Tag Archives: Tanganyika

The Mkumbara to Neu Hornow Cableway/Ropeway, Usambara Hills, German East Africa/Tanganyika.

The western Usambara Hills were characterised by precipitous cliffs and deep gorges. The provision of a rail link between Mkumbara and Neu Hornow was not considered practical.

A 9 km long ropeway was constructed, under the ownership of “the firm of Wilkens and Wiese, and designed to carry cedar from the Shume plateau to the railway, an enterprise that was never an economic success. The longest span of the ropeway, 907 metres, was said to be the longest in the world when it was built in the years 1910-1911.” [1: p75] Wood was transported via the Goatal/Ngoha Valley in the Schumewald/Shume Forest. [2]

The ropeway was constructed by Adolf Bleichert & Co. a German company primarily active in  cableway construction . It was founded in 1876 by Adolf Bleichert and was headquartered in Leipzig – Gohlis from 1881. [2] More information about Adolf Bleichert & Co. can be found here. [3]

What follows here is a translation of a German language text with the associated images. [4]

Wilkens & Wiese were aware, when negotiating with the German authorities for a concession to harvest timber saplings in the western Usambara hills and particularly the Schumewald forest, of the difficulty of connecting the steep high plateau with the railway in the plain. A railway or road would have been completely out of the question due to the sharp, steep, and heavily forested slopes of the hills. The only option available to them was a cableway to connect the high plateau with the then-planned station of Mkumbara on the Usambara Railway. In anticipation of the expected difficulties, the cableway was ordered from Adolf Bleichert & Co. in Leipzig-Gohlis andwork commenced in the spring of 1910. The location of the cable car and the timber concession of Wilkins & Wiese are shown below: [4: p17]

Site plan of the enterprises of the plantation company Wilkins and Wiese in Vestusambara. [4: p17]

The undertaking faced enormous difficulties due to the steep mountain slope. Furthermore, the rock was crumbly and easily weathered, so landslides often disrupted the work. A shortage of workers, the construction of new roads to transport building materials, and last but not least, Sandfly and Mosquitoes tormented workers and hindered completion throughout. Only through sheer energy and great sacrifice was it possible to complete the work. [4: p17-18]

By 1911, the system shown below was fully operational and transported sawn timber, beams, and logs from the sawmill located on the high plateau at 2000 m above sea level to the Mkumbara railway station on a regular schedule. Its horizontal length is 9.0 km, with a height difference of 1435m between the terminal stations. The greatest difference in elevation of the cableway is 1523 m, as shown in the longitudinal profile below. Due to the extremely unfavourable conditions, the line had to be divided into three sections, the uppermost of which first has to overcome a climb of about 90 metres. Therefore, a traction system had to be provided for all eventualities, which would assist if the gradient became too heavily congested with wagons. The cableway’s capacity was designed for ten tons per hour downhill and one tonne per hour uphill. [4: p18]

A longitudinal profile of the Mkumbara tto Neu Horow Cableway. [4: p18]

The line began at the loading station near the Neu-Hornow Sawmill at an altitude of approximately 2000 m above sea level. At about 1.2 km from the sawmill, it crosses the edge of the plateau. It then descends quite steeply to an altitude of 1290 m, where it turned through a 45° angle. From here, the cableway had to be routed to a breakpoint, where it turned once again seeking suitable locations for the support towers. The line then continued with two spans of more than 300 metres each to another breakpoint, the junction station at an altitude of 770 metres. Then the cableway heads for Mkumbara, crossing the uniquely beautiful Ngoha Valley with a free span of 100 metres. Prior to reaching the bottom station at an altitude of 68 metres, the railway has a tensioning and anchoring station at 660 metres and a double tensioning station at 170 metres. The journey of a load takes about one hour.

The Neu-Hornow sawmill has several standard frame saws on which logs can be cut into beams and processed into lumber. The loading station, shown below, is equipped with fixed hanging rails. In addition to the necessary guide rails for operation, it also has a storage area for empty hangers.

The cableway loading station at the sawmill. [4: p19]


The points at which hangers attach and detach from the cable/rope are visible on the right of the drawing. These points allow the incoming cars to detach automatically from the haul rope, while the outgoing cars automatically reconnect to the haul rope. Patented Bleichert apparatus is used as the attaching device. [4: p19]

The end guide pulley was equipped with two hand brakes, each with a disc diameter of approximately 2 metres, capable of braking 50 horsepower, with one serving as a safety brake. The brakes were only applied when the train was stationary. During operation, an automatic brake regulator controlled the train speed. This regulator (a hydraulic brake), along with the cableway’s drive system, was housed in a separate engine room next to the loading station. There was a 1.6 m³ reservoir on the roof of the loading station and two concrete tanks in front of the engine house, in which water supplies for the summer were collected. [4: p19]

The cableway needed both effective braking and a good quality drive system. Sometimes heavy loads had to climb the first length from the loading station without sufficient weight on the longer descent to balance the load. A higher capacity engine was required so that the cableway would also be used to generate electricity to power the sawmill. so a 50 PS electric motor was installed and performed well. [4: p19-20]

The hydraulic regulator consisted mainly of a capsule structure with a relieved throttle valve, which was driven by a belt from the cableway countershaft. The mechanism drew water from a reservoir and pushed it back into the box through slots of the regulating slide. The regulating slide is fully actuated by a centrifugal force governor, which may also be driven by the drive shaft of the track via a belt. As soon as the revolutions per minute of the countershaft begin to increase, the centrifugal force governor moves the regulating slide into action. [4: p20]

A General view of the Neu-Hornow sawmill. On the left in the foreground is the first support pillar of the cableway; in the middle is the loading station with the building for the drive and brake regulator. To the right of that is the sawmill with a grey timber drying shed. [4: p20]

The frame saw is driven by a Lanz Lokomobile/traction engine. (A Lokomobile was a portable, self-propelled, or towable steam-powered (or sometimes internal combustion) engine used historically to provide power to machinery like threshing machines or sawmills. Mounted on wheels or skids, these versatile, mobile power units were commonly used in agriculture and industry, frequently featuring a steam boiler and a single-cylinder engine.) [5]

Given the extremely difficult road conditions on the mountain, the firm Wilkins & Wiese undoubtedly acted uneconomically in choosing a locomobile as the drive system, because it was foreseeable that it would cause enormous difficulties to transport this large and heavy 10 hp machine up the mountain, and that the profit from saving on assembly costs compared to a stationary engine and boiler system to be transported disassembled would be far outweighed by the extraordinarily high transport costs of the fully assembled locomobile. [4: p21]

The locomobile/traction engine had to be transported 60 km from Mombo, the then terminus of the Usambara Railway, via Wilhelmsthal to Neu-Hornow. Two to three Europeans and 100 labourers worked continuously on the transportation of the traction engine for about seven months. The boiler was mounted on a railway wagon frame, which was then moved forward on a track. The track was then removed behind the wagon and reattached at the front. Depending on the difficulty of the route, distances of 100 to 1000 metres were covered daily.

