Plant will be handed over to purchasers FAS (free aboard ship) Glasgow-Lewis & Harris Welfare & Development Co., Bebington, Nr. Birkenhead’. Only two locomotives fit this description: Hudswell Clarke 597/1901 was delivered new to the Newcastle & Gateshead Water Co, and was later used by McAlpine on the Motherwell Corporation Culter Waterworks contract between 1903 and 1906. Its later history is not known for certain. The later machine, Hudswell Clarke 1037/1913, was supplied to McAlpine for work on the Pontstycil reservoir between 1913 and 1917. It was subsequently used on the Maentwrog Hydro-Electric reservoir contract, near Ffestiniog, from 1924-28, and must therefore have been retained by McAlpine.” [15: p12-13]
The Route
Lord Leverhulme’s Cannery sat to the East of Stornoway town centre. Appropriately, its address was Cannery Road. The building was never used as a cannery and later became a Harris Tweed Factory.
As Jolly mentions, material for the construction contract was excavated at a quarry at Dormitory which was to the West of the War Memorial (itself to the North of the town). Jolly also mentions that the temporary railway line ran close to Manor Farm (in the 21st century the Caber feidh Hotel occupies this site). His sketch map above shows the line running to the South of Manor Farm. This suggests that the line ran close to Willowglen Road, on its North side. Assuming that this is the case then the Contractor’s railway would have crossed Percival Road South close to its junction with Willowglen Road.
There has been mention of an incline leading from a point close to the War Memorial into Stornoway which may be a remnant of the line. [8]
The area from Dormitory to Manor Farm as shown on the 6″ Ordnance Survey Second Edition. [51]Pretty much the same area as it appears on Google Maps. [Google Maps, July 2025]A closer view of the area around Dormitory as it appears on the 6″ Ordnance Survey from the end of the 19th century. The contractor’s railway would have run East from the quarries close to Dormitory, probably parallel to and on the North side of what would eventually become the A858 (Willowglen Road). [52]A similar area in the 21st century. [Google Maps, July 2025]A closer view of the area around Manor Farm as it appears on the 6″ Ordnance Survey from the end of the 19th century. The contractor’s railway would have run West-East, probably parallel to and on the North side of what would eventually become the A858 (Willowglen Road). [52]Much the same area in the 21st century. [Google Maps, July 2025]
After crossing what is now called Percival Road South, the line crossed Macaulay Road and curved round through Goat Hill, passing the Poor House and the Hospital, running close to the pre-existing Fish Oil Works (adjacent to which Leverhulme’s Cannery was to be built). The Locomotive Shed was South and West of that location, as was a junction between the line from the quarry and the line East to Branahuie.
The line of the contractor’s railway heading East is not clear. The red-dotted line gives an idea of the possible alignment. Initially over open fields it has then been covered, by the extended cemetery at Sandwick and by housing developments. [Google Maps, July 2025]Some field boundaries support the assumed route but there is no guarantee that this is the actual line of the contractor’s railway. At the right side of this image the line has once agin been built over. [Google Maps, July 2025]The only indication as to the route of the line to the East of the built up area that I have been able to find is Jolly’s comment that the line ran alongside the A866. He says (above) that the line of the old railway is level while that of the road undulates. [Google Maps, July 2025]This view East along the A866 is taken from a point a little to the East of the end of the development visible at the left of the satellite image immediately above. It is not beyond the bounds of possibility that the land immediately to the right of the road was the route of the contractor’s railway, but it does not appear as though the highway undulates as much as Jolly suggests. [Google Streetview, September 2024]Further to the East, a relatively slight gradient is evident in the road but there is little evidence of an old railway formation alongside the road. [Google Streetview, September 2024]Jolly has the remaining length of the contractor’s railway to Branahuie continuing along the South side of the A866. [Google Maps, July 2025]
It is entirely possible that the road now evident in the 21st century is not that which was present in the 1920s. It is very likely that the road to Branahuie was a single track road in the 1920s and that the widening of the road has covered the formation from the contractor’s railway line.
Stornoway Waterworks Railway(2ft-gauge)
“Since the 1870s Stornoway’s water supply had come from Loch Airigh na Lic, about two miles west of the town, but by the mid-1930s this was proving insufficient for the population of around 5000 which was swelled by four or five hundred herring drifters operating out of the port during the season. Loch Mor an Stairr, five miles north-west of the town, was chosen to augment the supply as it was free from pollution and some distance from public roads. The exit from the Loch was between peat banks some 65ft apart, and it was across this that a concrete dam, 92ft long, was constructed. Pipes led at different levels to a small valve house on the north bank of the outlet stream, and a 9inch main then connect[ed] to the filter houses beside the main road.” [15: p8]
The Waterworks Railway. Another small extract [15: p9]
The work was facilitated by the construction of a 2ft-gauge railway line.
A Simplex locomotive was used on the Stornoway Waterworks Railway. This locomotive was a 20hp model built by Motor Rail Ltd. It operated on the 2-foot gauge line that served the Stornoway Waterworks. Its Works No. is not known. One source suggests No. 110U082 but the records at the Apedale Valley Light Railway have that works number attributed to a 3ft-gauge locomotive at the Bo’ness & Kinneil Rly. [10] It is worth noting that the Almond Valley Light Railway has a 2ft 6in-gauge example. [11]
The Stornoway Waterworks Railway was built in the 1930s and ran for approximately 1.5 miles between Stornoway Waterworks and Loch Mòr an Stàirr. It was used to transport materials during the conversion of the loch into a reservoir for the waterworks and for subsequent maintenance works. It was closed by the 1960s. [12][13]
Writing about the locomotive and the construction work in 1982, Ian B. Jolly states: “The Contractor for the dam and pipeline was G. Mackay & Son. of Edinburgh, who started work on the dam in 1935. Their work was completed mid-1936 when the pipeline was connected direct to the town’s mains – the filter house and covered reservoirs were completed within the next few years. … A locomotive-worked narrow gauge tramway was used by MacKay & Son to construct the dam. Rock was excavated and crushed in a small quarry east of the main road. across which it was transferred by lorry to the tramway terminus. Stone and other materials were then carried by rail to the site of the dam. The railway was left in-situ and used by Stornoway Town Council for maintenance of the dam for many years. The loco, a 20 h.p. bow-framed model built by the Motor Rail & Tramcar Company of Bedford, was in use until at least 1940 when Mr Alex Macleod, the fitter who maintained it. was called up for military service. By 1943 the engine had been removed and it had been reduced to a frame and wheels. in which form it is believed to have been in use, pushed by hand, until the early 1960’s as the line’s only item of rolling stock. The loco frame is now [1982] very delapidated and derailed about half a mile from the filter house. It was originally fitted with a Dorman 2JO two-cylinder petrol engine; not the later, but similar 2JOR engine. The axleboxes have ‘W D 1918’ cast on them. whilst the loco had been fitted with the narrow pattern of brake column. This suggests that it was built during late 1918 for the War Department Light Railways, but sold directly as Government Surplus. Motor Rail’s records throw no light on its identity – the only locomotives credited to G. Mackay & Son of Edinburgh are two 40 h.p. ‘protected’ machines: LR3057 4wPM MR 1336/1918 and LR3088 4wPM MR 1367/1918. Both were in the service of MacKay by 21st June 1924. MR 1336 was later with Inns & Co Ltd, Moor Mill Pits. Colney St, Herts. and MR 1367 was with Thomson & Brown Bros Ltd, of Edinburgh by 16th February 1933. There is no mention of a 20 h.p. loco but MacKay was obviously no stranger to Motor Rail & Tramcar Company products.” [15: p9]
Jolly further notes that “Rolling stock on the line at the time of the dam construction consisted of nine one-cubic-yard skips, a mixture of side and end tippers. The derelict remains of several [could in 1982] be seen at the foot of the bank beneath the filter house, one being a single end tipper. The axle boxes [were] marked ‘Du Croo & Brauns’ – the Dutch firm of railway equipment suppliers. … Most of the track from the roadside terminus to just beyond the loco [had by 1982] been removed without authority – probably for fencing posts! However, the track layout [could] be traced because the turnouts [had in 1982] been left in place. These [were] rivetted to corrugated steel sleepers, whilst the remaining track [was] spiked to wooden sleepers or clipped to corrugated steel sleepers.” [15: p11]
Of further interest, is the significant variation in rail cross-section and weight (between 14lb and 20lb per yard). Jolly also notes that, “On the lengths of prefabricated track where the rails [were tied accurately to gauge, three distinct gauges [could] be measured – 2ft, 60cm (1ft 11.5/8in) and 1ft 11½ in! The loco wheels [were] set to 60cm gauge.” [15: p11]
In 1982, only minimal earthworks were evident, with track following the undulation of the land but, says Jolly, “there is a rise of just over 25 feet from one end of the line to the other. The track terminates near the dam without so much as a buffer stop or siding. The remains of the loco and line will probably survive for many years to come, as scrapmen are unknown in the Outer Hebrides.” [15: p11]
Another photograph which shows remains of point work close to the Waterworks can be seen on the Railscot.co.uk website. [14]
Other Railways?
Jolly comments: “There appear to have been three other industrial railways in Lewis, lain D.A. Frew referred to the horse-worked system on the outskirts of Stornoway. This served the factory of the Lewis Chemical Co, promoted in the late nineteenth century to extract paraffin-oil from peat by a patent process. Garrabost Brickworks, about 8 miles east of Stornoway) is reputed to have had a short line. The brickworks is shown on the 1852 6in map but no railway, and the 1897 edition shows the works as ‘disused claypit. We were also told of Marybank Quarry, west of Stornoway, where there was a hand worked line from the rockface about 100 yards to the crusher. The quarry was operated in the few years before the last war by William Tawse of Aberdeen.” [15]
Other lines are referred to in a blog about the island accessed through the BBC website. The blog is entitled ‘Arnish Lighthouse’ and includes these words. … “Lewis did have railways, around the turn of the 19th/20th century. There was a railway from the quarry at Bennadrove to Stornoway. Posts related to this track can still be found in the Castle Grounds, opposite the Caberfeidh Hotel. … A trackbed was laid near Garrabost in Point, but a railway was never built.” [56]
I have not yet been able to find anything further about any of these short lines. There is an active quarry at Bennadrove. This is not far from Marybank
This final satellite image shows the relative locations of Marybank and Bennadrove to the West of Stornoway. It also encompasses most of the different line referred to in the immediate vicinity of Stornoway. [Google Maps, July 2025]
Records
Plans illustrating the surveyed railway routes proposed by Lord Leverhulme can be accessed at Tasglann nan Eilean Siar, the Hebridean Archives. [6]
References
Roger Hutchinson; The Soap Man: Lewis, Harris and Lord Leverhulme; Birlinn, Edinburgh, 2003 (latest reprint 2017).
We begin this article with a look at maps of the Piazza Raffeale de Ferrari and its immediate environs over the years around the turn of the 20th century. The Piazza became one of two focal points for tramways in the city (the other was Caricamento).
I found the series of maps interesting and they provoked a desire to find out more about the network of horse-drawn and later electric trams and tramways of Genoa. ….
Italian Wikipedia informs us that: “The first public transport in Genoa was provided by a horse bus service linking the city centre and Sampierdarena, that started in 1873. In 1878, the French company Compagnia Generale Francese de Tramways (CGFT, French General Company of Tramways) began to build a horse tram system.” [16][17]
Towards the end of the century, the new urban plan led to the construction of new roads with wider carriageways, principal among these were:
Via Assarotti connecting Piazza Corvetto to Piazza Manin;
Via XX Settembre, built between 1892 and 1899, widening Strada Giulia and connecting the Palazzo Ducale (Piazza de Ferrari) with Porta Pila and the banks of the River Bisagno (once the eastern boundary of the city);
Corso Buenos Aires, once outside the city walls, was lowered to the level of Ponte Pila and the new Via XX Settembre, to form a single artery that would connect the centre with the Albaro district;
Corso Torino, perpendicular to Corso Buenos Aires.
After this work was done, the city began to look more modern and the widened streets made room for tramways in the centre and East of the city. The municipal administration began to plan new lines, both towards the eastern suburbs and in the central districts of the city. [19]
The city welcomed competition and set up a series of concessions which were given to different groups: the French Company kept the Western concession; Val Bisagno and the hilly areas to two Swiss businessman (Bucher & Durrer); and the east of the city was granted to a group of local businessmen. [19][20: p66]
The two parties, other than the French, formed companies: Bucher created the Società di Ferrovie Elettriche e Funicolari (SFEF) in 1891. [20: p85] The Genoese entrepreneurs founded the Società Anonima Tramways Orientali (SATO) in 1894. [20: p120] The two companies took on the two concessions which envisaged electric traction on metre-gauge lines to accommodate running on the narrow winding streets of the city centre. [19]
“By 1894, SFEF had achieved no more than a single short electric tram line between Piazza Manin and Piazza Corvetto, whilst SATO had not progressed beyond the planning stage. The CGFT system had extended through the city and the Val Polcevera, but was still horse operated.” [16][17]
“In 1894, the German company Allgemeine Elektrizitäts Gesellschaft (AEG) … bought both the SFEF and SATO companies. The following year AEG created the company Officine Electrical Genovesi (OEG), … which took over the city’s existing electricity supply company, and the Società Unione Italiana Tramways Elettrici (UITE), … which purchased the CGFT’s concession. By the end of 1895, AEG had a monopoly of both electricity supply and public transport provision in the city.” [16][17] Under AEG’s “ownership, SFEF and SATO developed a tram network of more than 53 km (33 mi) reaching Nervi and Prato, whilst UITE electrified their lines to Voltri and Pontedecimo.” [16][17]
As we have already noted, the first electric traction line connected Piazza Corvetto to Piazza Manin, running along Via Assarotti. [20: p92] It was activated by SFEF on 14th May 1893 [20: p96] The single-track line was 800 metres long and ran on a constant gradient of 7% [20: p95]; the tickets cost 10 cents. The electrification (600 V DC) was via an overhead cable and was carried out by AEG of Berlin, which, as we have already seen, later acquired a significant shareholding in the company. [19][20: p86-87]
In subsequent years the SFEF network expanded rapidly; in 1895-96 the Monte line to the North of the city centre entered into service, including the Sant’Ugo spiral tunnel; in 1896 the line from Piazza Principe to Piazza Brignole was born. It included two tunnels in the Castelletto area. [21: p20] , In 1897, the Val Bisagno line up to Prato began operation. [19][21: p26]
The first SATO line entered into service on 26th July 1897, connecting Piazza Raibetta to Staglieno through the Circonvallazione a Mare, [20: p122] followed two years later by the long coastal line to Nervi. [20: p127] In 1900 the eastern trams reached the central Piazza de Ferrari, travelling along the new Via XX Settembre which was formed through widening of the old Via Giulia. [19][21: p53]
The two networks, SFEF and SATO, were technically compatible and the two companies, both controlled by AEG, soon unified the two networks. [20: p142]
“Finally in December 1901, AEG merged SFEF and SATO into an enlarged UITE.” [16][17]
The enlarged UITE found itself managing 70 km of network, divided between the 30 km of the ‘Western network’: (formerly the French Company) and the 40 km of the ‘Eastern network’ (formerly SFEF and SATO). [20: p170-171] The unification of the network led to an increase in overall traffic, symbolised by the creation of the vast ring terminus in Piazza de Ferrari in 1906. [20: p129]
This seems the right time to look again at the ‘ring terminus’ in Piazza de Raffeale Ferrari. ….
In 1908, after three years of construction work, Galleria Certosa (Certosa Tunnel) was put into use. It facilitated tram journeys to and from the Polcevera valley, avoiding the crossing of San Pier d’Arena. [19][21: p38] The tunnel connected Piazza Dinegro, in the port area, to the Rivarolo district in Val Polcevera. It was 1.76 km long. [22]
In 1934, Galleria Certosa was used every day by five lines: Tram No. 9 (San Giorgio-Rivarolo), tram No. 10 (San Giorgio-Bolzaneto), tram No. 11 (San Giorgio-Pontedecimo) and the two circular lines between San Giorgio and Sampierdarena. [22]
Having noted the construction of Galleria Certosa in the early years of the 20th century (above), it is worth looking at some other tunnels which were built to facilitate the movement of trams.
Galleria Vittorio Emanuele III (renamed Galleria Giuseppe Garibaldi on 27th November 1943)
There seems to be quite a story to the life of this tunnel! The first two photographs show the first tunnel. They focus on the portal in Piazza Della Zeccan.
These next two photographs show the tunnel as it was first widened in the form which preceded the Galleria Vittorio Emanuele III which had a much smaller bore.
Named after Christopher Columbus, whose house was nearby, the gallery was opened to the public in the 1930s and was hailed as the city’s gateway to the sea. It connected Piazza de Ferrari and Piazza della Vittoria.
Now long gone, there was a tram tunnel on Via Milano to the Southwest of the city centre. It took the tramway (and roadway) under San Benigno Hill. It was.built in 1878 by the Compagnia Generale Francese dei Tramways for its horse-drawn trams. Its Southwest portal was in Largo Laterna. Its Northeast portal is shown in the first image below.
In the early years of the 20th century, the municipal administration began to consider the idea of taking control of the tram service. In anticipation of this, in 1913, it built its own line from Marassi to Quezzi, known as Municipal Line A, it was operated by UITE on behalf of the Municipality. [19][21: p44]
Before the start of World War 1, the tram network provided these services: [19]
21 De Ferrari – Manin – Staglieno 22 De Ferrari – Manin 23 De Ferrari – Manin – Castelletto 24 De Ferrari – Manin – Castelletto – San Nicholo 25 Circuit in the hilly suburbs 26 Piazza Principe – Corso Ugo Bassi 27 De Ferrari – Zecca – Principe 28 Caricamento – De Ferrari – Galliera ‘Ospital 29 De Ferrari – Carignano 30 Circular Raibetta – Brignole – Corvetto – Raibetta 31 De Ferrari – Staglieno – Molassana – Prato 32 De Ferrari – Staglieno – Molassana 33 De Ferrari – Pila – Staglieno 34 Staglieno – Iassa 35 Pila – Staglieno 36 Pila – Staglieno – Molassana 37 De Ferrari – San Fruttuoso 38 De Ferrari – Foce 39 De Ferrari – San Francesco – Sturla – Priaruggia – Quinto – Nervi 40 De Ferrari – San Francesco – Sturla – Priaruggia – Quinto 41 De Ferrari – San Francesco – Sturla – Priaruggia 42 De Ferrari – San Francesco – Sturla 43 De Ferrari – Villa Raggio – Lido 44 De Ferrari – Tommaseo – San Martino – Borgoratti 45 De Ferrari – Tommaseo – San Martino – Sturla 46 De Ferrari – Tommaseo – San Martino 47 De Ferrari – Tommaseo 48 Raibetta – Pila
III. Municipal line:
A De Ferrari – Quezzi
The Western Network, particularly before World War One
Lines 1 to 11 constituted the Western Network. All of these lines had their city centre terminus at Piazza Caricamento. The Piazza is shown on the adjacent 1916 map.
The map shows part of the Port area of Genoa (Genova) in 1916 with a significant series of standard-gauge railway sidings in evidence (black lines) and some red lines which indicate the metre-gauge tram routes. Piazza Caricamento is close to the water halfway down the map extract. [31]
There were three main routes out of Piazza Caricamento, one of which followed the coast round to meet the lines on the East of the city. The other two shared the bulk of the services leaving the piazza. One of these two routes ran West through San Pier d’Arena (Sampierdarena), the other ran through Galleria Certosa.
The route to San Pier d’Arena (Sampierdarena) closely follows the coast and ran through the Galleria on Via Milano before the San Benigno Hill was raised to the ground.
Pictures of the Galleria can be seen earlier in this article.
West of the Galleria, the original tramway ran along what is now Via Giacomo Buranello (what was Via Vittorio Emanuele) to Sampierdarena. This route appears to the North of the SS1 on the satellite image below.
Before looking at line further West from Sampierdarena we need to note a line which was added to the network before WW1.
A second tramway was built which ran alongside the railway sidings on what is now the SS1, it was then Via Milano, towards Sampierdarena. The route is illustrated by the mid-20th century view below.
That route along Via Sampierdarena (Via Milano and Via Colombo) and then Via Pacinotti is illustrated at the bottom of the map below. After running along the centre of Via Sampierdarena, trams turned inland, heading Northwest to join the earlier route, West of Piazza Vittorio Veneto on Via Pacinotti.