This image shows the lengths that Wilkins & Weise had to go to, in order to get the locomobile/traction engine into position at Neu Hornow. [4: p21]

This required building roads and bridges, widening and reinforcing existing paths, and blasting rocks. Often the machine hovered over the abyss, in danger of plunging down and destroying months of expensive work. Using animals for transport was impossible because the tsetse fly was native to this area. Since the traction engine also had to provide the power for the initial commissioning of the cableway, the transport of the locomobile to Neu Hornow was a major contributor to delays in commissioning the cableway. [4: p21]

A glimpse into the loading station of the cableway at Neu Hornow. The station, except for the roof, was constructed entirely of iron to protect it from termites. The coupling points for Bleichert’s automatic clamping device, the “Automat,” are clearly visible at the front. This device was operated by the weight of the sling and the load. At the coupling points, the weights were supported during entry and exit by laterally arranged auxiliary rails, on which they ran with small rollers. Depending on whether the coupling rails rise or fall, the weight of the vehicles was raised or lowered, thus opening or closing the clamp. The haul rope was guided in such a way that it was gripped by the smooth coupling mechanism with a sling swivel. The process of coupling and uncoupling is therefore relatively simple: incoming cars require no operation at all, outgoing cars were pushed out of the station by hand and coupled themselves automatically to the haul rope. The coupling and uncoupling process was absolutely safe, and lifting the car’s weight by the auxiliary rails posed no risk of derailment, as the coupling rollers were only lifted by a very small amount and were also guided laterally, while the running gear did not lift off the track. [4: p22]

From the loading station, the track gradually ascended to its highest point. The log wagons, illustrated in the image below, which transported logs up to 14 m long and weighing up to 1000 kg, consisted of two carriages connected by the haulage rope. To increase the clamping force of the lead-weighted coupling mechanism on these steep inclines, stops were provided to the right and left of the suspension of the carriage, against which the suspension bracket was applied to inclines. Under the influence of a load, it acts like a lever on the pull piece of the clamp, thereby achieving a correspondingly increased clamping force, which ceases immediately when the incline decreases, so that the positive characteristics of the automatic coupling device reappear. Among these, the great protection afforded to the haul rope is particularly noteworthy, as the clamping force was not greater than absolutely necessary. The slings were designed with a lightweight construction, yet possessed the required stability during idling and when entering stations due to the use of a counterweight. [4: p22-23]

Timber-wagons on the cableway. [4: p23]

Platform wagons were used for transporting sawn timber down the valley and for transporting various goods up the mountain. These wagons were also used for passenger transport. [4: p23]

The highest point of the line was at 2011 m above sea level, 1591.2 m above the survey base, was reached 1.2 km from Neu-Hornow, 1523 m above the unloading station. To obtain the most favorable line alignment, a simple cut had to be made at the crossing over the edge of the plateau (shown in the image below). This presented no difficulties due to the firm clay layer, but in light of the heavy tropical downpours, special safety measures for the support foundations were required. For this purpose, the line was laid at an angle and equipped with a lateral drainage ditch. Sloping ditches were also dug in front of the supports to divert the water.  The slope of the ground followed the profile of the cableway and accordingly had a gradient of 1:2. [4: p24]

The summit of the line seen approaching from the loading station at Neu Hornow. [4: p24]

Ahead down the line was a gorge-like valley which the cableway panned on its way to the first ‘angle-station’, making use of a 30 metre high support stanchion.

The first angle-station (winkelstation) son the side of a promontory of rock above the gorge mentioned in the last paragraph. In order to accommodate the ‘winkelstation’, excavation was necessary at the top of the promontory. This cause difficulties as the ground proved friable and the easily crumbling and weathering rock fractured in two directions. Repeated collapses significantly delayed the completion of the cableway. Stability was finally achieved by building a significant retaining wall and by concreting the rock fissures. [4: p24-25]

This photograph was taken during construction of the first ‘winkelstation’. the cableway flanked the side of the promontory. ‘The ‘winkelstation’ is under construction to the right of this image. [4: p25]

The ‘winkelstation’ sat immediately above/behind the retaining wall and required some excavation of the rock to create a plateau. A sketch diagram appears below:

Winkelststaion No. 1. The cables of the first length line were tensioned by weights: the fully loaded cable, with 1 ton, the slack cable with 13 tons. The tension weights consisted of iron frames filled with concrete cubes. The cables of second length of the cableway were anchored in the winkelstation. The haul rope of the first line passed over the second line and was guided by deflection and guide rollers in the station. The station does not operate automatically. Automatic operation was omitted to keep costs as low as possible. Accordingly, each rope section had a coupling and attachment cleat, these operated in the same manner as at the loading station. A photograph pf this winkelstation appears below. [4: p26]
A photograph of Winkelstation No. 1. [4: p27]

Beyond Winkelstation No. 1, the railway crosses a short rocky ridge, then continues supported in the middle of two spans of approximately 300 metres each, across several hundred metres of steep gullies to the Willkelstation No. 2, located on the side of another rocky ridge and accessible only via difficult paths. Given the exceptionally unfavorable terrain, the central support between winkelstations I and II had to be maintained at a height of approximately 33 metres.

The section between the two Winkelstations is of particular interest because at the time of construction it was the steepest continuously operating cableway in the world. The location is shown in the image immediately below. Here, the gradient was 41° = 1 in 1.15 or 86 %). However, such inclined lifts with shuttle operation are not uncommon. The Bleichert company stated that this gradient was only surpassed by a few cable cars in the canton of Salzburg and the Wetterhorn lift near Grindelwald. The steepest gradient at the Wetterhorn lift, reaching up to 200%, corresponding to approximately 87°!

The steepest section of the cableway, shortly below Willkelstation No. 1 [4: p28]

Even funicular railways lag behind the Neu-Hornow cableway. The maximum gradient on a funicular railway is 70% on the Virgelbahn near Bolzano, which operates with a reciprocal carriage system.  The photograph below shows just how steep this section of the cableway is.

Over the longer spans, the haul rope had to be guided as far away as possible from the track rope to prevent entanglement in the track rope. Based on these considerations, a support design generally emerged that deviated from the normal design due to the large distance between the haul rope guide and the support shoe.

The steepest section of the cableway: 8 metre-long cedar beams are being transported down the gradient. Despite the gradient, the Bleichert coupling mechanism “Automat” holds the haul rope securely. Therefore, no safety or  multi-coupling devices are required. [4: p29]
The abnormal stanchion at the bottom of the steepest section of the cableway. [4: p30]

In Winkelstation No. 2 (shown diagrammatically below), the track cables of the second section are tensioned by counterweights. To gain the necessary space for the counterweights, a pit had to be blasted. The haul rope from Neu-Hornow terminates at this station. However, it is inextricably linked to the haul rope for the final section to Mkumbara, so that the speed of both ropes is the same.

Winkelstation No. 2. [4: p30]

Handbrake operation for the further descent did not seem reliable enough. It was much more practical to also apply the brake regulator installed in Neu-Hornow to the last section of the cableway. Furthermore, for this last section, with its relatively gentle gradient compared to the higher sections, there was a risk that the haul rope would stop if there was a large uphill load and a poorly occupied downhill section. Therefore, at Winkelstation No. 2, the traction cable of the upper two sections is guided around a pulley on the end guide shaft of the lower traction cable run, thus achieving the necessary positive connection. The traction cable of the upper section then passes over an end guide pulley mounted in a tensioning frame and is tensioned by tightening the tensioning lever due to weight distribution. In this station as well, the wagons are manually guided onto the following sections for the reason already mentioned. Winkelstation No. 2 is shown in the photograph immediately below. The coupling points are visible at the entry and exit points. To find space and support points for the installation, costly blasting and foundation work was also necessary here. [4: p28-29]

Winkelstation No. 2 [4: p31]
This photograph shows, dramatically, the length between Winkelstation No. 2 and Winkelstation No. 1 in the far distance. The longest span between support stanchions on the cableway was 900 metres which was the length closest to Winkelstation No. 2. It appears to good effect in this image! [4: p32]

As far as the terrain allowed, naturally existing support points were utilized. For example, just below Willkelstation No. 2, a support could be erected just before the drop into the Ngoha valley; however, beyond this point, no support was possible before the opposite valley edge, which was 210 m lower and 100 m away.