A map provided by the Marklinfan.com Forum which shows the new coastal tram route mentioned above. [92]
The Western Network’s Coastal Line(s)
At Sampierdarena the original lines of the Western network separated. Some lines continuing along the coast and others turning inland. The lines diverged at the West end of Piazza Vittorio Veneto. The coastal line ran along what is today Via Frederico Avio, then turned onto what is now Via Antonio Pacinotti, before turning West on what is now Via Raffaele Pieragostini, crossing the River Polcevera at Ponte di Cornigliano, running along Via Giovanni Ansaldo before joining Via Cornigliano at Piazza Andrea Massena.
We have followed the Western Network as far as we can along the coast. We now need to look at the line(s) of the Western network which ran up the valley of the River Polcevera from Sampierdarena.
To do this we need to return to Piazza Vittorio Veneto in Sampierdarena.
The Western Network and Val Polcevera (the Valley of the River Polcevera)
The lines to the North left Piazza Vittorio Veneto at its Western end, passing immediately through an underpass under the FS Standard Gauge railway.
In the 19th century the route was known as ‘Via Vittorio Emanuele’. In the early years of the 20th century the road was renamed ‘Via Umberto 1’. In 1935, the city gave the road the name ‘Via Milite Ignoto’ (the Unknown Soldier). This decision appears to have been short-live as very soon the road was divided into two lengths, the more southerly length becoming ‘Via Martiri Fascisti’, the remaining length, ‘Via delle Corporazioni’. After the end of Word War Two renaming again occurred. In 1945 the names which continue to be used in the 21st century were chosen – ‘Via Paolo Reti’ and ‘Via Walter Fillak ‘. Fillak and Reti were partisans in WW2. [59][66]
A view from above … This is Piazza (Via) Vittorio Emanuele seen from the West. The tram tracks can be seen heading away through the underpass in the foreground. [75]
The route of this part of the old tramway network begins at this rail underpass (where the street is now named, ‘Piazza Nicolo Montano’, having once been Via Nino Bixio), [65] before running along Via Paolo Reti and then Via Walter Fillak. Just beyond the underpass the railway station access left the road on the left. The first old postcard views below show this location.
Two pixelated, low definition images showing the bottom end of what was Via Umberto 1. One the left in both images is the incline leading to the Sampierdarena Railway Station forecourt. [59]A tram sits at a stop at Piazza Montano. This image was shared on the Foto Genova Antica Facebook Group by Annamaria Patti on 22nd May 2022. [3]Three further postcard views, of better quality, of the bottom end of Via Umberto 1, (c) Public Domain. [59][62][63]The view to the Northeast from the rail underpass in 2024. The station approach is on the left. The old tramway curved round to the left below the station approach’s retaining wall. [Google Streetview, August 2024]Just a little further along the old tram route. The retaining wall on the left supports the station approach road. The tramway ran on along what is now Via Paolo Reti. For some distance the road was flanked by a retaining wall supporting the FS standard-gauge railway. [Google Streetview, August 2024]
The adjacent Google satellite image shows roads over which the old tramway ran. In the bottom right is Piazza Nicolo Montano. It is also possible to make out the station approach ramp which has a number of cars parked on it. In the immediate vicinity of the passenger railway station, railway buildings can be seen separating Via Paolo Reti from the railway but very soon the road and the railway run side-by-side with the railway perhaps 2 to 3 metres above the road. Via Eustachio Degola passes under the railway just to the North of the station buildings. Towards the top of the satellite image, Via Paolo Reti can be seen turning away from the railway wall. [Google Maps, December 2024]
Via Paolo Reti (the former Via Umberto 1) turns away from the railway wall which is now much lower than it was near the station buildings. [Google Streetview, August 2024]Via Umberto 1, looking North from the bend visible in the photograph above where the road leaves the side of the railway, (c) Public Domain. [68]Via Paolo Reti (once Via Umberto 1) at the same location as the monochrome image above. [Google Streetview, August 2024]
The monochrome image below purports to show Piazza San Marino. As far as I can work out the piazza was historically, ‘Piazza Vittorio Emanuele III’ and later renamed for another partisan from World War 2 – ‘Piazza Ricardo Masnata’.
A relatively low quality image of Piazza San Marino and Via Umberto 1. The piazza later became Piazza Ricardo Masnata. This view looks North with a tram visible on the left, (c) Public Domain. [64]Piazza Ricardo Masnata, looking North. There is little to link this image from 2024 with the monochrome image above, other than the alignment of the roads and the shape of the piazza. However, at the centre of this image is a lower building which also appears in the monochrome image. [Google Streetview, August 2024]Via Umberto 1 looking North from what became Piazza Ricardo Masnata, (c) Public Domain. [67]The same location in the 21st century. [GVia Umberto 1, now Via Walter Fillak with a tram heading towards Genoa. [69]The same location on Via Walter Fillak in the 21st century. [Google Streetview, August 2024]
The line from Sampierdarena ran towards Certosa where, once Galleria Certosa was completed, it met the line through the tunnel.
The tramway followed Via Celesia through Rivarolo (Superior). Rivarolo and Via Celesia can be seen at the bottom of this extract from openstreetmap.org. [79]
This image from the early 20th century looks North along Via Celesia. Space on the street was clearly at a premium! [80]Via Celesia in the 21st century. [Google Streetview, August 2024]
North of Via Celesia, the tramway ran along Via Rivarolo.
This postcard shows the junction at the North end of Via Celesia, circa. 1920s. Via Rivarolo is ahead. This image was shared on the C’era una volta Genova Facebook Group by Mario Vanni on 18th August 2019, (c) Public Domain. [82]The smae location in the 21st century. [Google Streeetview, August 2024]This next extract from openstreetmap.org shows Via Rivarolo entering bottom-left. Trams ran on into Teglia on Via Teglia and continued on to Bolzaneto (in the top-right of this extract) along Via Constantino Reta. [79]This postcard view looks South along what is now Via Teglia (then Via Regina Margherita. This image was shared on the C’era una volta Genova Facebook Group by Elio Berneri on 19th October 2020, (c) Public Domain. [83]A very similar view at the same location in the 21st century. [Google Streetview, August 2024]Car 906 in service on line 7 Caricamento – Pontedecimo, one of the longest of the UITE, is seen here running in Bolzaneto. The photograph was taken facing North. In the background you can see another Tramcar, as well as a third on the track in the opposite direction, (c) Public Domain. [84]A similar North facing view in Bolzaneto in the 21st century. [Google Streetview, August 2024]A tram waits at Piazza del Municipio in Bolzaneto. This image was shared by Mario Vanni on the C’era una volta Genova Facebook Group on 8th July 2021, (c) Public Domain. [85]A very similat view of the same location in the 21st century. The road on which the bus is standing is now known as Via Pasquale Pastorino. [Google Streetview, August 2024]A few hundred metres to the Northeast is the area known as ‘Bratte’. A tram waits in the mid-20th century to set off for Caricamento. This image was shared on the C’era una volta Genova Facebook Group by Roberto Della Rocca on 12th December 2020. [86]A similar view at the same location in the 21st century. [Google Streetview, August 2024]
North of Bratte, Trams crossed the River Secca, a tributary of the Polcevera, following Via Ferriere Bruzzo and then continued North alongside the River Polcevera on Via San Quirico.
Tram No 79 leads a trailer car South on Via San Quirico in the first decades of the 20th century. It seems as though Ponte Tullio Barbieri can be seen behind the tram. This image was shared by Sergio De Nicolai on the C’era una volta Genova Facebook Group on 21st October 2018. [88]A similar location on Via San Quirico in the 21st century. [Google Streetview, July 2022]
Trams passed under the FS Standard-gauge lines close to Ponte Tullio Barbieri. [Google Streetview, June 2022]Trams ran on through the centre on San Quirico on Via San Quirico.Before returning to the side of the river, passing under the railway again. [Google Streetview, July 2022]
The next length of the journey is the last. Trams terminated at Pontedecimo. [79]
A tram and trailercar on Lungo Polcevera in Pontedecimo close to Pontedecimo Railway Station, This image was taken looking South along the river bank and was shared by Giorgio Gioli on the C’era una volta Genova Facebook Group on 4th November 2020. [89]This view looks South along the bank of the River Polcevera at a location similar to that in the image above. [Googler Streetview, January 2021]
The central piazza in Pontedecimo. The terminus of the tram service. This image was shared on the C’era una volta Genova Facebook Group by Roberto Cito on 29th October 2023. [87]Trams terminated in Pontedecimo. [Google Streetview, July 2022]The tram depot at Pontedecimo. This image was shared on the C’era una volta Genova Facebook Group by Alessandro Lombardo on 30th October 2019. [90]
The featured image at the head of this article shows BR No. 54445 with an permanent way train passing Culloden Moor Viaduct travelling towards Aviemore. [54]
The Inverness and Aviemore Direct Railway was built by the Highland Railway to provide a shorter and more direct route between Inverness and Aviemore, carrying its main line traffic to Perth and the south.
Earlier articles about the Highland Railway can be found here, [7] here, [8] here, [9] here, [10] and here. [11]
The original route via Forres and Dava, built by the Inverness & Perth Junction Railway (I&PJR), “ran over wild and remote terrain as far as Aviemore, and then on to Dunkeld. From there trains used the Perth and Dunkeld Railway to Stanley Junction, and from there the Scottish North Eastern Railway to Perth. This was a considerable improvement [over the only previously available route via Aberdeen], although operation of the line over the mountainous route was difficult. The traffic from east of Forres proved to be lighter than anticipated, and at the same time traffic from Inverness and from the Inverness and Ross-shire Railway became increasingly dominant. The deviation to Forres before turning south was now a serious liability.” [1]
It was also clear that the Great North of Scotland Railway (GNSR) was planning its own independent line between Elgin and Inverness and the West Highland Railway was known to be considering a line along the Great Glen connecting from Fort William to Inverness. “The Highland Railway was alarmed at both of these competitive encroachments into what it considered to be its own territory. It anticipated that Parliament would look favourably on them, if it could be shown that the Highland Railway was not taking adequate steps to improve its own line and its service to passengers and goods customers.” [1][2: p44][3: p103-104]
“The solution was a new line of 34 miles running directly south from Inverness, rejoining the existing Perth line at Aviemore. This became the Inverness and Aviemore Direct Railway, informally known as the Carr Bridge line, or later the Carrbridge line. Its authorising Act of Parliament was passed on 28th July 1884.” [1][3: p104][4: p175]
As Acworth noted in 1890, the Highland Railway “could never face a Parliamentary Committee and maintain that the existing facilities to Inverness were sufficient, when it had taken no steps to supply the additional accommodation whose necessity it had itself asserted only a few years before … The construction of the new road will mean to [the Highland Railway], in the first place, a capital expenditure of some hundreds of thousands of pounds; secondly the cost of working some thirty additional miles; thirdly no additional traffic whatever; and lastly, the reduction of the passenger fares by as many pence as the new road will be shorter in miles than the old.” [1][5:p74-75]
“Having received the authorisation, the Highland Railway did nothing to hasten actual construction, no doubt believing that the danger of encroachment had been staved off. In any event, for the Highland Railway this was the most important development of the decade. … The cut-off was 34 1⁄2 miles of new line between Aviemore and Inverness. For six years from obtaining the necessary Act on 28 July 1884, the company managed to stave off any real action, although by 1886 agreements about land acquisition were made with proprietors. Altogether four extensions of time to complete the line were granted: two before and two during construction.” [1][2: p44][3: p103-104]
The Inverness to Aviemore Direct Railway was opened in stages: the first, from Aviemore to Carr Bridge, opened on 8th July 1892 as a branch line operated by a tank engine, and carrying very little traffic. [1][3: p103-104] The line from Carr Bridge to Daviot opened on 19th July 1897.
The route was completed for through running by the opening between Daviot and Millburn Junction, Inverness, on 1st November 1898. [1] Ot should be noted that there is some ambiguity over the dates. [1: Note 1]
In October 1897, it was decided to install double track on the as-yet unopened section between Inverness and Daviot. This involved widening some completed single-track bridges. [1][2: p8]
“The new line incorporated the Highland Railway’s second-highest summit: Slochd at 1,315 feet. The Strathnairn Viaduct near Culloden Moor is Scotland’s longest masonry viaduct at 600 yards in length; there are 29 arches. [1][3: p133-134][4: p197] Major enlargement of the track facilities was also carried out at Millburn Junction in Inverness. The total cost of the line was almost a million pounds.” [1][2: p46]
From the opening of the direct line from Aviemore… “the traffic planners had to cater for two main lines into Inverness from the south. In the summer of 1909, seven scheduled trains ran each day between Perth and Inverness. The night train from Perth left at 12.50 a.m., with sleeping car from Glasgow, and travelled via Carrbridge, arriving at 5.10 a.m. A connecting train left Aviemore for Forres at 4.00 a.m., arriving also at 5.10 a.m. Nairn passengers went on to Inverness and changed trains there. At 5 a.m. another train left Perth, conveying sleeping cars from London and through carriages from southern railways, running via Carrbridge and arriving at Inverness at 8.35 a.m. This ran only from 1 July to 11 August. Fifteen minutes later the ‘normal’ night train from London left Perth, and arrived in Inverness at 9.08 a.m. This train was also noted as conveying Sleeping Carriages Euston to Strathpeffer.” [3: p184-185]
“A Forres connection left Aviemore at 8.25 a.m., arriving at 9.35 a.m. The Mail left Perth at 6.15 a.m. and reached Aviemore at 8.33 a.m. Here it divided, the direct Inverness portion arriving at 10.10 a.m., and the Grantown portion arriving in Inverness at 11.15 a.m. A Saturdays-only train left Perth at 9.25 a.m., reaching Inverness at 1.50 p.m.; its Forres connection left Aviemore at 12.45, arriving 1.56 p.m. A train for Inverness via Forres still left Perth at 11.50 a.m., running non-stop to Newtonmore, which it reached at 1.44 p.m.; Forres was reached at 3.25, and Inverness at 4.15. Only ten minutes later, the old Parliamentary left Perth, stopping at all stations (five on request only) and reaching Inverness via Carrbridge at 4:36 p.m.” [3: p185]
The Route
Inverness Railway Station was covered in the first article in this series. [7]
A single extract from the 6″ Ordnance Survey undertaken at the turn of the 20th century will suffice here. Note the Lochgorm Works at the top of the extract on either side of the loop which allowed East/West movements without trains needing to enter the station. The locomotive facilities centred on the roundhouse which can be seen on the right side of the extract. [12]This 21st century ESRI satellite image, provided by the NLS shows that non-rail uses now sit over the site of the old locomotive shed and turntable. The basic layout of the railway infrastructure remains as does part of the Works. [12]Heading East from Inverness Railway Station, two lines ran in parallel. Somewhat counter-intuitively the lines to Forres and beyond ran on the South side of the lines which will bear away South. [13]Both lines continue to be used in the 21st century, although it is difficult to make out the detail on this extract from the ESRI satellite imagery provided by the NLS. [13]This next extract from the 6″ Ordnance Survey of 1901 shows the more northerly set of lines rising up to bridge the lines to Forres. [14]On this next extract from the ESRI satellite imagery it is slightly easier to make out the two lines. [14]This extract from Google Maps shows the two lines crossing as they both pass under the modern A9. [Google Maps, June 2025]Looking West from the Harbour Road Crossing along the Forres line towards Inverness Railway Station. [Google Streetview, 2022]Looking East along the Forres line from the Harbour Road Crossing. [Google Streetview, 2022]Harbour Road looking North. The Inverness to Aviemore Direct Railway is bridging the road. [Google Streetview, 2022]This time looking South through the bridge. [Google Streetview, 2022]
The Inverness to Aviemore Direct Railway curves round to the Southeast.
Away from the coast line, the railway is in cutting. [18]The modern satellite image illustrates the growth of Inverness. The presence of the A9 is a significant change to the landscape. [18]Looking back round the curve towards Inverness Railway Station from the A96. [Google Streetview, 2023]Looking ahead along the Inverness to Aviemore Direct Railway from the bridge carrying the A96 over the line. [Google Streetview, 2023]For the first part of this journey along the line, each map extract overlaps with the previous extract. That is true for this and two further extracts. After that just discreet locations are featured. [15]The same area in the 21st century. [15]A further length of the line takes it as far as Caulfield Road North. [16]The same length of line in the 21st century. [16]Caulfield Road North is now National Cycle Route No. 7, this view looks Northeast along the cycleway through the bridge carrying the railway. [Google Streetview, 2022]The line continues East, at the turn of the 20th century its route was through open fields. [17]The same location in the 21st century. [17]
The line begins a wide curve round to the South to cross Culloden Moor. The line is initially in cutting, then on embankment and then in cutting on its approach to the site of Culloden Moor Railway Station.
The view Northeast along the railway from the bridge carrying Tower Road. [Google Streetview, 2022]The view Northwest along Culloden Road. The railway is carried over the road on a stone arch bridge. [Google Streetview, 2022]As the line curves around towards the Southeast it bridges a million nor road by means of another stone arch bridge.b[Google Streetview, 2022]By the time that the line passes under the B9006 it has already run through platform faces of the old Culloden Moor Railway Station to the North side of the road. The line is, by this time, almost on a North-South alignment. This view looks back North along the line from the bridge. [Google Streetview, 2022]Turning through 180°, this view looks South through the old station site towards the Strathnairn Viaduct (alternatively, the Nairn Viaduct or the Culloden Viaduct). [Google Streetview, 2022]Culloden Moor Railway Station as it appears on the 6″ Ordnance Survey undertaken at the turn of the 20th century. [19]The location of Culloden Moor Railway Station in the 21st century. [19]The 25″ Ordnance Survey from the turn of the 20th century. The footbridge was removed when it was realised that the road bridge was perfectly adequate for pedestrian access between platforms. The railway workers’ cottage and the Stationmaster’s House are on the West side of the line. [23]
Culloden Moor railway station served the village of Culloden from 1898 to 1965 (1967 for general goods). It was a two platform station just to the north of the Nairn Viaduct. Its location was closed too to the site of the Battle of Culloden. The platforms remain but the station buildings have gone.
The roadbridge at Culloden Moor Railway Station site, seen from what was once the station forecourt. The site to the East of the railway is, in the 21st century, occupied by Iain Cowie Plant Hire and Groundworks Ltd. [Google Streetview, 2008]
Both platforms at the station had a water column, with the water tank on the northbound platform. There was a bitumen depot adjacent to the station and in later years the goods sidings at the station could be seen filled with bitumen tank wagons. The depot closed towards the end of the 20th century. [22]
Another photograph of the station can be seen here. [25] South facing photographs of Culloden Moor Railway Station have the Culloden Viaduct appearing in the distance.
Culloden Viaduct (Nairn or Strathnairn Viaduct) as it is shown on the 6″ Ordnance Survey from the turn of the 20th century. [26]The same viaduct as it appears on the modern ESRI satellite imagery provided by the NLS. [26]Culloden Viaduct looking North-northwest from the minor road at the South side of the river valley. [Google Streetview, 2022]BR steam locomotive 54445 with up permanent way train passing Culloden Moor Viaduct. It should, of course, be noted that No. 54445 was an ex-Caledonian Class 113 4-4-0 locomotive. These locomotives were classed 3P by the LMS. Introduced in 1916, these locomotives worked to a boiler pressure of 175 psi, had 20″ x 26″ cylinders and driving wheels of 6′ 6″ diameter. The locos without their tenders weighed in at 61 tons 5 cwt and produced a tractive effort of 19,833 lbs. [54][55]
The viaduct was designed by Chief Engineer Murdoch Paterson and built by The Highland Railway. Twentynine arches carry the line over the valley of the River Nairn. It opened in 1889 and it remains the longest masonry viaduct in Scotland at 1800ft (549m) long and is a Category A listed building. [27}
South of the viaduct the railway head for a short distance to the Southwest before turning Southeast as it arrives at Daviot Railway Station.
En-route the line crosses a minor road and
Looking Northeast along the next minor road to be bridged by the line. [Google Streetview, 2022]Looking back towards Culloden from a bridge carrying the National Cycle Route No. 7 over the line. [Google Streetview, 2022]Turning through 180° and looking forward along the line towards Daviot Railway Station with the gorse in full bloom. [Google Streetview, 2022]Daviot Railway Station at the turn of the 20th century. The double -track line from Inverness became a single line to the Southeast of Daviot. [28]The same location in the 21st century. [28]
Daviot Railway Station opened on 19th July 1897.Ot was, for a short time the Northern terminus of the line from Aviemore until Culloden Moor Railway Station opened in 1898. Wikipedia tells us that, “on the northbound platform was the station building and to the southwest was the goods yard. There were two signal boxes: one to the north which was built, but never opened. The other signal box was to the south in between the goods sidings. It was relocated slightly to the north in 1952. The station closed on 3rd May 1965. The signal box closed in 1969. Only the platforms remain.” [29]
Ernie’s Railway Archive has an excellent photograph of the station which can be viewed here. [30]
Moy Railway Station was further to the Southeast and is shown here in an extract from the 6″ Ordnance Survey from the turn of the 20th century. [31]A satellite image of the same area in the 21st century. The A9 is the most significant change visible (running diagonally across the bottom third of the image). [31]
Wikipedia tells that Moy Railway Station opened on 19th July 1897. “The station building was situated on the southbound platform. Goods facilities were handled at the northeast. There were two signal boxes: north and east. Despite their names, they were both situated to the west. The station closed to both passengers and goods traffic on 3rd May 1965.” [32] The station building at Moy was of a very similar design to that at Culloden Moor and Daloit.
AmBaile has a monochrome image of the station at Moy. This image can be seen here. [34]
The modern A9 crosses the line of the railway to the East of Moy. The next two images are taken from the A9 road bridge.
Looking back along the railway from the A9 overbridge towards Moy. [Google Streetview, June 2022]Looking forward towards Tomatin from the A9 overbridge. [Google Streetview, June 2022]
Further Southeast we teach Tomatin Railway Station. …
The next station along the line was Tomatin Railway Station. [33]The same area in the 21st century. [33]
AmBaile has a few photographs of Tomatin station. These can be seen here, [35] here, [36] and here. [37]
Findhorn Viaduct is around 500 metres East of Tomatin village. …
The next significant location on the line is the Findhorn Viaduct which is approximately 500 metres East of the village of Tomatin and carries the line over the valley of the River Findhorn. [38]The same area as shown on the ESRI satellite imagery from the NLS. [38]
“The Findhorn Viaduct was designed and built for the Highland Railway between 1894 and 1897 by Murdoch Paterson, their chief engineer, and John Fowler, who was the consulting engineer and who also worked on the design of the Forth Rail Bridge. The viaduct was Fowler’s suggestion in order to create a more direct route; the railway company had originally planned a more circuitous route around the valley, over a mile longer. The steel for the lattice work was supplied by the Butterley Iron Company in Derbyshire, England. The granite for the piers was supplied by Kemnay Quarry in Kemnay, Aberdeenshire, which also supplied materials for the Forth Bridge. … It was opened to traffic on 19 July 1897. The viaduct is a Category B listed building, first listed in 1971, a status which grants it legal protection.” [39]
The highest point on the line between Inverness and Aviemore is at Slochd. The Slochd Summit “is a mountain pass on the A9 road and the Highland Main Line Railway. It is the highest point on the line between Inverness and Aviemore. An old military road and National Cycle Network Route No. 7 also go over the summit, the latter largely following the old A9. … Both the road and the railway have signs marking the spot – the A9 is at a height of 1,328 feet (405 m), while the railway reaches 1,315 feet (401 m). The Slochd Summit is the second highest place on the route from Inverness to Perth – the Pass of Drumochter at 1,500 feet (460 m) is higher and bleaker.” [41]
Close to Slochd the railway crosses the Allt Slochd Music on a high viaduct. [42]The same area in the 21st century. [42]The listed building record for the Allt Slochd Music Viaduct. [43]
A very short distance to the East of Slochd, the old A9 (now National Cycle Route No. 7 crosses the old railway. …
This photograph is taken from the old A9 road bridge and looks back along the railway towards Slochd. The modern A9 flyover sits above both the old road and the railway. [Google Streetview, June 2022]Looking ahead to the Northeast and Carrbridge from the old A9 bridge. [Google Streetview, June 2022]Bird’s eye view of the railway and old and new A9s at the same location. This image was shared on the Carr-bridge Past and Present Facebook Group by James Ross on 11th June 2024. [52]
East of the old A9 the railway curves round from a Northeasterly direction to a Southeasterly one. It runs down the valley of the Bogbain Burn crossing the stream a number of times on its descent. The modern A9 runs a little to the South of the railway. The Bogbain Burn joins the Allt nan Ceatharnuch and flows to the North of the railway before crossing under both the railway and the modern A9. It then flows into the River Dulnan to the West of the A9 and the railway.
The River Dulnan flows under the A9 and the railway and then flows down through Carrbridge. The railway station sits about 1.5 km Southwest of the village close to the river. It remains open in the 21st century.
An historical photograph of the railway bridge over the river can be seen here. [53]
The River Dulnan Railway Bridge, seen from the A9 to the South. [Google Streetview, 2023]Carrbridge Railway Station at the turn of the 20th century. The Dulnan River features in the top-left of the image. [44]The same area in the 21st century. The modern A9 runs parallel to the railway rather than through Carrbridge. [44]
Carrbridge Railway Station was opened on 8th July 1892 when the Highland Railway opened the line from Aviemore. For five years Carrbridge was the terminus of the line from Aviemore.
Northbound services commenced on 8th July 1897 when the line to Daviot was opened, the line through to Inverness opened on 1st November 1898. … The station was built with a passing loop on the otherwise single track railway, a signal box (automatic token-exchange apparatus was used) and several sidings on the north side of the line. The station building is thought to be by the architect William Roberts, dating from 1898. A camping coach was positioned here by the Scottish Region from 1954 to 1965. [46]
Southeast of Carrbridge Railway Station, the line curves through a heavily wooded landscape towards the South. It continues to be closely followed by the modern A9.
This wintery scene was recorded in March 2023 looking Northeast from the A9. The lack of foliage means that the railway can be seen close to the road. The location is the Knock of Kinveachy. The old A9 is still in use beyond the road as the B9153 and then the A95. [Google Streetview, March 2023]
The A9 begins to turn away from the railway to the South of Kinveachy. The A95 passes under the railway as it heads South, along the route of the old A9.. A new skew bridge carries the railway over the road in the 21st century.
This view looks Northeast along the A95 under the railway bridge. [Google Streetview, March 2023]The location of the old railway bridge over the A9, (the A95 in the 21st century). [47]The same location in the 21st century: the new A9 can be seen in the top-left corner of this satellite image. The A95 (the old A9] can be seen to the right of the railway at the top of the image,then passing under the railway before leaving the image bottom-left. [47]
The railway heads due South towards Aviemore, before drifting towards the West … Three small lochans were in the path of the railway. …
Now on the final run down to Aviemore. The old A9 can be seen on the left of this extract from the 6″ Ordnance Survey from the turn of the 20th century. The railway crosses the small Loch nan Carraigean.[48]This modern image shows that by the 21st century the lochans have almost completely disappeared. Just a small pool remains at the location of Loch nan Carraigean. A quarry has been opened up between the railway and what is now the A95. [48]Aviemore Railway Station at the turn of the 20th century. The River is the River Spey. There was very little to Aviemore at the turn of the 20th century – the Station and Hotel and a few private dwellings. [49]This 21st century satellite image of the same area shows a considerable amount of development to the North of the Railway Station. [49]This wider view shows just how significant the development over 125 years has been. The modern A9 can be seen on the left of this image. [49]
In the 21st century the journey between Inverness and Aviemore takes less than 45 minutes. The older main line through Forres was abandoned as part of the cuts which followed the Beeching Report in the mid-1960s. As we noted when looking at the route via Forres and Dava, a preservation railway is active at the southern end of that line and shares Aviemore Railway Station with ScotRail. The line is followed by walkers and cyclists on The Dava Way. [50]
An article about the Gorres/Dava route can be found here. [8]
The featured image at the head of this article (above) is Highland Railway No. 79, ‘Atholl’, a 4-4-0 Clyde Bogie, built Jun 1886. Renumbered 79A in Mar 1917. Only one of this class survived into LMS days No. 14278. It originally carried the name ‘Fife’ and was ‘Hughland Railway No. 82. Renamed ‘Durn’ in 1900, a name it retained until 1916, it was renumbered 82A in March 1917. It continued in service until being withdrawn in April 1930. [30][31]
Trains Illustrated No. 18 which was published in 1976 focussed on The Highland Railway. [1] The introductory article, ‘Highland Retrospect’, was written by Paul Drew. [1: p4-11]
The Highland Railway (HR) Company was based in Inverness. It was formed by merger, absorbing over 249 miles (401 km) of line. It continued to expand, reaching Wick and Thurso in the north and Kyle of Lochalsh in the west, eventually serving the counties of Caithness, Sutherland, Ross & Cromarty, Inverness, Perth, Nairn, Moray and Banff. Southwards, it connected with the Caledonian Railway at Stanley Junction, north of Perth, and eastward with the Great North of Scotland Railway at Boat of Garten, Elgin, Keith and Portessie. [6][7][17: p29]
The direct route to Perth eliminated the need for a change of trains at Aberdeen (and for some time a half mile journey between two railway stations). “It gave a route virtually under one ownership between Inverness and Perth. The greatest benefit lay, however, in the shortening of the journey. Previously the distance by rail from Perth to Inverness had been 198 miles; it now became 144. This represented a saving of about 24 hours in travelling time, and a corresponding reduction in fares.” [17: p27]
Earlier articles about the Highland Railway network can be found here, [15] here, [3] and here. [16] In this article:
we follow the line from Forres via Dava to Perth
We follow the line from Forres via Dava to Perth.
We look at the Highland Railway’s locomotive superintendents and at least some of its locomotives.
The First Direct Line Between Inverness and Perth (via Forres) – the Dava Line
H. A. Vallance tells us that “The proposed new railway through the Central Highlands left the Inverness & Aberdeen Junction Railway at Forres, 24 miles East of Inverness, and turning sharply to the south, rose steeply for several miles to cross the hills separating the valley of the Findhorn from Strathspey. Beyond the summit at Dava, 1,052 ft above sea level, the approximate course of [a] route surveyed … in 1845 was joined, and the line proceeded by way of Grantown and the west bank of the Spey to Kingussie. There followed a long, steep climb up the northern slopes of the Grampians to the head of the Druimuachdar Pass, and a corresponding descent to Blair Atholl and the Pass of Killiecrankie. Crossing the Tay near Dalguise, the line reached Dunkeld, where it made an end-on junction with the Perth & Dunkeld Railway.” [17: p24]
The enabling Act of Parliament was passed in July 1861 and authorised the Inverness & Perth Junction Railway to take control of the Perth & Dunkeld Railway. That amalgamation took place in February 1864.
Work on the line commenced in October 1861. The first 13 miles (Dunkeld-Pitlochry) opened at the beginning of June 1863. The length from Forrest to Aviemore opened in early August of the same year. The last length (Aviemore-Pitlochry) was completed in September, just a month or so later. Vallance notes: “The chief engineering works on the line were the seven-span masonry viaduct over the Divie, near Dunphail, 477 ft long and 105 ft high; the girder bridge across the Tay, near Dalguise, 515 ft long and 67 ft high; and the ten-span masonry viaduct, 54 ft high, over a deep ravine in the Pass of Killiecrankie. Smaller masonry bridges were required to carry the railway over the Bran, north of Dunkeld, and the Garry, at Struan. The latter was remarkable in that its main span crossed not only the river, but also the stone bridge carrying the Rannoch Road over the Garry at the same point. At Blair Atholl the Tilt was crossed by a single-span lattice-girder bridge. … Only two short tunnels were found necessary-one in the Pass of Killiecrankie (128 yd long), and the other north of Dunkeld (350 yd long). South of Dunkeld there was a third tunnel (310 yd long), but this had, of course, been constructed by the Perth & Dunkeld Railway. (These, incidentally, were the only three tunnels on the Highland Railway.)” [17: p25]
“Some time elapsed before all the stations were ready to be opened, a few of the smaller ones being left until communication between Forres and Perth had been established. Thus accommodation was not provided at Dava until 1864, and Killiecrankie and Dalnaspidal until 1865.” [17: p25]
We have already looked at the line from Keith to Inverness. The original station at Forres was not suited to becoming a significant junction station. Valance provides a sketch of the revisions at Forres. …
A sketch of the new Station and railway junction at Forres. A double junction was provided with the line from Inverness to Keith, so that trains could be run in each direction without reversing. To give each train platform accommodation, the new station was triangular in form and built to the south of the existing one. A new line was provided for the trains from east to west, leaving the original route as an auxiliary loop for goods trains and light engines. The old station building became the stationmaster’s residence. [17: p26-27]The junction at Forres as it appeared on the 25″ Ordnance Survey of 1868, published in 1870. The original alignment of the Inverness &Aberdeen Junction Railway runs diagonally across the top of the map extract. The old station was in the top right quadrant. The new triangular junction and accompanying station/platforms is at the centre of the image. [18]The same area in the 21st century as depicted on the ESRI satellite imagery from the NLS. [19]
Before following the route below, you may well enjoy watching a video about The Dava Way. … [28]
South of Forres, the line ran close to the Dallas Dhu Distillery which had its own sidings, before climbing to Dunphail Railway Station.
A short distance Southeast of Dunphail Railway Station, the line crossed the Divide Viaduct (over the river of the same name). The viaduct still stands in the 21st Century and carries the Cava Way footpath.
Some distance further South the line reached Dava Railway Station. … The NLS records of the 35″Ordnance Survey do not cover the line heading South, so we now use the 6″ Ordnance Survey. …
Dava Station, as recorded on the 6″ Ordnance Survey from the turn of the 20th century. [25]The same location in the 21st century, as recorded on the ESRI satellite imagery provided by the NLS. [25]
Wikipedia tells us that, “At 985 feet above sea-level, Dava was the third highest station on the Highland Railway network. The summit that followed south of the station (Dava Summit) reached 1,052 feet. … The station was located in sparsely populated moorland, along a dirt track near the junction of the A939 and A940 (OS Grid Reference NJ008389). The purpose of a station at this location was to provide a passing loop with water columns (15 miles from Forres and 16 miles to the first junction at Boat of Garten), although the station did provide some facilities for the area, such as a post office. … The surrounding area is wild moorland (The Dava Muir).” [26]
Goods services at Dava were first to end; the date of closure was 27th January 1964. A passenger service continued until 18th October 1965, when the station closed completely. The station building, stationmaster’s house, and platforms still remain. The stationmaster’s house is a private residence.
South of Dava the line ran almost due South and continues to provide the formation for the Dava Way [43] which reaches a summit between the locations of Dava Railway Station and Castle Grant Platform.
The 6″ Ordnance Survey shows the old military road and the railway line in close order. [45]Railmaponline.com shows the route of the old railway in green approaching Dava Way Summit. This is also the long distance footpath known as the Dava Way. [42]The route of the old railway runs alongside the A939 over Dava Summit an area of open moorland. In this South facing view, the fenceline beyond the Scots Pine marks the route [Google Streetview, June 2023]The route of the railway and that of the Old Military road sit side-by-side. [44]Further South the line of the old railway is very close to the modern road. [42]Looking Southeast from the A939, The fence line marks the line of the old railway. [Google Streetview, June 2023]
The line snaked back and forth following the contours as it travelled South.
The adjacent map shows two bridges close together and both associated with the Castle Grant Estate. The first we encounter travelling South along the line is a bridge carrying an estate road over the old railway. A picture of that bridge can be seen here. [50]
The adjacent map shows two bridges close together and both associated with the Castle Grant Estate. The first we encounter travelling South along the line is a bridge carrying an estate road over the old railway. A picture of that bridge can be seen here. [50] The second bridge is shown below.The old railway bridged the Old Military Road on its way down towards Grantown-on-Spey. [46]This extract from railmaponline.com shows the same area in the 21st century. [42]The bridge which carried the old railway over the Military Road remains in place carrying the Dava Way over the A939. This is the view looking South along the line of the old Military Road (A939). [Google Streetview, June 2023]The same bridge seen looking North. The building adjacent to the railway combined as a station building and a lodge for the Castle Grant estate. There was a platform on the East side of the line included at the railway’s expense as a private platform for the estate.At Lynmacgregor the railway bridged a minor road connecting the small hamlet to the Old Military Road. [48]
The same location is shown here on railmaponline.com’s satellite imagery. A length of embankment has been removed to allow a better alignment of the access road to Lynmacgregor. [42]
Close to what is now Grantown-on-Spey Caravan Park the railway crossed another minor road by means of a stone arch.
The area to the East of this bridge is now a Caravan Park. [47]The same location in the 21st century. [42]The view of the West face of the stone-arched bridge. [Google Streetview, August 2021]Grantown West Railway Station sat to the Southwest of the village. This was one of two stations serving the town. Grantown East Railway Station on the Great North of Scotland Railway (Strathspey Section) was on the South side of the River Spey at Speybridge. [49]
Photographs of the Station while in use can be found here [52] and here. [53]
This extract from the Ordnance Survey 1 inch to 1 mile 3rd Edition (surveyed between 1866 to 1870; revised: 1906; and published: 1908) shows the relative positions of the town/village and its two stations. [54]
This extract from railmaponline.com’s satellite imagery shows the location of Grantown West Railway Station in the 21st century with the line of the old railway superimposed. The village/town of Grantown is to the Northeast of this location. The station site is now an industrial estate. [42]
Looking South-southwest through the industrial estate which sits on the site of Grantown West Railway Station. [Google Streetview, August 2021]Looking North-northeast through the industrial estate which sits on the site of the old station. [Google Streetview, August 2021]
The Strathspey Steam Railway has an ambition to extend its line to the site of the old station. [55]
South of Grantown-on-Spey West Railway Station the line continued on embankment before crossing a minor road and stream as shown on the map extract below.
This extract from the OS 6″ mapping shows the location of the bridge mentioned above and shown in the photograph below. [56]The structure shown on the map extract above, seen from the East. A single span crosses both the road and the watercourse. [Google Streetview, August 2021]The line continued in a South-southwest direction. Its route is easy to make out on modern satellite imagery. [Google Maps, June 2025]The modern A95 crosses the line of the old railway which can be seen curving round towards the Southwest. [Goole Maps, June 2025]
This extract from the 6″ Ordnance Survey mapping which was surveyed in 1868 and published in 1873 shows the esrtwhile bridge which carried the railway over what was to become the A95. As can be seen above, the closure of the line permitted the road alignment to be considerably improved. [57]
A short distance further Southwest, the old line crossed the River Dulnain close to its confluence with the River Spey as this next extract from the 6″ Survey mapping surveyed between 1867 and 1871 and published in 1873/75 shows. [58]
The line continued on through Broomhill Station before it was met by the Great North of Scotland (Strathspey Section) line which had just bridged the River Spey. [59]
Broomhill Railway Station or Broomhill for Nethy Bridge Railway Station [60: p46] is a reconstructed railway station on the former Highland Railway main line [61] which was originally built to serve the small villages of Nethy Bridge and Dulnain Bridge in Strathspey. It is at present the eastern terminus of the heritage railway, the Strathspey Steam Railway. [55][62]
Boat of Garten railway station was a significant junction on the Highland Railway, and is now the headquarters of the Strathspey Railway. The station served as a link between the Highland Railway’s main line (Perth to Forres) and the Great North of Scotland Railway’s branch to Craigellachie. Today, the Strathspey Railway operates heritage steam trains between Aviemore, Boat of Garten, and Broomhill, using part of the original Highland Railway line.
In the 21st century, Aviemore Railway Station is owned by Network Rail and managed by ScotRail, is on the Highland Main Line, 83 miles 31 chains (134.2 kilometres) from Perth, between Kingussie and Carrbridge, and is also the southern terminus of the Strathspey preserved railway. [68]
The modern main line to the North of Aviemore, the later Highland Railway main line, will be covered in Part 3 of this series of articles.
South of Aviemore, the Highland line continued more than 83 miles South to Perth.
The line follows the River Spey southwards, passing to the East of Loch Alvie, through Kincraig Railway Station and then passing to the West of Loch Insh and through Kingussie Railway Station.
Southwest of Kingussie the railway continues to follow the River Spey. It bridges the Spey southwest of Newtonmore Railway Station and just to the East of Spey (Road) Bridge and then continues climbing alongside first the Spey and then a tributary, the River Truim, towards Dalwhinnie Distillery.
Beyond the confluence of the Spey and the Truim the railway runs Southwest up Glen Truim towards Dalwhinnie Distillery. Just a short distance Northeast of the Dalwhinnie Distillery the Truim splits into two. Google Maps records the name of the two arms being the River Truim. Both arms of the Truim pass under the railway.