This photograph looks down the line of the cableway to Mkumbara in the valley bottom. In the foreground, the first and second 300-meter spans between suspension stations I and II are visible. On the left side of the image, the Winkel station II with its white roofs can be seen. From here, the large span across the Ngoha Valley begins, behind which the first tensioning and anchoring station for section III is located. The line then descends further to the plain, intersecting the banks of the hills in front of the Usambara massif twice more, between which the second suspension station, Pangalliebeno, is located on the northern slope. [4: p33]

The railway descends from the first suspension station shown in the above photograph at a gradient of 1 in 3. Cuts had to be made in the affected ridges, the first of which, at support No. 59 (shown below), was particularly troublesome. Supports had to be spaced 10 m apart, and an allowance had to be made for very unstable ground where the substrata was highly fissured and where rockfalls were frequent. Due to the continued disruption, more than 6000 cubic metres of rock had to be moved,

Planed timber planks being carried past the site of the landslide shortly before dropping down the cableway onto the plain. This image gives a good idea of the terrain that the cableway travelled over/through. [4: p34]
The supports/stanchions, as the pictures show, are largely identical. This gave the advantage for the cableway that the individual elements could be interchanged as required. [4: p34]

About 100 metres before the lower terminal station there was a double tensioning station where the suspension cables leading to the lower station were tensioned because the lower terminal station did not offer enough space for the weights.  It was necessary to create pits for the tensioning weights. [4:p31]

From the tensioning station the cableway crossed level ground to reach the terminus in Mkumbara. [4: p32]

The terminal station (shown in plan and section below) was angled, due to the direction of the connecting track to the Usambara railway. The ground below the station was piled shaped to create a loading ramp from which the logs could be easily rolled into the railway wagons on the metre-gauge siding. The unloading of the cableway was carried out in the same way as the loading, using a mobile ‘table’ that was moved under the arriving logs and raised by a simple winch. The sling chains were then released. The table was then tilted towards the ramp, whereupon the logs rolled off in the desired direction. Sawn timber was unloaded by hand. To prevent any delays in railway operations, a siding was provided alongside the main line. [4: p32-33]

The lower terminal station of the cableway at Mkumbara. [4: p35]

Perhaps of interest is the fact that permission to operate the railway telephone was granted only after great difficulties and subject to revocation, because telephone lines longer than 500 m, even if they ran entirely on the owner’s land, were within the protected area of the Tanganyika postal monopoly. [4: p33]

All the railway structures were made of iron to protect against termites, and the telephone poles were made of Mannesmann tubing. [6] This increased the construction costs. Freight costs for shipping and rail transport were within the normal limits appropriate to the size of the project. [4: p33]

In contrast, the costs of transporting the components to the construction site from the then-terminus of the Usambara Railway at Mombo, the execution of the foundation work, and the procurement of cement, water, etc., required considerable expenditure, especially since, neither the cattle-herding Maasai from the surrounding areas were available to work, nor could draft or pack animals be kept due to the tsetse fly. Roads also had to be built almost everywhere for transporting the materials, along which the supports, station components, and building materials were hauled individually by porters, during which many a sack of cement and many a barrel of water leaked quite by accident along the way, thus becoming lighter. The wages were relatively low, amounting to 45 heller or 60 pfennigs per day with free board including rice. The workers’ housing was also inexpensive to build.  They consisted of reed sheds or reed huts, which, at best, were covered with clay. [4: p33-34]

Taking into account all the factors that delayed and complicated construction, it is understandable, despite the low wages paid to the workers, that the total construction costs exceeded those of the actual delivery of mechanical parts for the cableway many times over, and it does not seem implausible that the construction as a whole cost between 1.75 and 2 million marks according to one account, and between 2 and 2.5 million marks according to another. [34]

Writing in the early 20th century, Hand Wettich said, “The question must now be raised whether these considerable costs for a private branch line of 9 km in length will also achieve the desired success, but it can be stated that the system is already well on its way to doing so. In 1909/10, as already mentioned, 1240 cubic metres of cedar wood were exported from Neu-Hornow, and exports are constantly increasing.” [4: p34]

The influence of the timber transport cableway on plantation farming.

The company (Wilkens & Worse), which, like so many others, was only brought into being by the construction of the Usambara Railway, developed in a direction that was hardly expected beforehand. …. For logging, the construction of roads, field railways, and houses, the company Wilkens & Wiese needed to keep a large number of native workers and hauling the timber required the keeping of cattle. As an alternative, earlier in the 20th century, two stallions and ten Norman mares were purchased in Marseille and transported to Africa. Despite the contaminated coastal areas, they arrived safely in the mountains at that time. Initially, some animals died, but the majority began to acclimatize, as evidenced by a number of foals. The success of the stud farm was limited, although the animals were at least protected from the tsetse fly on the heights of western Usambara and otherwise found favourable conditions. The number of horses was insufficient for the needs of the sawmill. Therefore, oxen, the humped cattle of the country, were also raised. [4: p35]

Keeping people and animals forced the plantation society to engage in agriculture on the Usambara plateau. On the protected clearings, maize, turnips, and oats were cultivated. Barley, in particular, yielded exceptionally good harvests, albeit in a small area. Barley was preferable to oats as it was  less susceptible to damage from the numerous buffalo and wild boar of the Schummewald forest, which caused considerable damage to the oat fields. Potatoes yielded up to 100 hundredweight per acre, but the potato harvests were threatened by severe night frosts, which at the time of writing of Wettich’s article had destroyed almost the entire year’s crop. [4: p35-36]

These developments, which came about almost accidentally, provided the possibility that both arable and livestock farming on the plateau could meet needs across German East Africa. In addition, non-food crops might be able to be cultivated – hemp, rubber, tannins, coffee, quinine , cotton, etc. – all these could be transported to the plains via the cableway. Neu Hornow and its cableway seemed to have a very bright future.

The scale that Wilkens & Wiese’s plantation business had reached was demonstrated by the company’s development, which began 13 years before with two Europeans and 100 native-born workers and by 1907 employed 10 to 12 Europeans and about 2,500 native-born workers. The company was started with a maximum capital of 500,000 marks, which gradually grew to 3,500,000 marks (as of 1907). [4: p36]

Just as the firm Wilkens & Wiese secured its business by expanding from timber harvesting, so too other timber companies established or acquired plantations for the same reason. For example, Elie Deutsche Holz-Gesellschaft für Ostafrika (Elie German Timber Company for East Africa) took over the rubber plantation of 3,000 Manihot Glaziovii trees [7] established directly at Sigi by the former Sigi Export Company. The three- to four-year-old trunks were tapped for the first time in 1910. Wettich was unable to comment on the quality of the rubber harvest because it was only at the time en route to Hamburg; however, the company was convinced of a good future for its plantation. [4: p36]