It is worth noting before we travel on towards Perth that one of the most significant improvements to the journey to Perth “came with the doubling of sections of line, designed by the engineer Alexander Newlands, beginning with Blair Atholl to Dalnacardoch (8.25 miles (13.28 km)) in 1900, extended to Druimuachdar (8.5 miles (13.7 km)) in 1901 and Dalwhinnie (5.5 miles (8.9 km)) in 1909. [83] In the 1960s, many sections of the line were converted from double track to single track. In 1976, 23 miles (37 km) from Blair Atholl to Dalwhinnie was redoubled. [84][85] In March 2019 Network Rail completed a programme of works to increase capacity on the line and support the introduction of InterCity 125 sets on ScotRail services, with passing loops and platforms extended.” [82][86]
A few kilometres South of Dalwhinnie, the railway crosses a minor road just to the West of its junction with the A9,then crosses a small tributary of the River Truim, before it crosses the River Truim again.
Looking South from the minor road crossing. [Google Streetview, 2024]The bridge over the River Truim. [87]The same location in the 21st century. The unfenced road shown on the map extract is now the A9 trunk road [87]The same bridge, seen from the A9. [Google Streetview, 2023]
Continuing southwards the line crosses the River Truim again. …
The next bridge over the river. [88]The same location in the 21st century. [88]The best view of the bridge from the A9. [Google Streetview, 2024]
The line crosses the watershed and begins the drop down towards Perth.
The next bridge crosses the first of a series of streams which form the headwaters of the River Garry. [89]The same location in the 21st century. [89]The bridge on the map and satellite image above. [Google Streetview, 2024]
Over the next kilometre a number of such streams are crossed and the line passes through the location of Dalnaspidal Railway Station which closed in 1965.
Killiecrankie Railway Station in the years before closure. [110]
To the Southwest of Killiecrankie Railway Station the line passed through the Pass of Killiecrankie. First there was a short tunnel, then a viaduct over a tributary of the Tummel.
The tunnel can be seen to the top-left of this extract from the 6″ Ordnance Survey. The viaduct is towards the centre of the image. [103]The viaduct can easily be seen on this extract from the ESRI Satellite imagery from the NLS. [103]The East tunnel mouth sits immediately above the Northwest end of the viaduct. [111]The Viaduct sits alongside the River Garry. A plaque sits beneath the viaduct which reads: The National Trust for Scotland: Railway Viaduct – This 510′ long-viaduct, which was designed for the Inverness and Perth Junction (The Highland Railway from 1865) by Joseph Mitchell was completed in 2864;at a cost of £5,730. The ten masonry arches are each 35′ span and the extreme height to the parapet wall is 54′. [111]
The railway has followed the River Garry over some distance. The confluence of the River Garry with the River Tummel is just a short distance South of the Pass of Killiecrankie. From that point the railway follows the River Tummel through Pitlochry.
To the Southeast of Pitlochry the main line of the Highland Railway is met by the Aberfeldy branch which has just crossed the River Tummel. The Aberfeldy branch ran along the South side of the River Tay from Aberfeldy to close the confluence of the Tummel and the Tay. It bridged the Tay and soon also bridged the Tummel as shown in the OS extract below.
The River Tummel joins the River Tay a short distance South of Balliinluig Junction and the railway continues down the valley of the Tay, through what was Guay Railway Station and bridging the River Tay a little further South.
Images of Dalguise Railway Station can be seen here, [121] here [122] and here. [123]
Beyond Dalguise, a tunnel opposite Dunkeld House across the Tay, a bridge over the River Braan, a significant embankment and a length of cutting bring the line to a point across the Tay from Inverness. The line continues through the railway station at Birnam which serves Dunkeld and Little Dunkeld.
Opposite Dunkeld House and on the South side of the Tay, the line passes through a short tunnel and over the River Braan. [109]The same location on the 21st century ESRI satellite imagery from the NLS. The bridge is easier to make out than the tunnel to its West.[109]
Historic Environment Scotland describes Inver Viaduct, the bridge over the River Braan, as being a “large single segmental arch, flanked by crenellated refuges [with] large masonry approaches pierced for footpaths, octagonal flanking features.” [124]
Beyond Dunkeld and Birnam Railway Station, the line continues along the South side of the River Tay. Rail and river separate to make room for Murthly Castle. The line, first in tunnel and then skirting the South side of the Murthly estate.
“Murthly Castle has been owned by three families since the original tower house was built in 1450. The Stewarts of Grantully acquired the property in 1615 and the present owner is Thomas Steuart Fothringham, who lives at the property with his family. The castle was developed gradually over a 450-year period, with the latest addition dating from 1893. Consequently, there is a wide range of architectural styles both inside and outside.”[127]
The 6″ Ordnance Survey of 1901 shows the Murthly estate, the River Tay and the railway. The tunnel can be seen bottom-left of this map extract. [128]A similar area in the 21st century. [128]Murthly Village, Station and Asylum in 1901. [129]The same location in the 21st century. The station closed in 1965. [129]Murthly Railway Station close to the turn of the 20th century. [130]Abandoned signal box and level. Rossing at the closed Murthly Railway Station site. [131]The Highland Railway (HR) line met the Caledonian line at Stanley Junction. HR trains covered the remaining miles into Perth on Caledonian metals. [132]The same location in the 21st century. [132]
Stanley railway station was located around half-a-mile Southwest of Stanley Junction Station. It was opened in 1848 by the Scottish Midland Junction Railway which linked Perth and Arbroath. It closed in 1857, when the new station of Stanley Junction was built at the location where the Perth and Dunkeld Railway diverged from the Scottish Midland Junction Railway running between Perth and Arbroath. Stanley Junction Station closed to passengers in 1956 and to goods in 1965. [133][134]
Stanley Junction Railway Station. This image was shared on the Perth & Kinross Archive Facebook Page on 9th November 2022. [135]
Highland Railway trains ran through to Perth on the Caledonian Railway under an arrangement agreed by predecessor companies (the Perth and Dunkeld Railway and the Scottish Midland Junction Railway). The line ran through Stratford and Luncarty stations, before passing through Almond Valley Junction before the final approach to Perth General Railway Station.
Strathord Railway Station. [136]The same location in the 21st century. [136]Luncarty Railway Station and Bridge. [137]The same location in the 21st century. [137]Almond Valley Junction in 1901. Lines from Crieff and Methven met up the valley of the Almond. [138]The branch line’s route is marked by the narrow line of trees curving towards the top-left of this extract from the ESRI satellite imagery from the NLS. [138]The approach to Perth from the North in 1901. [139]The same area in the 21st century. [139]Looking North from the bridge carrying Long Causeway over the railway. Nothing remains of what was once a busy railway scene. [Google Streetview, 2022]Looking South from the same bridge. The bridge in the distance appears on the next map extract and satellite image. [Google Streetview, 2022]The final approach to Perth from the North in 1901. [139]The final bridge over the line before Perth Railway Station appears just below centre in both the map extract above and this satellite image. [139]Looking South into Perth Railway Station from the bridge carrying Glasgow Road over the station approach from the North. [Google Streetview, 2021]Perth Railway Station in 1901, as it appeared on the 6″ Ordnance Survey. [140]Perth Railway Station in the 21st century. [140]
Having travelled to the most southerly point on the Highland Railway’s network and having already noted that a more direct route from Inverness to Perth was to be constructed. We complete this particular article by acknowledging that the rivalry between the Highland Railway and the Great North of Scotland Railway (GNSR) was intense. Various schemes were developed by the GNSR, but none more significant than the GNSR backed abortive scheme of the Strathspey Strathdon & Deeside Junction Railway Co. (SS&DJ) to construct a 30-mile line from Ballater, terminus of the GNSR Deeside branch from Aberdeen, to Nethy Bridge on Speyside, on a GNSR branch from Craigellachie; the GNSR “was then to build its own line from Nethy Bridge directly to Inverness. The terrain was difficult on both sides of Nethy Bridge, especially to the east, where the alignment cut across a northern ridge of the Cairngorms that would have meant a 2000ft summit (compared with Druimuachdar’s 1484ft) or a long tunnel.” [1: p7] It is difficult to determine whether this route was ever a serious proposition, but “one effect of the SS&DJ proposal was to frighten the HR into building the costly Perth-Aviemore cut-off over Slochd so as to cut the Perth-Inverness passenger transit by reducing the distance compared with the route via Forres and Dava.” [1: p7]
However, the end of the 19th century saw the end of the more absurd, wasteful and expensive rivalry between the Highland and the Great North of Scotland Railway.
North and Northwest of Inverness
Drew comments: “True to its purpose of helping in the economic development of the Highlands, the I&AJR (as it was until it became the HR in 1865) started, in the early 1860s, … pushing north from Inverness. Once again local men and women co-operated. By means of working and eventually taking over lengths of line originally built as independent undertakings, the Farther North line of the HR reached Thurso, Britain’s most northerly town and 20 miles from Duncansby Head, in 1874; Penzance had been linked with London some years before, so that the British railway system extended from (near) John o’ Groats to (near) Land’s End in less than 50 years from the opening of the Stockton & Darlington.” [1: p7]
The Farther North Line or the Far North Line
In Drew’s opinion, “The Farther North line is unique in Britain. The distance by rail from Inverness to Wick, its terminus (Thurso being at the end of the branch) is 161 miles, because the line avoids expensive bridges over the sea lochs, serves as many population centres and even isolated estates as possible, and runs inland from Helmsdale roughly in a semicircle through some desolate country to near Wick. The object was partly to serve the population (sometimes by means of lonely roadhead stations) and partly to avoid the cliffs along the coast north of Helmsdale. By bridging the narrower sea lochs and keeping nearer to the coast the distance from Inverness to Wick could have been only a little over 100 miles; but that would have left many places in Ross & Cromarty, Sutherland and Caithness even farther from a railhead than they are. Less than 10 miles of the more direct route along the coast beyond Helmsdale would have been really difficult construction, probably involving some tunnelling under cliffs.” [1: p7]
Drew notes the problems encountered with snow on the Farther North line, particularly the problem of drifting snow on relatively open expanses of moorland and its propensity to collect in narrow railway cuttings. He highlights the resources needed and the logistics involved in clearing lines that were remote from highway links.
He also highlights the sterling efforts of the Highland Railway in maintaining supplies and communication to the Royal Navy in Inverness and at Scapa Flow during the first world war, over a line not built for such traffic flows. “One of the hardest tasks was the working of trains of coal, largely high-quality steam coal for the Fleet from South Wales, to Invergordon and Thurso; well over 200 miles of the journey was over the HR and most of that distance was single track. The HR had of course to borrow engines and wagons from other railways. The difficulties included lack of the running loops needed for such heavy two-way traffic.” [1: p9]
A daily scheduled service ran from Euston to Thurso during both world wars. Officially ‘The Euston to Thurso Naval Special’, this was known colloquially as the Misery Express or more positively, ‘The Jellicoe Express’. “It was named after Admiral Sir John Jellicoe and carried service personnel to and from Naval bases around the country, including Scapa Flow. WW100 Scotland reports that in WW1 it ran 15th February 1917 – 30th April 1919.” [9]
It was estimated to have transported some half a million service personnel during the two wars. (Helmsdale says half a million in each war). It was the longest scheduled rail service ever to run in the UK.
Crewe Station was one of the few scheduled stops on the 717-mile 21h 30m journey. This was a major refreshment stop, where over 300 women volunteers worked around the clock to provide refreshments in a canteen on Platform 6. Helmsdale was another refreshment stop.
From Royal Naval Association: “Although seen as a crucial piece of Naval infrastructure, the train also carried thousands of soldiers and airmen over the years.Travelling on it was usually a nightmare – invariably overcrowded, most passengers were unlikely to get a seat, and only the shortest of them could hope to get any sleep (if they could find an empty wire luggage-rack). In the Great War it left London at 6pm, arriving at Thurso at 3.30pm the following day. The southbound service took an hour longer …”
The Far North/Father North Line will be covered in detail in a future article. The early involvement of the 3rd Duke of Sutherland in the development of the line is covered here. [51]
The Skye Line from Dingwall to Kyle of Lochalsh “was also built to develop rural areas. Whereas the purpose of the Farther North was to help the mainland northerly counties (though the promoters had Orkney traffic via Scrabster in mind), the Skye line was regarded also as a means of helping economically depressed Skye and Lewis and Harris and such others of the Hebrides as sent fish and the products of sheep farming to the south.” [1: p9]
Drew tells us that “the line began as an independent undertaking, the Dingwall & Skye, and was to run for 63 miles across Ross & Cromarty and over the watershed between the Moray Firth and the Atlantic to Kyle of Lochalsh, terminus of the shortest ferry crossing to Skye and a site for berths for seagoing steamers to Lewis and other islands. Construction was delayed by opposition at Strathpeffer, which little spa had to be bypassed by a costly deviation. Consequently the line only reached Strome Ferry, at the head of the sea inlet, Loch Carron, in 1870. This remained the terminus and point of embarkation until 1897, when the Skye line eventually reached the Kyle, which became an admirably designed minor packet port, complete with a hotel.” [1: p9]
Drew explains that the HR absorbed the Dingwall & Skye (D&S) in 1880. “For 10 years from the opening of the railway to Strome Ferry the D&S and later the HR operated the shipping services between Strome Ferry and Skye and the Hebrides. They were eventually taken over by David MacBrayne and [then] … by Caledonian MacBrayne Ltd. … The HR later had high hopes of Strathpeffer; it built a special two-mile branch (now closed) thither from Fodderty Junction and during the season a Strathpeffer Spa express was run non-stop from Aviemore to Dingwall and on to the branch.” [1: p9]
The line to Skye and the Strathpeffer Branch are covered in detail elsewhere. The Strathpeffer Branch is dealt with here. [3] An article about the line to Kyle of Lochalsh will follow in due course.
Locomotives and Rolling Stock
To conclude this article we take a look at the notes Drew provides on the locomotives employed by the HR and further, at its locomotive superintendents: “The Jones Goods, placed in service from 1894, is one of the best-known types of engine of any British railway, famous as the first instance in the UK of the 4-6-0 wheel arrangement. … Experience with the Jones Goods helped their designer David Jones in developing the Castle Class express passenger 4-6-0s, which did not, however, enter service until 1906, after Jones had retired, and details were finalised by his successor Peter Drummond. … Struck by the magnificent work of the Castles on the HR main lines, the French State Railways in 1910 ordered 50 from the North British Locomotive Company. As illustrations show, Jones’s designs, despite the louvred chimneys, were handsome, and most were highly efficient. In view of the many steep gradients it might be asked why Jones did not develop six-coupled engines before. Indeed the HR and its constituents seemed to prefer Alexander Allan single drivers (successful on lines in flat country) long after greater adhesion became necessary.” [1: p11]
“Having developed successful goods and passenger 4-6-0s, the HR went ahead with others. The best was F. G. Smith’s River Class of 1915, which were not allowed to work on the HR after delivery, because of Smith’s failure to consult the civil engineer, and had to be sold to the CR. His successor Cumming was responsible for two fine designs of 4-6-0, the Cumming Goods and the Clans, but there was little opportunity for those engines to show their paces before grouping.” [1: p11]
William Barclay was HR locomotive superintendent from 1855 until 1965. “During Barclay’s incumbency, … various 2-2-2 and 2-4-0 locomotives were built, along with a solitary 0-4-0T. An 0-4-0ST was also inherited from the Findhorn Railway. Many of Barclay’s locomotives would later be rebuilt by Stroudley or Jones – most of the 2-2-2s ended up as 2-4-0s and one became a 2-2-2T, a pair of 2-4-0s became 4-4-0s and the 0-4-0T became an 0-4-2T. Only 4 much rebuilt Barclay locomotives (all 2-4-0s) were still in stock at the time of the Grouping.” [35]
William Stroudley was HR locomotive superintendent from 1865 to 1870. “His main achievement was high standards of maintenance and cleanliness that continued to be achieved long after his departure to the LB&SCR. He did much to rationalise the HR Lochgorm Works at Inverness.” [1: p11]
Stroudley was also responsible for the design of a small HR 0-6-0ST which, with modifications, was to become the well-known Terrier Class A1 of the LB&SCR. The HR 0-6-0STs were the first locomotives to be built at the Highland Railway’s Lochgorm works in Inverness. The design was the only new design of locomotive brought in during William Stroudley’s spell as the company’s locomotive superintendent. Only three of engines were built. [4] All three, built in 1869, 1872 and 1874, “started life with second hand boilers, shortened to suit, with the dome over the firebox. The side tanks had extra capacity in the form of an arched tank placed over the boiler. The cab was typical of Stroudley but there was no coal bunker; it was carried on the right hand side in the space alongside the firebox between the cab and the side tank, both sides having extended lower cab sides meeting the tanks.” [39]
“To the end the three engines showed strong Stroudley features, although they were not entirely identical, the oldest acquiring a larger side tank during rebuilding in 1917 and the rear toolbox was replaced by a small bunker on all three. Typically for HR engines all three were named: No. 56 – Balnain when new, then Dornoch in 1902; No. 57 – Lochgorm; No. 16 – St. Martin’s when new, then Fort George in 1899. … The later names reflect the move of the engines to those branches. Latterly in HR days they were nameless. The LMS correctly placed Nos 56 and 57, by then with duplicate list numbers as 56B and 57A, with similar powered tanks as Nos 16118 and 16119, by which time the former was working the Strathpeffer branch, but the third was numbered 16383 at the end of the other 0-6-0T classes this may have been due to mistaken identity. All three were soon running with LMS numbers, Inverness having applied LMS numbers and livery rather quickly, but they were condemned in 1927-32.” [39]
Until 1869, all locomotives were painted dark green. After that, passenger locomotives were painted yellow ochre , and freight locomotives were painted Scottish green (very dark green). From 1885, all locomotives were painted pea green with red and white trim. From 1903, the green became darker again, and the tenders now bore the inscription “Highland Railway.” [5]
David Jones was Locomotive Superintendent from 1870 to 1896. He “designed several classes of 4-4-0, and was also notable for introducing the 4-6-0 wheel arrangement to the UK. He also produced small numbers of 0-4-4ST, 2-4-0, 2-4-0T and 4-4-0T locomotives. Of 88 engines built to Jones’ design (including 3 built as late as 1917), 74 passed to the LMS in 1923. A small 2-4-0T purchased secondhand from the Duke of Sutherland also made into LMS ownership.” [35]
“Both the Highland Railway and the Glasgow & South Western Railway initially put [4-4-0 locomotives] on the line during 1873, the first in June, the latter in July – however the important distinction was that the former was produced by fitting a bogie to an outside cylinder 2-4-0 locomotive dating from 1858. The need arose from problems encountered on the recently opened (1870) line from Dingwall to Strome Ferry, the long fixed wheelbase of the six-wheelers sent there not taking kindly to the sharp curves. The first convert was No 10, followed by No 7, another 1858 engine, in 1875. The immediate result for the HR was the introduction of a new 4-4-0 type in 1874, the ’60’ class, which however was for main line work, ten being supplied by Dübs, and at the same time the most powerful passenger engines in Britain. For some years the Lechatelier counter pressure brake was fitted, later replaced by vacuum. With these engines came the well known louvred chimney, a feature of David Jones’ Highland engines. Seven more were built by the company at their Lochgorm Works in 1876-88, generally similar apart from boiler dimensions. Eight others, Nos 76-83, dated 1886, were the first products of the Clyde Locomotive Works of Glasgow (soon absorbed by Sharp, Stewart); the first engine of this batch was displayed at the 1886 Edinburgh Exhibition. The type was perpetuated in 1892 when Nos 89-100 were completed, essentially the same class, apart from further boiler enlargement. As main line engines all of these 4-4-0s were named, but confusion arises from numerous changes of names. Several lasted into LMS days, but only one of the Clyde built and six of the 1892 engines were allocated LMS numbers.” [10: p5]
“A smaller wheeled version of these classes for the Strome Ferry line, hence its popular title ‘Skye Bogie’, was only slowly multiplied, one engine in 1882, followed by eight in 1892-1901, the last four appearing after Peter Drummond had succeeded Jones in 1896. With the completion of No 48 in 1901 the use of the “Crewe” pattern of front end frames and cylinders on new building ceased. Unlike the ’60’ class these engines were not named. All passed into LMS stock but one was not given a new number. However one of them was a combination, identified as a ‘Skye Bogie’ because of its smaller coupled wheels, the frames originating from No 70, the 1882 engine, other components being derived from No 67 of the ’60’ class, the assemblage being completed in February 1923 and identified as No 67 Cromartie. Useful in their later days on branch lines it was 1930 before the last was condemned.” [10: p5]
“Despite introducing a large six-coupled goods in 1894 Jones retained the 4-4-0 wheel arrangement for the next express locomotives. Again the new class, ‘119’ or ‘Lochs’, were some of the most powerful passenger engines in the land. Most of the established features of Inverness engines were retained, the louvred chimney, the Stroudley cab but the frames and cylinders were conventional, not of the “Crewe” type. An early form of piston valves were used, not successfully for they were soon replaced by balanced slide valves. All were named after ‘Lochs’ in the Highland area. There were few changes in HR days although No. 130 ran with a feed water heater. Even after displacement as the premier main line engines by Drummond’s 4-6-0 ‘Castle’ class in 1900, the ‘Lochs’ remained almost confined to the HR main line right up to the Grouping. When further engines were urgently required during World War I the need was partly met by obtaining in 1917 three more of the class from North British in Glasgow, time and expenses being saved by using existing patterns and drawings.” [10: p5]
The ‘Lochs’ were Jones’ final design for the Highland Railway. 15 were originally built by Dubs & Co. in July-September 1896. [10: p16] With the 3 built in 1917, this meant that a total of 18 members of this class were built.