References

  1. M.F. Hill; Permanent Way Volume II: The Story of the Tanganyika Railways; East African Railways and Harbours, Nairobi, Kenya; Watson & Viney, Aylesbury & Slough, 1957.
  2. https://de.wikipedia.org/wiki/Adolf_Bleichert_%26_Co, accessed on 5th March 2025.
  3. https://web.archive.org/web/20081008211723/http://petervb.com/pdf/Clips_-_WireRopeNews.pdf, accessed on 5th March 2026.
  4. Hans Wettich; The development of Usambara under the influence of the East African Northern Railway and its private branch lines, with special consideration of the Mkumbara-Neu-Hornow cable car; Simion, Berlin 1911. Reprint from: Proceedings of the Association for the Promotion of Industry 90 (1911), Issue 6; via https://publikationen.ub.uni-frankfurt.de/frontdoor/index/index/docId/11924, accessed on 24th February 2026.
  5. https://commons.wikimedia.org/wiki/File:Lokomobile.jpg, accessed on 6th March 2026.
  6. Mannesmann tubing refers to high-quality, specialized steel tubes produced by Mannesmann Precision Tubes GmbH and Mannesmann Line Pipe GmbH, which are subsidiaries of the Salzgitter Group. The brand is known for pioneering the “Mannesmann process” for creating seamless steel tubes. Products are characterized by high dimensional accuracy, tight wall thickness tolerances, and minimal eccentricity. For more information see: https://en.wikipedia.org/wiki/Mannesmann and https://www.mannesmann-precision-tubes.com, accessed on 8th March 2026.
  7. Manihot Glaziovii is also known as Tree Cassava or Ceara Rubber Tree. It is a species of deciduous flowering plant in the spurge family, Euphorbiaceae, that is native to eastern Brazil. The tree cassava was used as a source of rubber, instead of Hevea brasiliensis throughout the world. The plant is introduced largely in the world, but now it is classified as one of the most highly invasive plants in the world. See: https://en.wikipedia.org/wiki/Manihot_carthaginensis_subsp._glaziovii, accessed on 8th March 2026.

Narrow-Gauge Industrial Lines in Tanganyika/Tanzania

The featured image for this article shows a train on the Kihuhui Bridge on the Sigi Railway in Tanganyika. [3]

Tanganyika (now part of Tanzania) possessed a dense network of industrial narrow-gauge railways, primarily developed during the German colonial era (German East Africa) in the late 19th and early 20th centuries to support plantation agriculture and forestry. While the main lines (Central Line and Usambara Railway) were built to 1,000 mm (metre) gauge, industrial, plantation, and forestry lines often used 600 mm (1 ft 11 5⁄8 in) or 750 mm (2 ft 5 1⁄2 in) gauge. [1]

Following World War I, the British administration deemed many of the 600 mm “light railways” to be economically inefficient compared to the, at the time, more efficient 1,000 mm metre-gauge lines, leading to a shift away from developing these smaller lines. [1]

Early Industrial Narrow Gauge lines included:

1. The Sigi Railway

The Sigi Railway (Sigi-Bahn) was a 23.5 km long, 750 mm gauge line opened in 1911 to transport timber from forests in the Usambara Mountains to the Tanga or Usambara Railway. It featured significant engineering challenges, including four switchbacks to handle steep terrain. [2]

Tanga and the Usambara Hills showing the metre-gauge line. [7: p2]
A closer view with a North point just to the West of North. The metre-gauge line runs diagonally across the map extract. The Sigi-bahn is represented by the dashed line leading to the centre of the hatched area. [7: p3]
A sketch map of the full length of the Sigi-bahn with Tengeni to the left and Sigi to the right. Beneath the German text in this extract, the gradient profile of the line is shown. [7: p15]
These three images divide the length of the Sigi-bahn shown above into three. [7: p15]
A train on the Kihuhui Bridge on the Sigi Railway in Tanganyika. [3]

The line ran from Tengeni station (185.9  m above sea level) near Muhesa (now Muheza), located on the Usambara railway, to Sigi (438  m above sea level) in the heavily forested eastern Usambara Mountains, which, until the railway’s construction, were rugged and difficult terrain. The area could only be reached by caravans via forest and mountain trails. This severely limited the transport of goods and merchandise. [3]

A travel group at Tengeni railway station in the Usambara Mountains, with trains of the Sigi Railway and the Usambara Railway in the background. [3]

Tengeni Railway Station. [10]

The Sigi Railway Management Concession of 29th April 1910, was a necessary precondition to the exploitation of the forestry concession held by the Deutsche-Holz-Gesellschaft fuer Ostafrika. The 23.5 km. railway’s operation was bound to the sawmills that provided the bulk of the traffic. [2]

Work began in 1904. The Sigi Export Company, during its ownership, constructed 17.6 km of track. Due to financial difficulties, work had to be interrupted repeatedly. After the German Timber Company for East Africa took over the logging rights to 12,000 hectares, the sawmill, and the railway, it completed the latter. The railway went into full operation on 1st September 1910. [3][4: p96]

The line had gradients of up to 40%, minimum radii of 40 m, and, as already noted, four switchbacks. It climbed 252 metres to the Sigi terminus. [4: p96] The switchbacks enabled a relatively uniform gradient to be achieved. The section of track near Sigi was particularly interesting. Here, the railway described almost complete circles to wind its way up the mountain slopes and featured three of its four switchbacks in a length of less than one kilometre. Where possible, engineering structures were avoided. Nevertheless, numerous smaller bridges had to be built over mountain streams, blasting operations carried out, and dams constructed, particularly along the upper section. The construction of a large steel girder bridge over the Kihuhui River gorge was unavoidable; this bridge had two masonry piers between its abutments. [3]

Between Tengeni and Fanussi, rails weighing only 10 kg/m and measuring 7 metres in length were used. The lower section of the line was ballasted only where absolutely necessary for structural or operational reasons, such as on the curves. In contrast, the upper section was fully ballasted and fitted with rails weighing 15 kg/m. Ballasting the entire line was planned but was not completed before the First World War. The rails rested on iron sleepers with shims. The sleeper design on the lower section was chosen to allow for the replacement of the existing rails with the new, heavier rail profile. The flat fishplates used in the first construction phase were reinforced by the addition of angle fishplates. [3]

The line originated at Tengeni station, 44km from Tanga on the Usambara Railway. The Tengeni station grounds belonged to the state and were leased to the German Colonial Railway Construction and Operating Company. With government approval, this company subleased the land for 30 years, first to the Sigi Export Company Ltd. and later to the German Timber Company for East Africa. The site included several buildings constructed by the Sigi Export Company, the railway’s administration building, a warehouse and goods shed, the engine shed, a carriage shed, and a water tower for supplying locomotives. There were also three houses for European employees, one of whom was a locomotive driver. The sawmill was located directly next to the station. The station building had a tiled roof. It contained a waiting room, an office for the local freight clerk, and a room rented by the Usambara Railway as a ticket office . A larger residential building with a corrugated iron roof served as overnight accommodation for travelers.