“Drummond brought about a complete transformation of Highland engine power, such that it seemed that his brother Dugald, and Eastleigh Works of the London & South Western Railway had taken over, for many of Peter Drummond’s classes corresponded to L&SWR designs. The first of two Drummond 4-4-0 classes appeared in 1898 and were slightly smaller than the ‘Loch’ class, being intended for services from Inverness to Aberdeen and Wick. The eight initial engines were supplied by Dübs in 1898, followed by a further nine built at Inverness in 1899-1901. Three more came from Glasgow in 1906, having tenders with a little more water capacity. Six-wheeled tenders had been supplied with the class but at times a few engines ran with eight-wheeled tenders from other classes. There was an important change from the previous Inverness adherence to Allan’s straight link valve gear to Stephenson link motion. Another change was the introduction of steam reversing gear, although the initial eight had lever reverse until modified. Nos 2 and 13-7 were dual braked and Nos 2 and 7 later ran with Smith’s feed water heaters. Named after Scottish mountains these engines were inevitably known as the ‘Ben’ class. The first engine arrived at Inverness with the name Ben Nevis to the embarrassment of the HR for that mountain is far away in rival North British Railway territory so a hurried change was made to Ben-y-Gloe.” [10: p5]
Highland Railway ‘Ben’ Class Locomotive No. 1, ‘Ben-y-Gloe’ was originally named ‘Ben Nevis’ but hastily changed to be named after a mountain in Highland Railway territory. There were actually two separate ‘Ben’ classes, usually referred to as the ‘Small Bens’ and the ‘Large Bens’: Highland Railway C and U classes. The designer was Peter Drummond, the builders were: Dübs & Co. (8 Small); HR Lochgorm Works (9 Small); and North British Locomotive Co. (3 Small, 6 Large). Serial numbers were: Dübs: 3686–3692 (Small); NBL: 17398–17400 (Small); 18269–18272 (Large); 18803–18804 (Large). Build dates were: 1898–1906 (Small); and 1908–1909 (Large). In total 20 small and 6 large ‘Bens’ were produced. [33]
Rowledge continues: “A further six 4-4-0s engines, intended for Inverness and Wick trains, were provided by North British in 1908-9, using the frames, etc., of the ‘Ben’ class but carrying a larger boiler, becoming the ‘New Ben’ class, using more mountain names. All had steam reversers and three, Nos 60, 61 and 63, were dual braked. Smith’s feed water heaters were fitted in 1914-5, feed water initially being heated by exhaust steam in a heat exchanger placed alongside the smokebox on the right hand side and then further heated by another set of tubes inside the smokebox. The four 1908 engines started with the same size 3,185 gallon tenders as the last three ‘Bens’, but two, Nos 61 and 63 soon received 3,200 gallon double bogie tenders taken from 0-6-0 tender engines, while the final pair had even larger 3,600 gallon bogie tenders.” [10: p5]
“Despite wartime conditions a new design appeared in 1916 after Cumming had replaced Drummond as Locomotive Superintendent, larger and heavier than any previous HR 4-4-0 class. Nos 73 and 74 were the first two-cylinder engines of this wheel arrangement in Britain to have outside cylinders and valve gear. Apart from the rejected ‘River’ class 4-6-0 engines this pair were the first on the HR to have superheaters, Drummond not having taken to this development while at Inverness. Both were built for the ‘Far North’ line but they also saw use between Inverness and Perth.” [10: p5]
From the late 1920s ‘Snaigow‘ and ‘Durn‘ were found mostly working from Aviemore Shed on goods or piloting work, having been replaced on the Far North line by Drummond ‘Castle’ 4-6-0s which were in turn displaced by new LMS ‘Crabs’ and Caledonian ‘Rivers’ sent to the Highland Railway. The arrival of Stanier ‘Black Fives’ sealed their fate and both were withdrawn in the mid-1930s. [10: p25]
“The LMS inherited 78 4-4-0s from the HR – this is the official figure contained in a document at the Public Record Office, Kew – but two others were still extant and are usually included in the initial LMS total. Only 61 were allotted new LMS numbers, the rest being written off during 1923. Despite early withdrawal of pre-1896 examples several of them, Nos 14271/2/5-9/82-5, appeared in the LMS red livery with large numerals on the tender.” [10: p5-7]
“The Highland [4-4-0] engines lasted rather better than those of the G&SWR, twelve passing into British Railways stock.” [10: p7] The last of these was ‘Small Ben’ No. 54398 ‘Ben Alder‘ which was scrapped in the second half of 1952.
Having followed the various 4-4-0 classes employed by the Highland Railway we return to look at David Jones’ other locomotive designs. As we noted above, he “was also notable for introducing the 4-6-0 wheel arrangement to the UK. He also produced small numbers of 0-4-4ST, 2-4-0, 2-4-0T and 4-4-0T locomotives. Of 88 engines built to Jones’ design (including 3 built as late as 1917), 74 passed to the LMS in 1923. A small 2-4-0T purchased secondhand from the Duke of Sutherland also made into LMS ownership.” [35]
David Jones Goods class “was notable as the first class with a 4-6-0 wheel arrangement in the British Isles. Fifteen were built, and one has survived to preservation.” [36]
Jones Goods locomotive, HR No. 103 survived into preservation. It is seen here at the Riverside Museum (replacing the preceding Glasgow Museum of Transport), Partick, Glasgow. [36][37]
Fifteen ‘Jones Goods’ locomotives “were built by Sharp, Stewart and Company and delivered between September and November 1894, numbered 103 to 117. At the time, these were the most powerful main line engines in the country. Originally intended principally as freight engines, they were often called upon for passenger duties during the wide fluctuations of traffic which occurred on the Highland Railway, particularly during the summer season.” [36]
“Although the type was a notable success for Jones, an accident while testing one of the locomotives caused one of his legs to be severely scalded. Although he recovered, he was permanently affected and by the end of December 1896 had retired due to ill-health.” [36]
Peter Drummond was Locomotive Superintendent of the HR from 1896 to 1912. Under Peter Drummond, new 0-4-4T, 0-6-0T, 0-6-4T, 0-6-0, 4-4-0 and 4-6-0 designs emerged. All 72 of his locomotives passed to the LMS. [35] Among those 72 locomotives were the class ’18’ 0-6-0s – these were the only class of 0-6-0 employed by the HR. The class was introduced on the network in 1900. [38: p9]
“Originally intended to be 15 in number the initial order was for only six when placed in 1899, Dübs supplying Nos 134-9. They were followed by Nos 18-21 in 1902, also from Dübs, and finally Nos 36 and 55 (the latter renumbered 37 in 1921) in 1907 from North British. The class replaced an assortment of small and old 2-4-0 locomotives. The 1902 engines had Drummond water tube fireboxes, but otherwise the “Barneys” (just how the class became so named is lost in the mists of the past) were quite conventional. The first six had double bogie tenders, the rest six wheeled, the former being exchanged with six wheeled tenders from the 4-4-0 ‘Ben’ class (LI No 117) in 1902-13.” [38: p9]
“Most of the ’18’ class, as far as is known, were used on goods trains to east and north of Inverness rather than the main line to Perth. Despite not having steam heating pipes there was some use on passenger trains, especially excursions. In LMS ownership the class was numbered 17693-17704. In 1938 Nos – 17693-5/7, 17703 were moved to the Glasgow area to assist with extra traffic to the Glasgow Exhibition, when more engines with vacuum brakes were needed, working from Dawsholm, – Motherwell and Hamilton. One No 17694, was even seen on an excursion train at Edinburgh which also crossed the Forth Bridge to Burntisland. Although returned to Inverness later that year five were again used in Glasgow during 1939, but not for long. A longer stay in that area began at the end of 1946 when the eight survivors moved to Corkerhill, some later being transferred to Ayrshire sheds. Seven passed into British Railways stock, the last being withdrawn in 1952.” [38: p9]
Generally, the HR had little use for shunting tank engines, but “In 1903/04 the Highland Railway built three powerful 0-6-0 tank engines at its Lochgorm works in Inverness. Numbered 22, 23 and 24 they were used for shunting duties. The boilers and wheels were taken from older locomotives that had been scrapped and this led to the engines being referred to as ‘Scrap Tanks‘.” [40] These outside cylinder engines were, “quite unremarkable except that the boilers were recovered from obsolete 2-4-0 locomotives together with re-usable wheels and other parts. The result was a class with rather larger wheels than usual for shunters. Normally there were two working at Inverness with the other at Perth. Once the boilers were no longer listed as a standard type their survival was limited and they were condemned in 1930-2.” [39: p9]
Frederick George Smith was Locomotive Superintendent from 1912 to 1915, his “brief tenure was cut short by a dispute over his sole design, the ‘River’ Class 4-6-0. Six locomotives were built, but they were (wrongly) considered to be too heavy for the Highland Railway, and were sold to the Caledonian Railway without being used.” [35]
Christopher Cumming was Locomotive Superintendent from 1915 to 1922. He designed one class of 4-4-0 and two types of 4-6-0, totalling 18 locomotives, which all passed to the LMS. [35]
The Future
One feature of Drew’s article, written at the end of 1975, was the palpable concern expressed over the security of rail services in the remote areas of Scotland and particularly from the Farther North Line and the Kyle of Lochalsh line. With hindsight, we can say that (although justified in the mid-1970s, with the bias of the time towards road transport) what was feared has not come to pass. [1: p8] Passenger rail services on both these lines (and on the West Highland line) continue, although we should note that Scotland’s Railway receives some of the highest public subsidy anywhere in the UK: two thirds of the costs come from the Scottish Government; and on average, tickets are 20% lower than the rest of the UK. [2]
References
Brian Stephenson (ed.); The Highland Railway; Trains Illustrated No. 18, Ian Allan, Shepperton, London, 1976.
J. W. P. Rowledge; The 4-4-0 Locomotives of the Glasgow & South Western and Highland Railways; in Brian Stephenson (ed.) Locomotives Illustrated No 117, January-February 1998, Ian Allan, Berkhamsted, Herts, 1997.
J. W. P. Rowledge; Scottish 0-6-0 and 2-6-0 Classes of the LMS; in Brian Stephenson (ed.), Locomotives Illustrated No 133, September-October 2000, Ian Allan, Berkhamsted, Herts, 2000.
J. W. P. Rowledge; LMS Scottish Goods Tank Locomotives; in Brian Stephenson (ed.), Locomotives Illustrated No 154, July – September 2004, Ian Allan, Berkhamsted, Herts, 2004.
Mike Bridge, ed.; TRACKatlas of Mainland Britain: A Comprehensive Geographic Atlas Showing the Rail Network of Great Britain (3rd ed.); Platform 5 Publishing Ltd., Sheffield, 2017, p94.
John Thomas & David Turnock; A Regional History of the Railways of Great Britain. Volume 15 North of Scotland; David St John Thomas, Newton Abbott, 1989, p236.
Highland doubling; in The Railway Magazine Issue No. 902, June 1976, p277.
Highland doubling starts; in The Railway Magazine Issue No. 905, September 1976, p476.
In Part 1, we looked at the railways in the North of Namibia, that article can be found here. [4] This article covers lines which left Windhoek and covered the South of the country.
After the aerial image immediately below, the next three images form a kind of ‘tryptic’ which shows the TransNamib train yard and station at Windhoek. Taken together they show the full site. …
The railway line from Windhoek to Nakop is 869 kilometres (540 miles) long. The section between Karasburg and Keetmanshoop was completed in 1909. In 1912, the 500 kilometres (310 miles) connection between Karasburg and Windhoek was completed, and the extension to Upinhton (South Africa) was built in 1915.
Windhoek
Aris
Rehoboth
Tses
Keetmanshoop
Karasburg
Nakop (border)
Upington
We start our look at the main line to the South, in Windhoek at the South end of the Railway Station site. …
The southern end of Windhoek Railway Station, seen from John Meinert Street. [Google Streetview, 2023]The line South, from John Meinert Street. [Google Streetview, 2023]A satellite image showing the railway South of John Meinert Street. [Google Maps, June 2025]The next length of the line to the South, to a point South of the B6. [Google Maps, June 2025]The bridge carrying the railway across the B6. [Google Streetview, 2024]
Although it does not look like it as yet, the line South of Windhoek Railway Station climbed relatively steeply as it meandered South. … As we will soon see, the landscape South of Windhoek is different to that to the North of the city.
The line to the South of the B6 curves round the residential area of Schmerenbeck Street. [Google Maps, June 2025]The line then crosses the Gammams River and under David Hosea Meroro Road. [Google Maps, June 2025]The triangular junction visible in this image gives access to the branch line to Gobabis. [Google Maps, June 2025]The line then passes under the B1. [Google Maps, June 2025]Looking South from the B1. [Google Streetview, 2024]Then the line(s) cross the Arebbusch River. [Google Maps, June 2025]The next road to bridge the line is the C26. [Google Maps, June 2025]The view from Mandume Ndemufayo Avenue (C26) back towards Windhoek Railway Station. [Google Streetview, 2024]The view South from Mandume Ndemufayo Avenue towards the hills. [Google Streetview, 2024]The line passes under the city’s Western Bypass. [Google Maps, June 2025]This view looks North-northeast under the Western Bypass towards Windhoek Railway Station. [Google Streetview, 2024]Looking South from the Western Bypass. [Google Streetview, 2024]The line passes once again under the B1 as the hills draw closer. [Google Maps, June 2025]The view South-southeast along the line from the bridge carrying the B1. [Google Streetview, 2022]
The line continues to wind its way into the hills passing under the B1 once again.
The line has deviated away from Birmingham as it finds its own way into the hills. It returns to pass under the modern road again. [Google Maps, June 2025]The view South from the bridge carrying the B1. The line enters a very short tunnel just to the South of the modern road. [Google Streetview, 2022]The B1 continues to climb as it heads South. The railway takes a different path as it gains height. It crosses over the B1 by means of this bridge. [Google Streetview, 2022]
Some kilometres further along the line it again crosses the B1. This time the road bridges the line. …
Both road and railway continue their journey South. Here their paths cross once again close to Aris Railway Station. [Google Maps, June 2025]Looking back along the line to the North from the bridge carrying the ,B1 over the line. [Google Streetview, 2022]And from the same bridge looking Southwest into Aris Railway Station. [Google Streetview, 2022]
The community of Aris and its railway station are shown in a YouTube video. …
Aria Railway Station. [9]
South of Aris Railway Station,two tracks run in parallel as far as Aris Quarry.
Part of Aris Quarry appears at the bottom left of this satellite image. The other significant part of the Quarry sits to the Southeast on the East side of the B1. [Google Maps, June 2025]At Aris Quarry Google Maps shows three bogie hopper wagons at a short wharf. [Google Maps, June 2025]T
The line continues South towards Rehoboth. …
Quarry land sits on the East side of the line at the top of this next satellite image. A dry water ourselves can be seen across the image. It is bridged by the line. [Google Streetview, June 2025]A closer image of the truss girder bridge crossing the dry river bed. [Google Maps, June 2025]
The next two satellite images show the line heading further South. …
For a short distance the B2 runs close to the railway. The railway then crosses two more dry watercourses. [Google Maps, June 2025]The view from the B1 at the top of the satellite image immediately above looking West. The railway line can be seen between the road and the mountain. [Google Streetview, January 2024]The bridge over the first dry river. [Google Maps, June 2025]The bridge over the second of the two watercourses. [Google Maps, June 2025]
The next three satellite images take the line further South, running on its own course with the B1 away to the East. …
Close to the bottom of this image the railway crosses the Oanob River’s watercourse on a causeway with a short trestle bridgeThe truss girder bridge over the dry watercourse of the Oanab River. [Google Maps, June 2025]
More satellite images take us further South. The first two of these extracts from Google’s satellite imagery shows the line running past the Omeya Golf and Residential Oasis. More about this relatively recent development can be found here [6] and here. [7]. The second of these is a sales video produced to attract investors and house sales.
Four satellite images take the line South to a point where it once again runs alongside the B!. [Google Maps, June 2025]Looking Southwest from the B1 just to the North of its junction with the D1427. The railway can be seen in the background behind the closest trees. [Google Streetview, 2022]
Over the next 3 or 4 kilometres the line and the road run in parallel, with little worthy of note, Before the railway moves away to the West of the road once again. …
The line crosses four dry watercourses before turning towards the East. [Google Maps, June 2025]Running Southeast the line follows the fourth of the watercourses and passes under the B1. [Google Maps, June 2025]Looking North-northwest along the line from the bridge carrying the B1. [Google Streetview, November 2023]Looking Southeast from the same bridge towards Rehoboth. [Google Streetview, November 2023]
The line wanders its way through the hills to the East of the B1 passing from the Khomas Region of Namibia into the Hardap Region. It runs through a number of small townships close to the Usip River before entering Vogelpan where Rehoboth Railway Station was sited.
Just a short distance to the South of Vogelpan, the railway turns East to cross the Usip River. ….
Looking East from the C25, the line can be seen curving away to the East and crossing the channel of the Usio River. [Google Streetview, August 2024]
Just to the South of the location of the photograph immediately above, the C25 itself turns East and crosses both the Usip River watercourse and then the railway.
Looking North-northwest from the ungated crossing on the C25, along the railway back towards Vogelpan. [Google Streetview, August 2024]Looking Southwest from the same crossing on the C25, along the railway. [Google Streetview, August 2024]
The railway runs down the East side of the River Usip passed its confluence with the Oanob River. It then continues alongside the Oanob (less than 1 kilometre to the East of the river).
The railway then crossed the Oanob River and a tributary in quick succession. [Google Maps, June 2025]
The first bridge (on the left) crosses the Oanob, the second (on the right, crosses the tributary). [Google Maps, June 2025]
The line then continues, pretty much in a South-southeast direction for some considerable distance with little to remark on. It passes close to Duinevelde on its way South before reaching Kalkrand.
The railway is seen here (from a minor road to the West of the line) approaching Kalkrand from the North. [Google Streetview, August 2024]Kalkrand and its railway Station. [Google Maps, June 2025]Kalkrand Station and passing loop. [Google Maps, June 2025]
Southeast of Kalkrand the line and the B1 run in parallel. …
Road and rail together head Southeast. [Google Maps, June 2025]Looking East from the B1, the railway can be seen on a parallel course. [Google Streetview, December 2023]The dunes of the Kalahari desert appear in the top right of this satellite image. [Google Maps, June 2025]The railway flirts with the edge of the dunes. [Google Maps, June 2025]
The line continues in a generally Southeasterly direction along the edge of the Kalahari Desert before beginning to swing round to the South and heading into rougher terrain where its route is dictated by the contours. …
Sidings some distance to the South of Kalkrand. [Google Maps, June 2025]These two structures carry the line over watercourses in the hills [Google Maps, June 2025]
Further South and again on flatter terrain the line passes under the C20. …
Looking North-northwest along the line from the bridge carrying the C20. [Google Streetview January 2018]Looking Southeast along the line from the same bridge. The various storage tanks on the horizon are associated with Agrimark, an agricultural and retail store, part of the Agrimark, Namibia network. Adjacent to it, between the Fish River and the B1 and on the West side of the Fish River is an area of irrigated fields. [Google Streetview, January 2028]Agrimark’s storage facility heralds the arrival of Southbound strains in the town of Mariental. [Google Maps, June 2025]Looking North from an ungated crossing on a minor road North of Mariental. [Google Streetview, December 2023]Looking South from the same crossing. [Google Streetview, December 2023]Further South, the line bridges a dry watercourse which is a tributary of the Fish River. [Google Maps, June 2025]The bridge in the above satellite image as seen from the B1. [Google Streetview, December 2023]Immediately North of Mariental, the crosses another minor road at an ungated crossing. This view looks North along the line. [Google Streetview, December 2023]Looking South towards Mariental from the same crossing. [Google Streetview, December 2023]Mariental Township with the Railway Station at the centre of the satellite image. [Google Maps, June 2025]
The next two extracts from Google Maps satellite imagery cover the length of the railway station site. …
These two images show the station site at Mariental. [Google Maps, June 2025]Mariental Station building. [Google Streetview, December 2023]The station building at Mariental seen fromt he Southwest. [Google Streetview, December 2023]Immediately South of Mariental, the B1 and the railway run South side by side. [Google Maps, June 2025]The railway seen from the B1 to the South of Mariental. [Google Streetview, January 2018]A series of different culverts and bridges support the line over historic channels most of which are dry. [Google Streetview, December 2023]The next road crossing is that for the C18. It is another ungated crossing and this is the view looking North along the line. [Google Streetview, August 2024]And this is the view South along the line at the same crossing. [Google Streetview, August 2024]Looking Northwest from the ungated crossing over the D1068 at Asab. [Google Streetview, December 2023]Looking Southeast from the D1068 into the station at Asab. [Google Streetview, December 2023]Asab Trading CentreSidings/Station and River. [Google Maps, June 2025]Looking North from the B1 along the Asab River watercourse with the railway bridge close to the centre of the image. [Google Streetview, December 2023]
The line continues Southeast from Asab. …
Looking North from the D3919Looking Southeast from the D3919The line continues Southeast towards Tses on a shallow embankment with bridge openings for run-off water over dry watercourses. [Google Streetview, December 2023]Tses Township, the B1 and the railway. [Google Maps, June 2025]Tses Station and Sidings. [Google Earth, June 2025]Tses River Bridge. [Google Earth, June 2025]Looking Northeast from the bridge carrying the B1 along the line, back towards Tses. {Google Streetview, December 2023]Looking ahead to the Southwest from the same bridge. [Google Streetview, December 2023]
The line continues in a generally Southwards direction. Again, when hills are encountered it curves its way along the contours to limit gradients. …
Another glimpse of the line from the bridge carrying the B1 across the outfall channel from the Van Rym Dam on the approaches to Keetmanshoop. [Google Streetview, December 2023]
Keetmanshoop is the next significant settlement on the line. …
4-8-0 Locomotive Class 7A, No. 1011 was built by Neilson & Co, of Glasgow, Scotland, as works no 4930 in 1896, it was brought to Keepmanshoop in 1980 to be plinthed. The first Class 7 locomotives were commissioned by the Cape Government Railways and delivered by Dübs & Co of England in 1892. Follow-up batches were built by Sharp, Stewart & Co, Neilson, Reid & Co, and North British Locomotive Company. They had a wheel arrangement of 4-8-0, coupled wheels of 3’6 3/4″ (1086mm diameter) and Stephenson link valve gear. The engines were originally powered by saturated steam, but many were later reboilered and converted to use superheated steam. Modifications such as larger boilers, increases in cylinder diameters and larger cabs resulted in the reclassification of the locomotives as Class 7A, 7B, 7C, 7D, 7E and 7F. More information and photographs can be seen here. [16]
More modern facilities at Keetmanshoop Railway Station. [Google Maps, June 2025]A view of these modern facilties from the Southwest on Darn Viljoen. [Google Streetview, December 2023]The yard and turning triangle at Keetmanshoop Station. [Google Maps, June 2025]The view of the yard from the corner of 3rd Street and 12th Avenue. [Google Streetview, December 2023]
While Keetmanshoop railway Station sits approximately on an East-West alignment is is approached bey means of a sharp curve from the North and trains leaving to the South take a sharp curve to the South within the township.
The line continues to the South. … First five images following the line South and Southwest as far the point where the B4 bridges the line.
A series of five images following the line, first along Railway Street and then at an ungated crossing, all photographs are taken looking Southwest. [Google Streetview, December 2023]The length of line covered by the images above. The B4 crosses the line at the bottom of the image. [Google Maps, June 2025]
Two pictures now taken from the bridge carrying the B4 over the line.
Looking North from the B$ towards Keetmanshoop. [Google Streetview, December 2023]Looking South along the line ahead. [Google Streetview, December 2023]Looking Southwest along the line from an ungated crossing to the South side of the B4. [Google Streetview, December 2023]Looking Southwest along the line from another ungated crossing to the South side of the B4. [Google Streetview, December 2023]The B4 and the railway run parallel, perhaps around 100 metres apart. This photograph shows the line running parallel to the road. Google Streetview, December 2023]Another view looking Southwest, this time from the ungated crossing which takes the C12 dirt road over the line. [Google Streetview, January 2018]The last image showed a passing loop to the Southwest of the C12. This satellite image shows the full length of the loop. [Google Maps, June 2025]
The railway turns to the West as it approaches the junction at Seeheim.
The railway continues to run parallel to the C12/M28 for some considerable distance. Close to Grunau the C12/M28 meets the B1 which crosses the line at 90°.
Looking East-southeast from the bridge carrying the B1 over the railway, the passing loop at Grunau is just ahead. [Google Streetview, 2024]The railway station/passing loop at Grunau to the East of the B1. [Google Maps, June 2025]Grunau Railway Station, (c) Pgallert (2010) and licenced for reuse under a Creative Commons Licence (CC BY-SA 3.0). [17]
East of Grunau, the line runs parallel to the B3 in a southeasterly direction. …
The line bridges the dry watercourse of the Hom River and other watercourses as it head Southeast. [Google Maps, June, 2025]The Hom River Railway Bridge, seen from the B3 Bridge over the same watercourse. [Google Streetview, 2022]Two further structures carry the Railway over dry watercourses on the run into Karasburg. [Google Streetview, The railway enters Karasburg from the North West, curving round to close to a North-South alignment through the railway Station and then sharply curving round to the Northeast as it leaves the town. [Google Maps, June 2025]The view South towards Karasburg Railway Station from the ungated crossing at Hendrik Snyman Street. [Google Streetview, January 2024]Karasburg Railway Station. [Google Maps, June 2025]The turning triangle at Karasburg. [Google Maps, June 2025]
Karasburg Railway Station: passenger platform/building (note the painted edge of the platform) and goods shed. The pictures below come from 1914/1915 and show different aspects of Karasburg Railway Station at that time.
Karasburg Railway Station in 1914/1915. [18]The engine shed in 1914/1915. [18]Pointwork at Karasburg Railway Station. [18]
As we have already noted the line South of Karasburg Station curves sharply to the Northeast and passing under the M21.
Looking West from the bridge carrying the M21 over the railway at Karasburg. [Google Streetview, September 2024]Looking East from the bridge carrying the M21 over the railway at Karasburg. The curve shown on the last image continues as the railway turns to the Northeast. [Google Streetview, September 2024]East of Karasburg the line runs Northeast alongside the B3 before head East on the South side of the road. [Google Maps, June 2025]
For some distance the line runs along the South side of the B3. separating from it close to Nuwefontein. The B3 crosses to the East side of the Ham River, with the railway remaining on the West side of the river, before crossing it close to De Villiersputs. Near Grondorner, the line crosses the D237 and bridges a tributary of the Ham River before running alongside the D237 in a Southeasterly direction. After a number of kilometres, the D237 turns away to the South and the line continues in a generally easterly direction.
Meandering to the north and then again to the East the line of the B3 once again and runs on its South side towards Ariamsvlei, the border with South Africa and Nakop, beyond the border in South Africa.
Ariamsvlei Railway Station and marshalling yard. [Google Maps, June 2025]The view East into the site from the B3. Google Streetview, January 2024]Ariamsvlei turning triangle and border post which is just about 10 kilometers from the border with South Africa. Nakop is beyond the border. [Google Maps, June 2025]
Beyond Ariamsvlei is the border crossing at Nakop and the South African town of Upington.
Windhoek-Gobabis
Next we look at a line which ran East from Winhoek. … The railway line from Windhoek to Gobabis is 228 kilometres (142 miles) long and was completed in 1930. [10]
Windhoek (capital – junction)
Neudamm
Omitara
Gobabis (branch railhead)
The line to Gobabis leaves the Windhoek-Nakop line at a triangular junction and headed East.
The Windhoek-Gobabis line leaves the line to Nakop in the Southern suburbs of Windhoek. [Google Maps, June 2025]
These next fourteen satellite images show the line wandering back and forth through the Windhoek suburbs. …
These fourteen extracts from Google’s satellite imagery take the Gobabis line to a point to the East of Sam Nujoma Drive. [Google Maps, June 2025]Looking South-southwest from Andries de Wet Street along the line of the railway towards Windhoek Railway Station, showing the substantial bridge under construction in 2024. [Google Streetview, 2024]
The journey beyond Andries de Wet Street continues, the next satellite images cover larger areas than the ones above. ….
This sequence of three images takes the line out into open country beyond the Avis River. [Google Maps, June 2025][The truss girder bridge which carries the railway over the B6 and the channel of the Avis River. [Google Streetview, January 2024]
The line continues in a generally easterly direction, although the contours of the terrain mean that the railway has to meander back and forth to find the most advantageous route. These next extracts from Google’s satellite imagery show the way that the line picks its way through the landscape. …
This sequence of four extracts from Google’s satellite imagery takes the line as far as Finkenstein. [Google Maps, June 2025]The railway station close to Finkenstein Estate/Village. [Google Maps, June 2025]The view Southwest from the D1527 along the railway towards Windhoek. Google Streetview, January 2024]Looking Northeast into the railway station site from the D1527. [Google Streetview, January 2025]A little further to the Northwest on the D1527, this is the view looking Northeast into the goods yard at the station. [Google Streetview, January 2024]
These two images indicate the area of the Finkenstein Estate/Village and current proposals for new housing and a new major road.The location of the railway Station can be seen at the top-left of the satellite image. The lighter swathe of ground curving across the top half of the satellite image is the construction site for the new road. [5][Google Maps, June 2025]
The line continues East from Finkenstein. At the right side of this satellite image the railway passes under the B6. [Google Maps, June 2025]Looking Northwest from the B6 along the line towards Finkenstein Station and Windhoek. [Google Streetview, January 2024]Looking Southeast from the B6 the line is curving to the South. The older road bridge is visible alongside the B6. Not too far ahead, out of shot are the construction works for the road that will replace the B6 as a main artery. [Google Streetview, January 2024.
Almost immediately after passing under the B6, the line begins to swing round from a Southwestern trajectory to the Northeast and then the East before passingunder the B6 again.
Looking Northeast from the bridge carrying the B6 over the line. The line can be seen curving round to the East. [Google Streetview, January 2018]
The line heads generally in an easterly direction and after some kilometeres runs alongside the Seeis River. …
The railway bridges the Seeis River and then runs alongside both the river and the B6. [Google Maps, June 2025]
Stocking close to the South bank of the Seeis River, the line passes to the North of Sonnleiton Village before rejoining the B6 close to Windhoek Airport, and passing to the South of the airport and then running immediately alongside the B6.
The railway seen looking North from the B6. [Google Streetview, January 2024]
Both road and rail pass just to the North of the settlement of Seeis where an old railway station was sited, and then over the River Seeis.
The Seeis River Railway Bridge seen looking North from the B6. [Google Streetview, January 2024]
The video below shows the settlement, its station and its bridge. ….
YouTube video of Seeis and its station and railway bridge. [19]
Beyond Seeis the railway continues alongside the B6 in a Northeasterly direction. The railway then turns further Northwest and leaves the B6, finding its own path towards Gobabis. It crossed the D1535 at an ungated crossing. The D1535 then runs alongside the railway heading Northeast before crossing the railway again at another ungated crossing. Both road and railway cross the Wit Nossob River and run along its North bank, crossing tributaries enroute before passing to the North of the Otjivero Reservoir and Dam.
The Otjivero Reservoir and Dam. The railway runs on the North side of the D1535. [Google Maps, June 2025]
Beyond the Dam, the railway follows the C29 heading East. it pulls away a little to the North to create room for Omitara Railway Station.
Omitara Railway Station. [Google Maps, June 2025]Omitara Railway Station, (c) Arche-foto, Burkhart Rüchel, and licenced for reuse under a Creative Commons Licence, (CC BY-SA 3.0). [20]
Northeast of the station the railway passes under the C29, continuing to run East-northeast and then East, and then Southeast, before returning to run alongside the B6 once again. Following the line on satellite images has become increasingly difficult.
This photograph looks along the line to the East from the ungated crossing over the D1658 at Grunental. As can be seen the line is significantly over grown by grasses. [Google Maps, June 2025]Looking East at the ungated crossing over the D1663, the line seems to be disappearing into the sand. {Google Streetview, January 2024]Looking West at the old station site at Witvlei. [Google Streetview, January 2024]Looking East at the old station site at Witvlei. [Google Streetview, January 2024]Witvlei Railway Station in 2018, (c) Hp.Baumeler and licenced for reuse under a Creative Commons licence, (CC BY-SA 4.0). [21]Looking back West along the apparently little used line towards Windhoek from the bridge carrying the B6 across the line. [Google Streetview, January 2024]Looking East from the same bridge, the line is almost indistinguishable from the surrounding grassland. [Google Streetview, January 2024]
The railway continues to the East, with the B6 running parallel to it on the North side.
Looking East along the line at an ungated crossing on a minor road.The B6 can be seen over to the left. [Google Streetview, January 2024]
The line begins to turn to the Southeast before crossing the Black Nossob River. …
Looking Southeast from the C30 along the line towards the Black Nossob River. [Google Streetview, January 2024]The railway bridge over the Black Nossob River close to Gobabis. [Google Maps, June 2025]Gobabis: the Black Nossob River, Reservoir and Dams are on the left side of the satellite image. The railway bridge over the Black Nossob can just be picked out in the extreme top left of the image. The railway flanks the reservoir and lake before passing under the B6 and then, after the B6 has turned through 90, under the B6 (Gobabis Bypass).Looking West from the B6 bridgeLooking East from the B6 BridgeLooking Northwest from the Gobabis Bypass (B6).Looking Southeast from the Gobabis Bypass (B6).Looking Northwest from Heroes Lane/Creamery Lane, GobabisLooking Southeast from Heroes Lane/Creamery Lane, Gobabis
The 6 images immediately above show the railways approach to Gobabis Railway Station.
Gobabis Railway Station. Its turning triangle can be seen at the right side of this satellite image. A series of sidings are the end of the line, these are just Southeast of the turning triangle. [Google Maps, June 2025]Gobabis Railway Station Buildings. [Google Maps, June 2025]Gobabis Railway Station Building, (c) Hp. Baumeler and licenced for reuse under a Creative Commons Licence (CC BY-SA 4.0). [22]Plinthed at Gobabis Railway station is this 5 man railway inspection car (c) Hp. Baumeler and licenced for reuse under a Creative Commons Licence (CC BY-SA 4.0). [23]Also plinthed at Gobabis Railway station is this small flatbed wagon, (c) Hp. Baumeler and licenced for reuse under a Creative Commons Licence (CC BY-SA 4.0). [24]
Gobabis is the end of the line. There are plans on the drawing board for a Trans-Kalahari Railway Line which would extend East from the current livestock railhead at Gobabis and may well be electrified. [25][26][27]
Seeheim-Lüderitz
The final length of line to be looked at is that from Seeheim to Lüderitz. The railway line from Seeheim to Lüderitz is 318 kilometres (198 miles) long. The connection between Lüderitz and Aus was completed in 1906, and the extension to Seeheim was completed in 1908. [1] The service between Aus and Lüderitz was decommissioned in 1997, due to poor track condition, and there is no regular passenger service between Seeheim and Aus. … The line to Lüderitz was rehabilitated in the 2010s and was scheduled for reopening in 2017. Test trains ran to Lüderitz in 2014 and Lüderitz Harbour in 2018. [2]
We begin this journey from Seeheim Railway Junction. …
Seeheim Railway Junction. [Google Maps, June 2025]
Just a short distance Northwest along the line from the junction is Seeheim Railway Station.
Seeheim Railway Station in very early days. [30]The train yard at Seeheim in the 21st century. The Skaap River is just to the South of the Yard. [Google Maps, June 2025]A short distance beyond the limits of the old railway station the line crosses the Fish River close to its confluence with the Skaap River. This is a very early postcard image of the railway bridge over the river. A modern image of the bridge in the distance on this postcard can be seen here. [29] [30]The bridges over the Fish River in the 21st century. [Google Maps, June 2025]After crossing the Fish River the railway follows its North bank. [Google Streetview, June 2025]It crosses a couple of tributaries before heading away from the river into the hills. [Google Streetview, June 2025]
After some distance winding through the hills, the railway line crosses the D463, bridges a dry watercourse and crosses an open area of sand before again winding its way through more hills, heading Northwest. The next image shows the location of Sandverhaar Railway Station and Bridge.
Sandverhaar Railway Station and Bridge. [Google Maps, June 2025]Sandverhaar Railway Station, (c) Matthias Bruhin & Hp.Baumeler and licenced for reuse under a Creative Commons Licence (CC BY-SA 4.0). [28]
A little further West the line crosses the dry watercourse of the Gurib River and close to Alte Kalkofen Lodge passes through the site of what was Simplon Railway Station.
Immediately to the West of Simplon Station, the line crosses the D462 and then, for a short while runs alongside the B4.
At Goageb, the line passes under the B4 before entering the Railway Station.
Looking Northwest from the first bridge at Goageb carrying the B4 over the line. [Google Streetview, December 2023]
Goageb Railway Station. [Google Maps, June 2025]
The Station building and water tower at Goageb, seen from the Southwest. [Google Streetview, December 2023]The platform, station building and water tower at Goageb, seen from the Northwest. [Google Streetview, December 2023]
The bridge carrying the railway over the Konkiep River, seen from the bridge carrying the B4 over the river. [Google Streetview, December 2023]Looking back towards the railway’s bridge over the river. [Google Streetview, January 2018]Looking Northwest from the bridge carrying the B4, along the line of the railway. [Google Streetview, January 2018]
After passing under the B4 the railway turns North-northwest and runs alongside the B4 for a few kilometres before the road turned away toward the North. The railway then turns to the Southwest.
A relatively short passing loop is provided seemingly in the middle of nowhere! [Google Maps, June 2025]
These next few pictures give a sequence of satellite images or views in sequence along the line.
An igneous rock intrusion alongside the lineA dramatic S-curve on the line. Another unnamed passing loopThe line running South of and close to the B4Looking South from the B4 along the D446 we can see an ungated crossing . The railway is around 100 metres from the B4 at this location.A memorial has been placed at the site of a prisoner of war camp from the First World War to the East of the town of Aus.The Camp near Aus for German prisoners of war 1915-1919, (c) Public Domain. [31]Immediately to the South of the Kriegsgefangenen Denkmal, the war memorial, there are a series of sidings/passing loop on the railway as shown here. [Google Maps, June 2025]A gantry crane, cabins and water tower are present at this location. [Google Streetview, January 2024]
It is only a short distance from this location to the township of Aus. The railway passes, first, under the C13 and then into Aus.
The view Northwest from the C13 towards Aus. [Google Streetview, January 2024]Aus Railway Station. [Google Maps, June 2025]Looking Southeast, this photograph shows the railway line curving round into Aus Railway Station. [Google Streetview, December 2023]The Southeast end of Aus Railway Station site. [Google Streetview, December 2023]The station approach at Aus. [Google Streetview, December 2023]
Northwest of the railway station, the line crosses the C13 again, this time at an ungated crossing.
Looking Southeast from the ungated crossing on the C13 towards Aus Railway Station. [Google Streetview, December 2023]Looking Northwest along the line from the ungated crossing on the C13 – Aus church is prominent in this photograph. [Google Streetview, December 2023]
West of Aus the railway passes through the hills. The next image looks backthrough those hills towards Aus.
Looking back towards Aus we can see the line meandering through the hills. [Google Streetview, November 2021]Looking ahead along the railway line from the same location. [Google Streetview, November 2021]
The next station on the line was at Garub. The station is no longer in use.
Garub Railway Station: the station hose and water tower remain standing. [Google Maps, June 2025]Garub Railway Station building in the early 1900s, (c) Public Domain. [32]
This image of Garub station house and water tower is embedded here from fineartamerica.com. [33]
To the West of Garub the B4 and the railway run in parallel. The line passes through another abandoned stations at Tsaukaib, Haalenburg, Rotkop and Grasplatz.