Apart from the two termini, there were no train stations or stops. Trains would stop on the open track if necessary. The entire route was equipped with a telephone line for train safety. [3]

In the German era, the operation of the Sigi-bahn was the responsibility of the German Colonial Railway Construction and Operating Company (DKEBBG), which had also been operating the Usambara Railway since 1905. The DKEBBG was a subsidiary of Lenz & Co. in Berlin, which built and operated numerous narrow-gauge railways of various track gauges throughout the German Empire. [5]

The choice of a narrower gauge compared to the Usambara Railway meant that all goods had to be transshipped at the connecting station of Tengeni for onward transport. [4: p96] In addition to freight traffic, there was also public passenger traffic, as roads were scarce in the area around Sigi and Amani. Thus, from the beginning, the terminus at Sigi also served the Imperial Biological-Agricultural Institute in Amani. [4: p96]

Given the route, the average travel speed was only about 10 km/h, so a train needed 2 hours and 20 minutes for the entire journey. Trains were usually mixed passenger/goods services , with connections in Tengeni to and from trains on the Usambara Railway. Pure freight trains were also used when needed. The scheduled trains also carried mail and parcels to and from Sigi. Due to their more pleasant climate compared to the lowlands and coastal region, the Usambara Mountains were a popular destination for Europeans living in the colony. Their abundance of game also made them a popular destination for hunters. Therefore, the DKEBBG even offered special trains for tour groups from Tanga to Tengeni with a connection to Sigi. The DKEBBG advertised the journeys on the Sigi Railway:

A special trip to Tengeni for the purpose of touring the Sigi Railway and visiting the Agricultural Institute in Amani would cost 100 Rp. On the Sigi Railway, scheduled trains should be used at the regular fares. (One way 3.50 Rp.). … A trip on the Sigi Railway, with its magnificent views of wild gorges, lush valleys, and dense primeval forests, is one of the most beautiful excursions in German East Africa.” [6]

In passenger transport there was a so-called “European class” (1st class, fare 15 Heller per tariff kilometre = 3.50 Rupees/total distance) and a so-called “native class” (3rd class, fare 2.5 Heller per tariff kilometer = 0.60 Rupees/total distance). With the full commissioning of the line, great expectations were placed on its role in opening up the entire East Usambara region. [7]

The up and downs of the Sigi railway after World War I were due to differing opinions with regard to the economics of the line. Operational experience in the 1920s ultimately served to convince the Government of Tanganyika that there could be no place in the Territory’s pattern of communications for the operations of light railways. [2]

Locomotives in Older Literature

According to the few available sources, the Sigi Railway had two wood-fired tank locomotives of different power levels at its disposal for operating the line . The existence of at least one six-coupled locomotive is confirmed by the photograph of a train on the Kihuhui Bridge which is included in this article. [3]

Details regarding the locomotives vary depending on the source.

The locomotives were reportedly built by Orenstein & Koppel in 1904. [5: p6][8: p26] According to another source, the locomotives were built in 1910. [9] The more powerful of the two locomotives had an output of 50 hp, weighed 10.7 ton, had a driving wheel diameter of 580 mm, and a wheelbase of only 1,400 mm. The less powerful locomotive was said to have produced only 40 hp, with otherwise similar dimensions. The 40 hp locomotive could only haul a load of 13–14 t on the inclines. This meant that it was not even capable of pulling the regular, relatively light trains on the line alone, without the assistance of the second locomotive. [3]

According to a third source, the locomotive fleet consisted of one 45 hp and one 60 hp triple-coupled tender locomotive. [3][7: p15]

Locomotives in More Recent Evidence

The more powerful of the two locomotives, however, was clearly a Mallet locomotive of the B’Bn4vt type. A photograph in the collection of the colonial image archive of the University of Frankfurt/Main confirms the error in older literature. The locomotive pictured, with its Kobel chimney and the tender box for timber “attached” to the cab, is unequivocally a Mallet locomotive and not a six-coupled locomotive with a rigid chassis. The photograph is captioned ‘Train of the Sigibahn’. [3]

Between 1902 and 1912, the Orenstein & Koppel company built a large number of Mallets for plantation and narrow-gauge railways worldwide. At least visually, the Sigibahn locomotive is very similar to locomotive number 13 of the Frankfurt Feldbahn Museum (Orenstein & Koppel , B’Bn4vt, 1909/3902, 30 hp, 600 mm gauge, ex. locomotive 4 of the Gending/Java sugar factory, 600 mm gauge) and to the locomotive of the Statfold Barn Railway /England (Orenstein & Koppel, B’Bn4vt, 1905/1473, 60 hp, 762 mm gauge, ex. locomotive 5 of the Pakis Baru/Java sugar factory, 750 mm gauge). [3]

A Mallet locomotive was better suited to the railway’s needs than a six-coupled engine, as it could be fitted with a significantly larger boiler and four cylinders instead of two for increased power. The articulated design of the running gear resulted in good running characteristics, even on the tight curves of the track. [3]

Questions regarding the Sigibahn locomotives remain unanswered. However, it is established that the locomotives bore the operating numbers No. 1 and No. 2. The original bronze number plate of locomotive No. 2, bearing the inscription “Deutsche Holzgesellschaft für Ostafrika No. 2” (German Timber Company for East Africa No. 2) , has been preserved, is part of the collection of the German Historical Museum in Berlin, and is displayed there in the permanent exhibition. [3]

The locomotive plaque for Sigi-Bahn No. 2! © Kleinbahnen, and licenced for reuse under a Creative Commons licence (CC BY-SA,4.0). [3]

Passenger Coaches and Goods Wagons

The rolling stock was modest. It consisted of only two-axle, semi-open passenger cars [5: p6-7] , two covered and two open two-axle goods wagons, six four-axle stake wagons for timber transport, two firewood wagons, and three track maintenance cars. When there was a large influx of passengers, they were occasionally transported in the two-axle open and covered freight cars. [3]

Despite the steep gradients, all the Sigi-bahn carriages were equipped only with hand brakes. This necessitated the presence of brakemen on the trains . The manufacturers of the carriages and wagons are not known. [3]

And Finally ….

In searching for more information about the Sigi-bahn I have discovered a forum link to a walk along the length of the Sigi-bahn which can be found here. [11]

.

2. Sisal Plantation Railways

Numerous privately owned, 600 mm gauge light railways operated throughout the coastal and Tanga regions, linking sisal estates to factories and main-line stations.

Sisal plantation railways were a crucial component of the German colonial agricultural economy in the late 19th and early 20th centuries. These networks, often referred to as ‘light railways’, were used to transport harvested sisal leaves from the fields to processing factories and eventually to the Tanga port for export.

Sisal cultivation and associated small-gauge rail systems were introduced in the 1890s, with a significant boost in the very early 20th century. The plantation lines typically branched off from the main Usambara Railway.

The Tanga sisal plantations contributed to over 60% of the country’s GDP at their peak. The history of these railways is deeply linked to the development of the Tanga port and the early industrialization of the region.

A sisal plantation in German East Africa in 1906, ©  Walther Dobbertin (CC BY-SA 3.0 de). [14]

An illustration of one of these lines in use in the 1960s can be seen here. [12]

Sisal production in the country peaked in 1964 with around 250,000 tonnes in production from regions from all over the country such as Tanga, Morogoro, Arusha, Mwanza and Shinyanga. In 1967 following the Arusha Declaration most of the sisal estates were nationalized by the government. This began the downfall of the sisal industry as bureaucracy, over-centralization and lack of experience caused the production to fall rapidly. Furthermore, with the increasing popularity of Synthetic Nylon fibers, drove the world price for sisal down resulting in the foreclosure of many sisal factories. By the end of Ujamaa and President Nyerere’s rule, sisal production had fallen from 235,000 tonnes in 1964 to 32,000 in 1985, less than 15% of the country’s peak.” [14]

Remnants of the 600mm-gauge lines appear in different places in the sisal fields of Tanzania. Two examples can be found on these links – here [13] and here. [15]

.

3. Later Industrial Lines

Later Industrial Narrow Gauge Lines included:

A. The Southern Province Railway

The Southern Province Railway was a short-lived, metre-gauge railway constructed by the Overseas Food Corporation to support the “Groundnut Scheme.” It ran for 212 km from the port of Mtwara to Nachingwea, and including various branch lines was a network of about 275 km of railway, all told. It was abandoned in 1963.