The abandoned Railway Station at Haalenburg, looking West. [Google Streetview, December 2023]The abandoned Railway Station at Grasplatz, looking West. [Google Streetview, December 2023]
At Kolmanskop there is an entire derelict mining station which once had its own railway station.
Kolmanskop seen from the B4 with the railway visible in front of the buildings.Kolmanskop mining village. The diamond mine was to the South of the village. [Google Maps, June 2025]
As well as its railway link to Lüderitz, Kolmanskop was “the terminus of two private narrow-gauge electrified railway lines that served the diamond mining industry further south. One ran 119 kilometres (74 mi) via Pomona to Bogenfels. It was completed in 1913 but destroyed during World War I in 1915 by South African troops. The other railway line, 7 kilometres (4.3 mi) long and completed in 1920, led to Charlottental. Both were powered by a 1.5 megawatts (2,000 hp) power station in Lüderitz, then assumed to be the largest in Africa.” [1][34][35]
Kolmanskop is only a short distance from the Atlantic Ocean. The B4 and the railway find their own way down to the coast at Lüderitz.
The railway meanders down to the coast following the contours to keep the gradient to a minimum. On the way it passes through a turning triangle. It not obvious why the turning triangle is located at this point on the railway. However, not far beyond the triangle there are a series of sidings/loops which are shown in the image below.
Goods transfer sidings/loops close to Lüderitz. [Google Maps, June 2025]Just short of Lüderitz, the Atlantic is on the left of this photograph, the railway on the right. [Google Streetview, January 2024]Looking West along the railway towards the centre of Lüderitz. [Google Streetview, January 2024]Further along the line and looking North. [Google Streetview, January 2024]Looking North from the ungated crossing at Bay Street. The platform of the passenger station is just ahead. [Google Streetview, January 2024]
A photograph looking Southwest through the station can be viewed here. [37]
The engine shed and yard in Lüderitz are to the Northeast of the station platforms. [Google Maps, June 2025]The gates to the port at Lüderitz seen from Hafen Street. [Google Streetview, January 2024]The Port of Lüderitz. [Google Maps, June 2025]An aerial view of the port. [36]
The Railway Magazine of February 1952 carried an article by Charles E. Lee about railways in what was German South West Africa. This encouraged me to have a look at the history of Namibia’s railways and their condition and extent in the 21st century. The 1952 article also caught my attention because Manchester Diocese (I was a priest in Manchester Diocese before retirement) is linked with the Diocese of Namibia.
The territory was formally colonized by Germany between 1884-1890. It covered an area of 835,100 sq. km. It was a settler colony and had attracted around 3,000 German settlers by 1903, who primarily settled in the central high grounds. [2]
German South West Africa, now known as Namibia, was a German colony from 1884 to 1915. It was not a province within the German Empire but a separate colonial territory. From 1891, the capital was Windhoek, which also serves as the capital of modern-day Namibia. [2]
The arrival of German settlers disrupted the existing socioeconomic balance and led to conflicts, particularly with the Herero and Nama people.
“In 1883 Franz Adolf Lüderitz, a merchant from Bremen, Germany, established a trading post in southwest Africa at Angra Pequena, which he renamed Lüderitzbucht. He also acquired the adjacent coastal area, which he named Lüderitzland. These areas were constituted the first German colony under German protection on April 24, 1884. The German occupation subsequently extended inland. By the latter 1880s the German Colonial Company for the South realized that it was incapable of administering the territory, and the German government immediately took over the colony’s administration. As a result of the Zanzibar Treaty (1890) between Germany and Great Britain, German South West Africa acquired the Caprivi Strip (named after the German chancellor Graf Leo von Caprivi), a tract of land 280 miles (450 km) long in the extreme northeast of the territory; the colony thus gained access to the Zambezi River.” [3]
German colonial rule was harsh, leading to insurrections and resistance. “Major Theodor Leutwein, governor of the colony in 1894–1904, suppressed insurrections of the Khoekhoe (1894) and of the Hereros (1896). In 1904, however, the Hereros fomented a far more dangerous rebellion. The German force, at first only 750 strong and supported only by one artillery battery, had to face an army of some 8,000 men equipped with modern weapons. Reinforcements increased the German force, ultimately under the command of General Lothar von Trotha, and resulted in a decisive German victory on the Waterberg River. Further Khoekhoe rebellions were put down in 1904–07.” [3]
German South West Africa was occupied by the South African Union Defence Force in 1915 during World War I, and Germany formally ceded the territory under the Treaty of Versailles in 1919. Its administration was taken over by the Union of South Africa (part of the British Empire) and the territory was administered as South West Africa under a League of Nations mandate. It became independent as Namibia on 21st March 1990. [2]
The Railways
The railways in German South West Africa played a crucial role in the colonial administration and the First World War campaign. The German colonial authorities built a railway network between 1897 and 1914 to enable colonial territorialization and facilitate the extraction of resources. [4]
Charles E. Lee tells that “under the German regime, the first railway in South West Africa was the Northern State Railway (NSR), as it was then called, built to a gauge of 60 cm. (1 ft. 11 in.) between Swakopmund and Windhoek, via Jackalswater and Karibib, a distance of 238 miles. This line was begun in 1897 and was built by a German Military Brigade from Europe. It was first intended to be worked by animal power – Argentine mules or Cape donkeys – but steam traction was soon adopted. The first section (15 miles) was opened to traffic from Swakopmund in January 1898. By the end of that year 68 miles were ballasted and 54 open. In July 1900, the line was opened to Karibib, 121 miles, and the whole railway completed to Windhoek, a further 117 miles, in June, 1902. The curves and gradients were very severe, the gradient out of the Khan River gorge, for instance, being 1 in 19 with curves of 180 ft. radius. The rails weighed about 19 lb. a yard and were laid on iron sleepers. There were iron girder bridges at Khan River, Dorst River, and Kubas. The only good and plentiful water supplies were at Swakopmund and Karibib.” [1: p121]
Wikipedia tells us that there was actually an earlier line than the one Lee talks about. It was a small mining rail line at Cape Cross in 1895. [5] “Soon afterwards, the ox-cart transport system totally collapsed, in the wake of a rinderpest epidemic in 1897. As it was necessary to react quickly to the now extremely precarious transport situation, decisions were made: to build a railway line from the German port of Swakopmund to Windhoek (the Staatsbahn); to use existing, 600 mm (1 ft 11 5⁄8 in) gauge military Feldbahn material; and to entrust a railway brigade with the construction work, which began in September 1897.” [5]
Wikipedia continues: “Construction of the railways connecting with the Staatsbahn was aimed partly at military strategic objectives following the uprising of the Herero and Nama, and partly at economic requirements. … By World War I, the following lines had been developed (listed by the first year of full operation):” [5]
1902: Swakopmund–Windhoek line, 600 mm (1 ft 11 5⁄8 in) gauge, Karibib–Windhoek section re-gauged in 1911 to 3 ft 6 in (1,067 mm) gauge. [5]
1906: Otavibahn, 600 mm gauge. [5]
1905: Onguati–Karibib branch. [5]
1908: Otavi–Grootfontein branch. [5]
1907: Lüderitzbahn, 3 ft 6 in (1,067 mm). [5]
1909: Seeheim–Kalkfontein branch. [5]
ca 1911: Kolmannskuppe–Elisabethbucht–Bogenfels, industrial railway of the diamond fields. This 600mm gauge railway was electrified from 1911 (the only electric railway in Namibia’s history). Diamond mining in the region gradually moved south. The northern part of the line as far as Pomona was abandoned in 1931, and some of its materials were used for the extension of the railway towards Oranjemund. The southern section was operated with diesel traction. This line no longer exists. [5]
1912: Windhoek–Keetmanshoop railway, 3 ft 6 in (1,067 mm) gauge. [5]
1912: Rehoboth shuttle, 600 mm (1 ft 11 5⁄8 in) gauge (questionable). [6][7][2][5]
1914: Otjiwarongo–Outjo–Okahakana, 600 mm gauge (project started, but not completed due to the war). [5]
Lee talks of the formation, by the Otavi Mining & Railway Company, an Anglo-German syndicate owning the copper mines at Otavi and Tsumeb, of a railway: “This company was formed in Berlin in 1900, in accordance with an arrangement between the South-West Afrika Company, the Disconto-Gesellschaft of Berlin, and the Exploration Company. The first intention was to build a 3 ft. 6 in. gauge railway from Port Alexander in Portuguese West Africa to run in a south-easterly direction up the Muende River Valley and via Etosha Pan to the Tsumeb Copper Mines, and later to extend this line to Rhodesia to form a trans-African railway. Eventually it was decided to form a 60 cm. gauge line entirely in German territory connecting Swakopmund with Tsumeb, a distance of 351 miles. Construction was undertaken by Arthur Koppel & Co. and was begun in November 1903, but was delayed by the Herero War, and the work completed on 25th August 1906. This undertaking, called the Otavi Railway, had the distinction of being the longest narrow-gauge railway in the world. Branches were laid subsequently from Otavi to Grootfontein (56 miles) and from Onguati to Karibib on the State Railway (9 miles). The cost is stated to have been about £2,400 a mile, or roundly £1,000,000 in total. The railway was bought by the German Imperial Government in 1910 for £1,250,000, but the management was left in the hands of the company under a 30-year lease, terminable after 10 years.” [1: p121]
This line was well constructed, and well ballasted. It had a ruling gradient of 1 in 66 and minimum curvature of 150 metres. The permanent way consisted of steel rails in 30-ft. lengths, 30 lb. a yard, laid on steel sleepers weighing about 26 lb. each. “From Swakopmund, for a distance of 68 miles, the line rises steadily on a grade of 1 in 66 to Ebony Station, where it reaches an altitude of 3,500 ft. (On the down journey, the last 40 miles into Swakop-mund can be run by gravity.) From Ebony there is a regular fall to Usakos, which is 2,640 ft. above sea level. From Usakos it climbs 690 ft. in 13 miles to Onguati, and continues to rise until it attains its greatest elevation near Kalk-feld, where the summit is 5,200 ft.” [1: p121]
“The Otavi Railway, like the State Railway, was built to the 2 ft-gauge, though a difference of 1 centimetre in the wheel gauges is stated to have prevented the free interchange of rolling-stock. The widening to 3 ft. 6 in. of the gauge between Swakopmund and Omaruru had been voted by the German Railway Board, but the work had not been put in hand by the outbreak of the 1914 war. A new branch projected at the same period was the Ovamboland Line, the first aim of which was to provide Ovambo labour for the South. The Landesrat in November 1913, approved a line of 2 ft-gauge, but on earthworks and bridges wide enough for a 3ft. 6in. gauge track, to run from Otjiwarongo (on the Otavi Railways) to Outjo and Okahakana.” [1: p121]
Railways in South West Africa from Swakopmund, mainly German- built, included the 361 miles to Tsumeb, opened in 1906, and the longest narrow-gauge railway in the world. The gauge at the Southern end was widened in 1915. [1: p122]
A sum of £450,000 was allowed for the line from Otjiwarongo to Outjo and Okahakana “in the German Loan Estimates for 1914-15. The first section, including the 55 miles from Otjiwarongo to Amiab Poort, was to cost £250,000. Construction was begun, and the line was laid for 22 miles before the outbreak of hostilities in the first world war.” [1: p123]
“Railway developments south of Windhoek, on the 3 ft. 6 in. gauge, made it desirable to convert the earlier 2ft. lines. During 1911, the section from Karibib to Windhoek was converted to 3 ft. 6 in. gauge at a cost of £550,000, with the Bechstein-Koppel Gesellschaft as contractor. The ruling gradient [was] 1 in 66 with a minimum curvature of 656 ft. This work was completed during 1913. The Swakop River at Okahandja [was] spanned by a bridge 350 ft. long, and there [was] a smaller bridge at Otjihavera. About the same time, the coastward section from Karibib to Swakopmund was practically abandoned in favour of the alternative route provided by the Otavi Railway. In fact, the settlers in the Swakop Valley, who asked for a short railway to link them with Swakopmund, were promised in November 1913, that the material from the disused 92 miles of the State line between Swakopmund and Kubas would be used for this purpose, but it was not done.” [1: p123]
An image showing an armoured train in South West Africa during World War I, 1914-1918, can be found here [29] The South African army invaded the German colony of South West Africa in March 1915 overrunning the much smaller German forces.
Wikipedia tells us that, “With the outbreak of World War I, the German Schutztruppe military unit retreated from the coast, and withdrew into the inland. In the process, the Schutztruppe destroyed the Otavibahn, and the old Staatsbahn towards Karibib, as far as Rössing.” [5]
The Staatsbahn was abandoned but this was not the case with the Otavibahn. In 1914, “British troops … moved forward from the British enclave of Walvis Bay, and by the end of 1914 they had built a 37 km (23 mi) long 3 ft 6 in (1,067 mm) railway to Swakopmund. The Otavibahn was also reconstructed in 3 ft 6 in (1,067 mm) as far as Usakos, and the section between Usakos and Karibib was realigned. The network north of Usakos remained in 600 mm (1 ft 11 5⁄8 in) gauge; the workshop for both gauges was consolidated in Usakos, and the one in Karibib was closed.” [5]
Lee tells us that by 1917 the Staatsbahn line from Karibib to the coast had ceased to exist. “the line between Karibib and Rossing (95 miles), the 10-mile branch from Jakalswater (built to carry water from the Swakop River at Riet), and the Kubas military line (4.5 miles), were lifted and removed to provide material for Tanganyika and the Union of South Africa.” [1: p123]
Lee goes on to confirm that the Union forces, in the course of their invasion of German South West Africa, “laid a 3 ft. 6 in. line for 100 miles inland from Swakopmund to Kranzberg along the original track of the Otavi line, which the Germans had wrecked in their retreat. This was completed in August, 1915. The construction of a new 12.5-mile section, of the same gauge, from Kranzberg to Karibib, was completed in July 1915, and again connected the Otavi Railway with the [NSR]. Thus, in August 1915, there was continuous communication of uniform gauge for the first time from Swakopmund to points south of Windhoek. As strategic railways had meanwhile linked the Union Railways with those of South-West Africa on 25th June 1915, a through railway of 1,635 miles was provided between Walvis Bay and Cape Town.” [1: p123]
Also during the first world war, a new railway from South Africa was constructed – “as an extension of the De Aar-Prieska Railway – to achieve a secure supply route for … South African troops. In 1916, the line was connected to the German network at Kalkfontein (now Karasburg).” [5]
“With the linking of the Kranzberg-Tsumeb 2ft-gauge line to the workshops at Usakos by means of a third rail between Usakos and Kranzberg on the 3-ft. 6-in. gauge track of improved location, the 9-mile section from Karibib to Onguati was no longer of value, and it was uplifted in 1924.” [1: p123]
“The former Otavi Railway system [was] therefore represented [in 1952] by about 100 miles of 3 ft. 6 in. line on the coastward section, part of the main railway system of South-West Africa, and 307 miles of 2ft-gauge farther inland. [In 1952, there were] also various private branch lines (some disused) connected with the 2ft section. [In 1952], the present main line of this gauge [was] from Kranzberg to Tsumeb, some 251 miles, on which one train in each direction [was] run two days a week.” [1: p123]
Wikipedia continues: Under South African/British occupation, the following lines were established (listed by first year of full operation): [5][10]
1914: Walvis Bay–Swakopmund in 3 ft 6 in (1,067 mm). [5]
1915: Swakopmund–Karibib: Reconstruction in 3 ft 6 in (1,067 mm). [5]
1915/1916: (De Aar)–Nakop (border)–Kalkfontein in 3 ft 6 in (1,067 mm). [5]
1921: Otjiwaronge–Outjo 600mm gauge (based on German preparations). [5]
1929: Windhoek–Gobabis railway in 3 ft 6 in (1,067 mm). [5]
From 1958: the Otavibahn north of Usakos was gradually regauged to 3 ft 6 in (1,067 mm), with the new line being laid parallel to the existing line, but largely on new foundations; the new line was in operation from 1961. [5]
“From August 1915 the Namibian railway network was operated de facto by South African Railways, and this arrangement became official in 1922. … From 1959, steam locomotives were gradually replaced by diesel locomotives, for which an engine-house was built in Windhoek. This made operations very much easier, because water is in short supply in Namibia, and the coal needed to heat the water in the steam locomotives also had to be procured from the Transvaal.” [5]
The Namibian Network in the 21st century
In the 21st century, the rail network of Namibia is operated by TransNamib. As of 2017, the Namibian rail network consisted of 2,687 km of tracks. [11]
The railway line from Windhoek to Kranzberg is 210 kilometres (130 miles) long and was completed in 1902. [10]
Windhoek (capital – junction)
Okahandja
Karibib (proposed cement works)
Kranzberg (junction Tsumeb v Windhoek)
After the aerial image immediately below, the next three images form a kind of ‘tryptic’ which shows the TransNamib train yard and station at Windhoek. Taken together they show the full site. …
Wikipedia tells us that “the station was built in a Cape Dutch-style and is located on Bahnhof Street. An additional northern wing was constructed by South African Railways in 1929 to match the existing style of the building. … The station also houses the small Trans-Namib Railroad Museum which outlines Namibian transport history, particularly that of the railway. Opened on 1st July 1993, the exhibition consists of a wide range of railway equipment, maps and related items which date back to German colonial times. Another part of the exhibition is dedicated to Namibian Airways history and Namibian Maritime history. … Across from the entrance [to the station] stands the German locomotive ‘Poor Ole Joe’, one half of a South West African Zwillinge, No 154A, the sole surviving specimen of this type of steam locomotive. It was originally shipped to Swakopmund in 1899 and reassembled for the run to Windhoek” [23][24]
Namibia Scientific Society posted the following on Facebook on 9th June 2020: Poor Ole Joe is a 600mm-gauge steam locomotive “and was manufactured in 1900 by Henschel & Sohn GmbH, Kassel, Germany, under the serial number 5376. It was put into operation in 1904 and operated on the Swakopmund – Windhoek route. The steam locomotive was taken out of service in 1939 after traveling approximately 371,000 miles.” [25]
There is some uncertainty over the date of fabrication of the locomotive. Perhaps the two years mentioned relate to a date when the locomotive was shipped from the factory and the date of completion of the reassembly in Swakopmund?
The railway line from Kranzberg to Walvis Bay is 201 kilometres (125 miles) long. The section between Kranzberg and Swakopmund was completed in 1902. In 1914, an extension to Walvis Bay was commissioned; the rails were laid close to the shore of the Atlantic Ocean. In 1980, this extension was replaced by an alternative route behind the dunes that allowed for higher axle load. [10]
Kranzberg (junction Tsumeb v Windhoek)
Usakos
Arandis (crossing loop)
Swakopmund
Walvis Bay (port)
Looking back Northeast towards Kranzberg Railway Station from the B2. [Google Streetview, 2024]Looking Southwest along the railway towards Usakos’, Arandis and Swakopmund. [Google Streetview, 2024]
Key locations along the line to Swakopmund are illustrated below: …
Before having a look at the Rossing Uranium Mine, it is worth a quick diversion Northwest of the station and marshalling yard shown above. The Namibia Institute of Mining & Technology is host to a plinthed display of a locomotive and carriages from the old 2ft-gauge railways of Namibia.
This image shows a complete (but short) 2ft-gauge train at the Namibia Institute of Mining and Technology. [Google Streetview, 2024],
This train was once on display in Windhoek. It was moved to the Namibia Institute of Mining Technology (NIMT) outside Arandis. and restored with the help of Wesbank Transport and AWH Engineering, Rigging and Rentals. The locomotive, is a Henschel Hb 56. The locomotive and its wagons were in use between Usakos and Tsumeb between 1906 and 1959. The South African Railways then donated it to the National Museum in Windhoek and in 1964 it was placed in front of the Alte Feste, but it was too close to the Reiterdenkmal and was moved in 1974 to the southern side. The train consists of the locomotive, a coal wagon, a closed goods wagon, a passenger coach for first and second class and a wagon in which the conductor travelled with the mailbags, milk and cream cans that were picked up along the route. The passenger coach could transport 16 passengers. The first-class passengers could sit on upholstered seats while the second-class passengers sat on plain wooden benches. The two classes were divided by a small washroom. The conductor’s wagon was destroyed in 2007 when it was set alight by a homeless person who slept in the train and made a fire. The boilermaker and carpentry students at NIMT renovated the train. [35]
“The locomotive is from the class Hb 0-6-2T. Of the 15 locomotives built by Henschel for the Otavi line between 1905 and 1908, six were absorbed into the SAR. The engines had Allan valve gear and often ran with an auxiliary tender attached which contained both coal and water.” [36]
Walvis Bay was a British enclave in German South West Africa. The first narrow gauge railway in the British ruled Cape Colony was in Walvis Bay. Initially projected merely to connect the jetty with the town, the Walvis Bay Railway was opened in 1899 and ran for twelve miles up north to the German border at Plum. [17]
“On 6th March 1899 the Agent General for the Cape of Good Hope ordered a “Sirdar” class locomotive named ‘Hope’ which was almost as long in transit to Walvis Bay – where it arrived on 22nd August 1899 on board the British barque Primera – as it had been in the building. Because of the extremely light nature of the track (12 lb. rail with sleepers spaced three feet apart) HOPE was provided with an additional pair of carrying wheels at both ends. Thus the standard 0-4-0T type was converted to a 2-4-2T type. Even so the maximum axle load of ‘Hope’ in working order would be about 1¾ tons, which is considerably more than today’s suggested figure for this category of track of 1 ton 4 cwt. Within six years the railway was virtually moribund and by 1915, ‘Hope’ had been laid aside and forgotten. That was because the Germans preferred to use their own harbour in Swakopmund.” [17][18]
Two works photographs of ‘Hope’: in the one with the valance (wheel cover) raised, one of the smaller carrying wheels can just be made out on the left of the picture. [17][18]
Kranzberg-Otavi
The railway line from Kranzberg to Otavi is 328 kilometres (204 miles) long and was completed in 1906. [10]
Kranzberg (junction Tsumeb v Windhoek)
Omaruru
Kalkfeld (short siding)
Otjiwarongo (junction for Outjo)
Otavi
Kranzberg Railway Station has already been featured above. The next images show the line from there to Otavi. …
Kranzberg Railway Station. [Google Streetview, June 2025]
The loop allows trains from Windhoek to access the route to Otavi without reversing. That line running towards Otavi sets off from Kranzberg in a Northeasterly direction crossing a series of dry watercourses and gradually taking a more northerly course before encountering the D2315 (a dirt road).
From Kalkfeld the line heads in a generally Northeasterly direction towards Otjiwarongo.
As on the earlier length of the line, we see it crossing a number of dry river beds. [Google Maps , June 2025]
The next five images are a sequence which shows a long passing loop, perhaps halfway towards Otjiwarongo.
A sequence of five images shows a passing loop. The sequence has the Northeast end of the loop in the first of the five images and the Southwest end of the loop in the fifth image, immediately above. [Google Maps, June 2025]
The next five images show a sequence of structures over dry river beds
Five bridges spanning dry watercourses. [Google Maps, June 2025]This photograph is taken from the C33 which has followed the railway Northeast towards Otjiwarongo. [Google Streetview, 2024]Approaching Otjiwarongo, this photograph faces East-northeast from alongside an ungated crossing around 50 metres Southeast of the C33. [Google Streetview, 2024]This photograph faces East-northeast along the approach to Otjiwarongo Railway Station. The road from which it is taken is the C38. [Google Streetview, 2024]Otjiwarongo Railway Station is a junction station with line onward to Otavi and Outjo. [Google Maps, June 2025]Otjiwarongo Railway Station building. [Google Streetview, 2024]Otjiwarongo Goods Shed. [Google Streetview, 2024]In 1912, Henschel built three 2-8-2 tender engines No. 40, No. 41 and No. 42 for the Otavi line for use on the Swakopmund-Karabib section. No. 41 is plinthed outside Otjiwarongo Railway Station. Like many other SWA locos they had dust covers to protect the motion. The carrying wheels were arranged as radial axles. As there were no separate bogie truck, the axle boxes were guided in such a way that the wheels could move radially with respect to the frame. At that time the railway was a 2ft-gauge line [Google Streetview, 2024] More information can be found here. [39]
The line to Otavi continues heading Northeast. …
The line to the Northeast of Otjiwarongo Railway Station. [Google Maps, June 2025]At the bottom-left of the image above the line crosses the C33 at an ungated crossing. [Google Streetview, 2024]A closer satellite view of the length of the line to the Northeast of the C33. A few sidings serve industries to the South of the line. The road at the centre of the image running North-South is Industria Street. [Google Maps, June 2024]Looking West from Industria Street. [Google Streetview, 2024]Looking Northeast from Industria Street. [Google Streetview, 2024]The B1 to the Northeast of Otjiwarongo bridges the line. This view looks Southwest towards the railway station. [Google StrLooking Northeast towards Otavi from the B1. [Google Streetview, 2024]The line runs parallel to the B1 heading Northeast. [Google Streetview, July 2024]Looking Northeast along the line from an ungated crossing at the D2430. The B1 can just be seen on the left of this image. [Google Streetview, July 2024]A little further Northeast this photograph, taken from the B1, shows a minor road crossing the railway at an ungated crossing. [Google Streetview, July 2024]As we travel Northeast, the landscape becomes greener. This another view looking East from the B1 and shows another ungated crossing of a minor road. [Google Streetview, July 2024]The line passing under the B1. The landscape has changed. The line is running through dense shrubs and small trees. [Google Streetview, 2024]In Otavi, this is Phyllis Street. It crosses the line at the Southwest end of the station site. [Google Streetview, 2024]Otavi Railway Station and turning triangle. [Google Maps, June 2025]Otavi Railway Station building. [Google Streetview, 2024]
It is worth noting here that the original gauge of the line from the coast to Otavi and Tsumeb was originally built to 2ft-gauge. Later it was converted to 3ft 6in gauge. The line was built for the Otavi Mining and Railway Company (Otavi Minen- und Eisenbahn-Gesellschaft or OMEG). The company was founded was a railway and mining company in German Son 6th April 1900 in Berlin with the Disconto-Gesellschaft and the South West Africa Company as major shareholders. [41]
The first locomotives designed for regular service were fifteen 22-tonne 0-6-2T locos built by Arn. Jung. [41][42: p45] Henschel & Sohn built twelve locomotives similar to the Jung design and three 0-6-0T locos. [41][42: p45] Twenty 8-wheel auxiliary tenders carrying 8 cubic metres of water and 3.5 tonnes of coal were built to enable these tank locomotives to complete longer runs. [41][42: p45][43: p65] Henschel & Sohn built three HD class 2-8-2 in 1912 with separate 8-wheel tenders for long-distance running. [42: p47] These locomotives weighed 59 tonnes (including the 26-tonne tender) and remained in service for 50 years as the 2-8-2 type became standard for the railway. [41]
By 1913, train service included 4 express trains, 14 mixed trains, and 29 freight trains each week. [42: p39] Express and mixed trains included a baggage car, a car for African passengers, and a coach for first and second class passengers. [42: p39] The passenger coaches carried concrete ballast in a depressed center section to minimize the possibility of wind tipping a lightly loaded car off the rails. [43] Express trains stopped only at designated stations, but other trains would stop at intermediate points when transport was required. [42: p39] Equipment included: 96 low-side ore gondolas; 55 high-side gondolas; 20 limestone gondolas; 20 boxcars; 12 tank cars; 4 stock cars; 3 passenger coaches; and an executive business car with a kitchen, a bathroom, and an office convertible to a bedroom at night. [41][42: p42][43: p65]
There were also some self-powered steam rail cars with a coal bunker, a mail compartment, 2 compartments for Europeans, and 4 for Africans. [41][42: p36]
Otavi-Grootfontein
The railway line from Otavi to Grootfontein is 91 kilometres (57 miles) long and was completed in 1908. [10]
Otavi (junction for Grootfontein)
Grootfontein (branch terminus)
Otavi Railway Station, seen from the C39 at the Northeast end of the station site. [Google Streetview, 2024]From the same location on the C39 a wider view shows the sidings at Otavi Railway Station [Google Streetview, 2024]Turning through approximately 180° and looking Northeast, the line to Tsumeb runs towards the hills at the left of the image. The line to Grootfontein curves away to the right. [Google Streetview, 2024]Looking back towards Otavi Railway Station from the ungated crossing on Josef Buchholz Avenue. [Google Streetview, 2024]Turning through 180°, this is the view Sputheast from Josef Buchholz Avenue towards Grootfontein. [Google Streetview, 2024]Heading Southeast out of Otavi the line to Grootfontein passes under the B1. This is the view along the line from the road and bridge. [Google Streetview, 2024]Out of Otavi, the line soon starts to accompany the B8 in its journey East. This photograph is taken from the B8 and shows an ungated crossing on a minor road. [Google Streetview, 2024]An ungated crossing provides access from the B8 into Kombat. The road is the D2863. This is the view East at the crossing. [Google Streetview, 2024]
The line turns away from the B8, to the North. As it does so it crosses the D2860 at an ungated crossing.
The line to Grootfontein crosses the D2860 at an ungated crossing. [Google Streetview, 2024]
The line follows the D2860 and then the D2905 before passing under the B8, as it heads for Grootfontein.
An ungated crossing to the South of the D2905. [Google Streetview, 2024]The view ahead along the line towards Grootfontein from the B8. [Google Streetview, 2024]The view towards Grootfontein from a minor road ungated crossing. [Google Streetview, 2024]Much closer to Grootfontein, another view East along the railway. [Google Streetview, 2024]The fuel depot at Grootfontein. [Google Maps, June 2025]Grootfontein Railway Station. [Google Maps, June 2025]Grootfontein Railway Station in 2007. This image was shared on the African Railway Station Stopping Places Facebook Page In 2012. [46]Grootfontein Railway Station goods depot in 2007. This image was shared on the African Railway Station Stopping Places Facebook Page In 2012. [47]
Grootfontein railway station is being converted into a logistics hub for business with the DRC and Zambia.
At the moment, trucks from the DRC, Zambia or Namibia travel about 2,500 kilometres from Walvis Bay harbour to Lubumbashi. With the introduction of the Grootfontein hub, these trucks will travel a distance of about 1,400 kilometres. TransNamib is prepared to dedicate four trains a week for this business idea. [44]
Immediately to the East of the railway station the line turns to the South and is clearly not well used and significantly overgrown in places. [Google Maps, June 2025]After a few hundred metres the line turns to the East. It can clearly be made out towards the bottom of this satellite image. [Google Maps, June 2025]The line continues East and passes under the D2830. [Google Maps, June 2025]Looking West from the D2830, a short length of the line can be seen just to the right of the centre of this image. [Google Streetview, 2024]To the East of the D2380 a series of sidings still exist. [Google Maps, June 2025]It is harder to make out the sidings in this view. The photograph looks East from the D2380. [Google Streetview, 2024]These last two satellite images show the extent of the tracks in the industrial area to the East of the D2380. [Google Maps, June 2025]
Otjiwarongo-Outjo
Otjiwarongo (junction for Outjo)
Outjo (railhead)
Otjiwarongo Railway Station is illustrated above. The railway line from Otjiwarongo to Outjo is 69 kilometres (43 mi) long. The first 26 kilometres (16 mi) were completed under the German colonial administration in 1914/1915; the railway line was named Amboland Railway in reference to the territory of the Ovambo people. The link to Outjo was completed in 1921 under South African rule. [10]
The branch line to Outjo can be seen turning away North from the line to Otavi. [Google Maps, June 2025]The branch line crossed the C33 at an ungated crossing. This photograph looks South from the C33 towards Otjiwarongo Railway Station. [Google Streetview, 2024]Turning through 120°, or perhaps more, standing on the C33, the rails of the line to Outjo disappear into the vegetation. The line has clearly not been used for some time. However, we will see that much of the line to Outjo remains in place and perhaps could be renovated should the need arise. [Google Streetview, 2024]The line curves round towards the West. On the way it appears often out of the undergrowth. Here, this minor road crosses the old railway and the signs still stand proudly either side of the line, either side of the railway. [Google Streetview, 2924]The road shown above appears bottom-right of this image. [Google Maps, June 2025]The line then heads Southwest for a while before gradually turning through the West to the Northwest. [Google Maps, June 2025]The line appears out of the brush quite often and sometimes for significant distances, as these two. [Google Maps, June 2025]These two images are typical of what can be seen on satellite imagery. The line appears out of the brush quite often and sometimes for significant distances. [Google Maps, June 2025]The line turns through West to Northwest. [Google Maps, June 2025]It continues, Northwest. [Google Maps, June 2025]One passing loop appears out of the undergrowth. [Google Maps, June 2025]Here it can be seen crossing another minor road. [Google Maps, June 2025]And then a tarmac road. All crossings are ungated. [Google Maps, June 2025]
The next series of six photographs show sidings parallel to the running line. This location is more than just a passing loop but I have not been able to establish whether a specific local industry was the reason for the sidings. The photographs run in sequence Southeast to Northwest. …
The last of six photographs of sidings adjacent to the line to Outjo. [Google Maps, June 2025]
The next sequence of four photographs shows a passing loop on the line. In sequence, these photographs run from the Southeast to the Northwest. …
In 2005, a new 89 km section of Northern Railway from Tsumeb to Oshivelo was opened by President Sam Nujoma, as part of the “Northern Extension” of the railway link from Kranzberg to Otavi. Construction on the project’s second phase, a 59 km stretch from Ondangwa to Oshikango on the Angolan border at a cost of about N$329m, was scheduled to be completed by December 2007. Ondangwa Station opened in 2006 for freight.
In phase 3, a 58 km branch from Ondangwa to Oshakati was constructed at an estimated cost of N$220m, for completion in December 2008. For the future a connection from Oshikango to a point near Cassinga is planned on Angola’s southern railway system. [11][13][14]
The Ondangwa-Oshikango line was officially opened by President Hifikepunye Pohamba in July 2012. In order to keep the system operational and safe, provincial governor Usko Nghaamwa implored local residents to stop stealing railroad ties and sections of the wire fence. [11][15]
Otavi (junction for Grootfontein)
Tsumeb
Ondangwa (junction)
Oniipa (road bridge)
Onjdiva [11][14]
Namacunde [11][16]
Oshakati
Oshikango (Angolan border)
The C39 crossed the railway immediately to the North of Otavi Railway Station. as we have already noted, this view from the ungated crossing shows the branch to Grootfontein heading away to the right and the line North-northeast to Tsumeb heading for the distant hills. [Google Streetview, 2024]
The journey towards Tsumeb runs uneventfully over flat ground surrounded by shrub and small trees, heading North-northeast, until it reaches Ohorongo Cement Works.
An aerial view of the works can be found here. [48] That view looks North across the Works and shows the railway and a dedicated branch to the Works in the background.
Ohorongo Cement Works. [Google Maps, June 2025]The passing loop and access to the cement works’ private sidings. [Google Maps, June 2025]The dedicated siding can be seen leaving the main line at the Southwest end of the passing loop. [Google Maps, June 2025]The siding curves round along the Northeast side of the Works. [Google Maps, June 2025]The siding ends towards the Northeast corner of thecsite
The railway continues Northeast over largely unremarkable flat terrain, before turning East, encountering one arm of the B1 and then a triangular junction.
Encyclopedia Britannica tells us that “In 1851 Sir Francis Galton, a British explorer, made note of copper ore deposits in the vicinity of what later became the town of Tsumeb. An Anglo-German company acquired mining rights for the Tsumeb area in 1903. Southwest of Tsumeb is the site of the final German troop surrender to South African forces in World War I. The town remained a small copper-mining centre until the Tsumeb mine was purchased in 1947 by a largely U.S.-based corporation. It has since been developed as a planned company town (although ownership of the mine has changed hands several times), exploiting mineral deposits that include significant amounts of lead and copper as well as zinc, cadmium, silver, and germanium (a metalloid element used as a semiconductor). An integrated copper and lead smelter treats concentrates from Tsumeb and other mines. Owambo labourers are the chief contract workers.” [50]
The mine, owned by Dundee Precious Metals sits to the East of the B1.
The line to the North of Tsumeb left the triangular junction to the West of the town heading first to the West and then to the Northwest and then directly North alongside the D3007, before turning West-northwest again.
Looking Southeast from the B1 towards Tsumeb. [Google Streetview, 2022]Looking Northwest from the B1 along the line towards Omuthiya. [Google Streetview, 2022]The ungated crossing at the D3007. [Google Maps, June 2025]
After a few kilometres on a West-northwest heading, the line then turns to the North-northwest and runs parallel to the B1 for some considerable distance.
The B1 and the railway converge and head North-northwest. [Google Maps, June 2025]The line seen from the B1. [Google Streetview, 2022]The ungated crossing on the D3004. [Google Streetview, 2022]The view North-northwest along the line from the ungated crossing on the D3001. [Google Streetview, 2022]The line diverges from the B1 just to the South of the River Owambo. Both the railway and the road cross the river in this satellite image. [Google Maps, June 2025]Triangle on the South side of the D3610 at Oshivelo. [Google Maps, June 2025]Oshivelo Railway Station on the North side of the D3610. [Google Maps, June 2025]Looking South East from an ungated crossing just to the Southeast of Omuthiya Railway Station. [Google Streetview, 2022]Looking Northwest from the same ungated crossing into the site of Omuthiya Railway Station. [Google Streetview, 2022]Omuthiya Railway Station. [Google Maps, June 2025]The line to the Northwest of the station, seen from the South. [Google Streetview, 2024]A short distance further up the line looking back towards Omuthiya. [Google Streetview, 2024]A little further Northwest again, this time looking North towards Ondangwa. [Google Streetview, 2024]Looking back towards Omuthiya from the ungated crossing on the D3603. [Google Streetview, 2024]At the same ungated crossing, this photograph is taken looking forward towards Ondangwa. [Google Streetview, 2024]Two culverts then take the line over the dry channel of the River Gwashigam. [Google Maps, June 2025]Looking back Southeast from the bridge carrying the D3622 over the line on the approach to 0ndangwa. [Google Streetview, 2024]Looking Northwest from the same bridge towards Ondangwa. [Google Streetview, June 2025]Looking South-southeast from an ungated minor dirt road crossing closer to Ondangwa Railway Station. [Google Streetview, 2024]Looking West-northwest towards Ondangwa. [Google Streetview, 2024]Ondangwa Railway Station and turning triangle. [Google Maps, June 2025]Looking South East from the B1 overbridge into the site of Ondangwa Railway Station. [Google Streetview, 2024]The view West from the same bridge across the turning triangle, the arm on the right leads to the line heading North towards the Angolan border. [Google Maps, June 2025]Fuel depots sit alongside the line as it heads North. [Google Maps, June 2025]Looking North from the bridge carrying the C45 over the railway which is now closing in on the railhead on the Angolan border. [Google Streetview, 2024]
The next three images are a sequence of North-facing photogra
The last photograph on the northern line is a satellite image showing the railhead
The railhead in Oshikango at the Namibia/Angola border. [Google Maps, June 2025]
References
Charles E. Lee; The Longest Narrow-Gauge Railway; in The Railway Magazine, February 1952, Tothill Press, Westminster, London, p121-123.
Helmut Schroeter; Die Eisenbahnen der ehemaligen deutschen Schutzgebiete Afrikas und ihre Fahrzeuge = Die Fahrzeuge der deutschen Eisenbahnen 7 [The Railways of the former German Protectorates in Africa and their Rolling Stock = the Rolling Stock of the German Railways 7]. (in German); Verkehrswissenschaftliche Lehrmittelgesellschaft, Frankfurt am Main, 1961.
Helmut Schroeter and Roel Ramaer; Die Eisenbahnen in den einst deutschen Schutzgebieten: Ostafrika, Südwestafrika, Kamerun, Togo und die Schantung-Eisenbahn: damals und heute [German colonial railways: East Africa, Southwest Africa, Cameroon, Togo and the Shantung Railway: then and now] (in German and in English); Röhr-Verlag, Krefeld, 1993.
Brenda Bravenboer and Walter Rusch; The First 100 Years of State Railways in Namibia; TransNamib Museum, Windhoek, 1997.
According to Schroeter; Bravenboer does not mention this line.
Frederic J. Shaw; Little Railways of the World; Howell-North, Berkeley, California, 1958.
Dick Andrews; Extra Narrow Gauge Junction: Otavi Ry., State Northern Ry. in South Africa [sic]; in Narrow Gauge and Short Line Gazette, Volume 16 No. 1, 1991, p63–66.
Roger Farnworth 
.jpg){kind=link}
_58_class_Garratt_locomotive_no_5804.png){kind=link}
.jpg){kind=link}
.jpg){kind=link}
.jpg){kind=link}
.jpg){kind=link}
.jpg){kind=link}
.jpg,baccrssef){kind=link}