An extract from a photograph of a map of Tanzania’s railways taken by Kiplimo Koech. This photograph was shared by Kiplimo Koech on Facebook on 12th February 2026. [23]

When the British Government “decided to press on with the Groundnuts Scheme in the Southern Province, the railways were faced with the problem of transport to and from the areas of cultivation which were then only vaguely defined. As soon as more was known about the approximate location of the areas to be cultivated around Nachingwea, and the estimated tonnages to be moved between there and the coast, the railways undertook a study of the best way of meeting the requirements at the lowest transport cost. There were three possible sites for port works – Lindi creek, Mikindani and the Mtwara creek which forms part of the port of Mikindani. Reasonably reliable information was available about the route along the Lukuledi valley between the areas to be cultivated and Lindi creek, but very little was known of the topography and soil conditions over about so miles at the coastal end of the most likely route to Mikindani and Mtwara. No sort of survey was practicable at the time, as all routes were impassable due to the rains. Such information as was available indicated that this section would not be easy to build. The limitations of Lindi as a port were fully recognised, but the Administration of the Railways recommended the use of Lindi in the early years of the scheme, leaving the route to Mikindani and Mtwara and the possible sites for port works to be surveyed later when more precise information would be available regarding costs and tonnages to be carried. However, the Ministry of Food decided to send out im-mediately its own mission to consider the site of the port works. The mission, which arrived in the February of 1947, consisted solely of specialists in port operation and construction. They recommended that the port works be in Mtwara creek on the grounds that it offered the best harbour facilities. Little or no account was taken of the extra cost of railway construction, the possible difficulties of the route and the extra time which would be needed for construction. The Ministry of Food, which was solely responsible for financing the new port and railways, accepted the advice of the mission.” [22: p268-269]

The Railways’ Administration had meantime accepted responsibility for the survey and construction of the railway on behalf of the Managing Agents. In view of the importance attached to the Groundnuts Scheme, the Administration decided that the quickest way of providing a route of adequate capacity from the sea was to proceed from Mkwaya at the head of Lindi creek through Ruo – 20 kilometres from Mkwaya – to Nachingwea. The survey began in May, and the first contracts for earthworks were placed before the end of the year. The survey of the line from Mtwara to Ruo was entrusted to Messrs. Paulings, who were later awarded the contract for the construction. The Mkwaya-Ruo-Nachingwea route was opened to traffic on 25th October 1949. Due to physical difficulties, work on the Mtwara-Ruo section proceeded slowly, and due to financial difficulties Messrs. Paulings were released from the contract in the September of 1950. The Mtwara-Ruo section was finally opened on 17th January 1954.” [22: p269]

In addition to the needs of the Groundnuts Scheme, the railways were also concerned with the provision of transport for minerals. In 1947, a siding, about nine miles long, was built from the Mwanza branch to serve the Mwadui mine of Williamson’s Diamonds Limited. The Company met the cost of bridging and earthworks. … In the Mpanda district of the Western Province, a large lead-silver-copper-gold ore body was found, and Uruwira Minerals Ltd. undertook the development of a lead mine. It was decided, in 1946, to investigate the possibility of a branch line from Kaliua on the Central line to Mpanda. After a preliminary survey it was decided to build the line, about 131 miles long. By the end of 1947, 70 miles of the location survey had been completed, and contracts for the earthworks of the first 40 miles had been placed with three local contractors. This line was opened to traffic in the August of 1950.” [22: p269]

In 1947, the goods stock on the Central and Tanga lines was in a normal state of repair and overhauls were up to schedule. Like all other railways, the Tanganyika Railways suffered at the time from a shortage of tyres, springs, couplers and brake spares, but the percentage of wagons withdrawn from service due to this shortage was not abnormal. As soon as a preliminary estimate of the tonnage to be moved in connection with the Groundnuts Scheme was available, it was clear that the factor restricting the Railways’ capacity would be the supply of wagons. Orders were placed immediately in the United Kingdom but, in spite of the priorities obtained, deliveries could not be expected for two to three years. A world-wide search was made for metre-gauge stock to meet the demand during the intervening period – and to some extent as part of the permanent equipment. As a result wagons belonging to the War Office were found at Shaiba, near Basra, and at El Shatt, near Suez. A technical officer was sent to select suitable wagons, and those chosen were shipped as soon as possible. Because of exposure to the weather and pilfering they needed a fair amount of rehabilitation. By concentrating the resources of the Dar es Salaam workshops on the shipments as they arrived, the wagons were quickly put into service. In all, 430 four-wheeled wagons were obtained from Shaiba and El Shatt towards the end of 1947 and early in 1948. Later the purchase of these wagons was criticised as an unnecessarily expensive and unsatisfactory way of meeting a very urgent need. An effective answer is that 414 of the wagons were still giving good service in 1957.” [22: p269-270]

Moreover, these wagons arrived in Dar es Salaam at a most opportune moment. They enabled the ordinary traffic offered, the extraordinary traffic of the Groundnuts Scheme and construction materials to be carried during the very difficult time until the new wagons on order arrived in 1949 and 1950. The strain imposed on the capacity of the port and the Central line was exacerbated by the fact that the import of materials and equipment for the Groundnuts Scheme far exceeded the original estimates.” [22: p270]

As it turned out, the Southern Province Railway was an overambitious scheme. Construction took place between 1948 and 1954 to facilitate the export of groundnuts under the British Overseas Food Corporation’s ambitious post-World War II agricultural initiative. The scheme was intended to produce up to 600,000 tons of peanuts annually on cleared bushland to address food shortages and generate revenue, the project exemplified colonial-era top-down development but collapsed amid unsuitable clay-heavy soils, erratic rainfall, mechanical breakdowns, and overestimation of yields, resulting in total losses of £36 million by 1951. Despite the scheme’s abandonment, the single-track railway—built with 60-lb rails on untreated wooden sleepers—was completed and opened for limited traffic in January 1954, initially subsidized through joint guarantees by the Overseas Food Corporation and the Tanganyika government against predictable operating deficits. Its brief operation underscored the perils of ignoring local ecological realities in favour of imported heavy machinery and rapid mechanized clearing, leaving behind underutilized infrastructure that highlighted systemic flaws in mid-20th century imperial planning rather than delivering sustained economic benefits. [16]

The failure of the Groundnut Scheme should probably have resulted in a decision to abandon the construction of the railway, but the thinking at the time was that the line would promote significant growth in southern Tanganyika. After completion of the line and its branches it was placed under the same management as all the other metre-gauge lines in East Africa, with the East African Railways (EAR). [16]

Rather than being designed with longevity in mind. The network, focused on freight haulage with minimal passenger elements. It spanned key agricultural zones but was engineered for temporary use (for instance, using untreated timber for sleepers rather than steel), reflecting the scheme’s optimistic projections for short-term financial gain, rather than long-term viability! [16]

Initially the railway was steam-powered (series RV/21and NZ/22) but under EAR diesel-power was introduced (series 80 and 81).

East African Railways publicity photograph of No. 2217, circa. 1953. In the late 1940s, two of this 4-8-0 Class (TR NZ Class/EAR Class 22) were transferred to the Southern Province Railway, © Public Domain. [20]
East African Railways publicity photograph of TR No. 252, circa.  1953. These were a 4-8-2 development of the 2-8-2 TR MK class. The eight members of the RV class were built by Vulcan Foundry, in Newton-le-Willows, Lancashire. [21]

Two American outline diesel locomotives at work on the Southern Province Railway in the mid-20th century. [19]

The Route of the Southern Province Railway

This schematic route plan is provided by Wikipedia. It shows the main line from Mtwara Harbour to Nachingwea and notes two link lines to the Sisal Plantations at Muta-Narunyu and Karimjee. Two branch lines, one to Masasi, the other to Lindi Creek are also shown. The second of which also provided a link to Lindi Sisal Plantation. [17]

Two bridges over the River Lukuledi are also shown.

Mtwara Port was deepened by the British in 1948-1954. It was functional but underutilized for many years due to poor transport infrastructure. However, in 2010-2011 the increased activity in oil and natural gas exploration caused a surge in activity. It has recently seen major upgrades. The port has a special economic zone attached to it and In December 2015 Alistair Freeports Limited injected $700,000 to upgrade the Export processing zone around the port area. [18]

Mtwara City, Port and Gas Plant. [Google Maps, February 2026]

As we noted earlier, the Port at Mtwara was built towards the end of the construction period as an extension to the original length of line which ran to a port at Lindi. The extension to Mtwara came off the line to Lindi about 27 km from Lindi and ran for about 106 km serving Mikindani and Mtwara.

There is little that I have been able to find online which can be used to confirm the route of the railway. Some assumptions on the alignment of the railway can be made, whether they are warranted or not I cannot tell.

Should further information come to light the remaining paragraphs and images about this line will need to be revised. I can see two possibilities for the route of the old line:

1. The first possibility depends on an assumption that the railway will have been built with an access road alongside it. That access road is likely to have survived and become part of the lasting road network of the Southern Province of Tanzania.

Determining, conclusively, the route of the line would have been helped by the availability of Google Streetview images in the Southern Province of Tanzania. However Google Streetview imagery in Tanzania is primarily focused on key tourist locations and specific, curated, or partnered locations rather than comprehensive nationwide road mapping. Major coverage areas include Gombe National Park, parts of Zanzibar, and the Ngorongoro Crater rim.

The next three images show a possible route of the line between Mtwara and Nachingwea the grey lines are roads T6 running West from Mtwara, the T7 in the East serving Lindi, the T6 from Mingoyo to Nangana and the Nachingwea Road, West of Nangana.

In the West there was a branch line serving Lukuledi and Masasi, the route of which is much less clear.

These three map extracts come from Kartaview. They show the full length of the presumed route of the line from Mtwara to Nachingwea. The branch line to Lindi is also shown. The branch line to Masasi ran through Lukuledi, but its likely route is less easy to establish. [24]

My presumptions about the possible route of the line are called into question by the schematic route plan above which places the junction between the line serving Lindi at a place called Ruo. We will come back to this below.

In the East, close to Mtwara Port the alignment of the railway is very difficult to determine as it has probably been built over.

It is probable that the line from Mtwara Port followed the line of the T6 Northwest along the coast before turning inland adjacent to or on the line of the T6 which runs in a straight line South-southwest as far as its junction with Tanu Avenue. Here the road runs through reverse curves before continuing South-southwest. [24]
As we have already noted, a most likely alignment for the old railway follows the T6 as it travels West, through Mikindani and on towards a  junction at Mingoyo. [24]

Assuming that my assumptions are correct, at Mingoyo the line from Mtwara appears to have met the line to/from Lindi. It appears that trains to Lindi would have progressed without reversing. Trains to Nachingwea would have required the locomotive to run-round its train.

Assuming that the alignment of the T6 and T7 roads roughly follow the line of the old railway, it can be surmised that each train heading West would have required the locomotive to run-round the train. [24]
The coastal town and port of Lindi was the original eastern terminus of the Southern Province Railway which approached the town from the South along the route of the present T7 road shown grey on the map extract. [24]
A similar area as it appears on Google Maps satellite imagery. The line would have approached Lindi from the South on or alongside the T7 road which Google Maps annotates B2, then probably crossing Lindi Town Small Bridge and entering the Port from the South. [Google Maps, March 2026]
Lindi Town Small Bridge seen from above. [Google Maps, March 2026]
Lindi Town Small Bridge in the 21st century, © Khalid Sakewa and licenced for reuse under a Creative Commons licence (CC BY-SA 4.0). [25]

The remainder of the route West to Nachingwea would follow the line suggested, along the T6 and the Nachingwea Road.

2. Given the complications associated with a junction at Mingoyo and the fact that the Wikipedia schematic route diagram indicates that the junction between the Lindi line and the Mtwara line was 14 km further South at Ruo, it is possible that the line from Lindi followed the T7 to Mingoyo and the the T6 to Mkwaya. However, following this route takes the line in the wrong direction to access a junction at Ruo which is some distance away to the Southwest. No bridge is shown over the Lukuledi river on the Lindi branch on the schematic route plan above. However, neither is a bridge shown on the line to/from Mtwara.

If, Ruo was the location of the junction at least one bridge over the Lukuledi would have been necessary, most probably on the line from Lindi.

As is painfully obvious from the paragraphs above, without significantly more information we can only speculate on the actual route of the old line.

.

B. Port of Bujumbura Railway: While operating within the neighbouring territory of Burundi, a 600 mm gauge industrial railway ran from 1947 to 1982 within the port area, servicing Lake Tanganyika traffic. Its operator was Office Congolais des Chemins des fer des Grands Lacs (CFL). [26] At the time of its construction, Bujumbura was known as Usumbura. [27]

.

References

  1. https://en.wikipedia.org/wiki/History_of_rail_transport_in_Tanzania, accessed on 24th February 2026.
  2. https://www.africabib.org/rec.php?RID=187591229, accessed on 24th February 2026.
  3. https://de.wikipedia.org/wiki/Sigibahn, accessed on 24th February 2026.
  4. Franz Baltzer; The Colonial Railways with Special Consideration of Africa; Berlin 1916. Reprint, Leipzig 2008.
  5. Helmut Schroeter; The Railways of the Former German Protectorates of Africa and Their Vehicles = The Vehicles of the German Railways, Volume 7; Frankfurt 1961.
  6. German Colonial Railway Construction and Operating Company (ed.): German East Africa – from the coast to Kilimanjaro with the Usambara Railway; Heinicke, Berlin, 1914.
  7. Hans Wettich; The development of Usambara under the influence of the East African Northern Railway and its private branch lines, with special consideration of the Mkumbara-Neu-Hornow cable car; Simion, Berlin 1911. Reprint from: Proceedings of the Association for the Promotion of Industry 90 (1911), Issue 6; via https://publikationen.ub.uni-frankfurt.de/frontdoor/index/index/docId/11924, accessed on 24th February 2026.
  8. Helmut Schroeter and Roel Ramaer; The railways in the former German protectorates then and now; Krefeld, 1993.
  9. Roel Ramaer; Gari la Moshi – Steam Locomotives of the East African Railways; Malmo 2009.
  10. https://postimg.cc/dDHY1F8g, accessed on 24th February 2026.
  11. https://bimmelbahn-forum.de/forum/index.php?thread/22289-die-sigibahn-auf-den-spuren-einer-privaten-schmalspurbahn-in-ostafrika, accessed on 24th February 2026.
  12. https://www.gettyimages.com.au/detail/news-photo/sisal-leaves-are-taken-by-rail-for-processing-on-a-news-photo/88556860, accessed on 24th February 2026.
  13. https://www.alamy.com/stock-photo-tanzania-tanga-usambara-mountains-sisal-farming-and-industry-dd-ruhinda-103739156.html, accessed on 24th February 2026.
  14. https://en.wikipedia.org/wiki/Sisal_production_in_Tanzania, accessed on 24th February 2026.
  15. https://www.alamy.com/stock-photo-tanzania-tanga-usambara-mountains-sisal-farming-and-industry-dd-ruhinda-103738452.html, accessed on 24th February 2026.
  16. https://grokipedia.com/page/southern_province_railway, accessed on 24th February 2026.
  17. https://en.wikipedia.org/wiki/Southern_Province_Railway, accessed on 24th February 2026.
  18. https://en.wikipedia.org/wiki/Mtwara_Port, accessed on 25th February 2026.
  19. https://unitedrepublicoftanzania.com/economy-of-tanzania/infrastructure-in-tanzania/railway-in-tanzania/mtwaras-forgotten-rails-a-glimpse-into-tanzanias-industrial-past, accessed on 25th February 2026.
  20. https://en.wikipedia.org/wiki/TR_NZ_class, accessed on 25th February 2026.
  21. https://en.wikipedia.org/wiki/TR_RV_class, accessed on 25th February 2026.
  22. M.F. Hill; Permanent Way – Volume II – The Story of Tanganyika Railways; East African Railways and Habours, Nairobi, Kenya; Watson & Viney, Aylesbury & Slough, 1957, p268-272.
  23. https://www.facebook.com/share/p/1DyhZvKm8J, accessed on 26th February 2026.
  24. https://kartaview.orghttps://kartaview.org, accessed on 4th March 2026.
  25. https://upload.wikimedia.org/wikipedia/commons/thumb/1/11/Little_Lindi_Town_Bridge.jpg/500px-Little_Lindi_Town_Bridge.jpg, accessed on 4th March 2026.
  26. https://en.wikipedia.org/wiki/History_of_rail_transport_in_Burundi, accessed on 4th March 2026.
  27. https://www.sinfin.net/railways/world/burundi.html, accessed on 4th March 2026.

Mallet Locomotives in East Africa

Ethiopia/Eritrea

The 950mm-gauge line from Massawa on the coast, inland to Agordot, was built during colonial occupation by the Italians with some steep gradients which meant that Mallets were considered to be suitable motive power.

The line should not be confused with the metre-gauge line running from Djibouti to Addis Ababa. A metre-gauge railway that was originally built by the French from 1894 to 1917 which has since been replaced by a Chinese built standard-gauge line. [5]

In 1907, Maffei built three 0-4-4-0T locomotives for the Massawa to Agerdot line.

Ansaldo the “supplied twenty five further engines of the same class between 1911 and 1915, and in 1931 and 1939 Asmara shops assembled a nominal three new engines from d components of earlier withdrawn engines. All these were standard European narrow-gauge Mallet tanks, saturated, slide-valved and with inside frames.” [1: p64]

In the mid-1930s, a series of fifteen larger 0-4-4-0T locomotives were built. These were “built to a superheated, simple expansion design, of which ten had piston valves and Walschearts gear and the other five, Caprotti poppet valves driven from outside cardan shafts.” [1: p65] A later series of “eight engines built by Analdo in 1938 reverted to compound expansion, retaining the superheater and piston valve features.” [1: p65]

The last of the Eritrean Mallets was built in their own shops in 1963, making it the last Mallet built in the world. [6]

The line closed in 1975. Eritrea was occupied by Ethiopia for many years. After gaining independence in 1993, some of the former railway staff started to rebuild their totally destroyed railway. Some of the Mallets, built by Ansaldo (Italy) in 1938, were brought back to life. Also one of the small Breda built shunters, two diesel locos and two diesel railcars (one from 1935) were put back into working order. [7]

A section of the line, between Massawa, on the coast, and Asmara, was reopened in 2003 and has offered an opportunity for Mallet locomotives to be seen in operation in East Africa. Indeed, an internet search using Google brings to light a list of videos of locomotives heading tourist trains in the Eritrean landscape.

Wikipedia notes that the line has a track-gauge of 950mm and that locomotives operate over a 118 km section of the old line. Italian law from 1879 officially determined track gauges, specifying the use of 1,500 mm (4 ft 11 1⁄16 in) and 1,000 mm (3 ft 3 3⁄8 in) gauge track measured from the centre of the rails, or 1,445 mm (4 ft 8 7⁄8 in) and 950 mm (3 ft 1 3⁄8 in), respectively, on the inside faces. [4]

Between Arbaroba and Asmara in November 2008, a single coach is headed by one of the surviving Mallet locomotives. This is an extract from an image on Wikimedia Commons (public domain). [13]

Steam operation on the line is over, no regular services are provided but occasional tours still take place with plenty of caveats about the availability of any form of propulsion. An example is a German-speaking tour planned (as of 24th March 2024) for November 2024. [8]

Tanzania (Tanganyika)

The metre-gauge line inland from Dar-es-Salaam was built by the Ost Afrika Eisenbahn Gesellschaft (East African Railway Co.). A.E. Durrant tells us that its first main line power “was a class of typical German lokalbahn 0-4-4-0T Mallets, built by Henschel in 1905-7. These were supplemented in 1908 by four larger 2-4-4-0Ts from the same builder, after which the railway turned to straight eight-coupled tank and tender engines.” [1: p67]

R. Ramaer notes that the first locomotives used by the Usambara Eissenbahn (UE) on the Tanga Line were five 0-4-2 locos which arrived on the line in 1893. Rising traffic loads led the UE “To look for something more substantial and in 1900, Jung supplied five compound Mallet 0-4-4-0T’s as numbers 1-5, later renumbered 6-10. … To provide enough space for the firebox and ashpan, the rigid high-pressure part, comprising the third and fourth axles, had outside frames, whereas the low-pressure part had inside frames.” [9: p19]

UE engine No. 1 (0-4-4-0T – supplied by Jung) with an early passenger train ready for departure at Tanga station in 1890. This image was posted on the Urithi Tanga Museum Facebook Page [10] and is also reproduced in R. Ramaer’s book. [9: p19]
UE Mallet 0-4-4-0T No. 8 heading a passenger train at Mombo. This image was shared in error on the Old Asmara Eritrea Facebook Page.  [11] It also appears in R. Ramaer’s book. [9: p20]

On the Central Line (Ost Afrikanische Eisenbahn Gesellschaft – or OAEG) which ran inland from Dar-es-Salaam, construction work started in 1905 and the first locomotives used by the OAEG were four 0-4-0T engines built by Henschel, a further four of these locomotives were supplied in 1909. These small engines had a surprisingly long life. Mallets were first supplied in 1905 by Henschel and were suitable for both coal and oil firing. These were 0-4-4-0T locos (four supplied in 1905 and one supplied in 1907). “The problem with this type of engine was the restricted tractive effort and running was not satisfactory because of the lack of a leading pony truck. … Therefore Henschel supplied a second batch of four locomotives in 1908 as 2-4-4-0Ts with larger boilers and cylinders. They also had a higher working pressure of 14 atmospheres (200lb/sq in) in comparison to 12 atmospheres (170lb/sq in)  for the earlier engines, while the bunker capacity had been increased from 1.2 to 2.2 tonnes of coal. (Oil fuel had been discarded).” [9: p21-23]

OAEG 2-4-4-0T No. 27, in the last Mallet class to be built for German East Africa. This locomotive appears in the Wikipedia list of Henschel steam locomotives. [12] It also appears in R. Ramaer’s book [9: p23] and A.E. Durrant’s book. [1: p66]

Kenya-Uganda

An ‘MT’ class locomotive in ex-Works condition at the Queen’s Park works of the North British Locomotive Co. [1: p66]

Mallets were the first articulated locomotives to operate in East Africa. Mallets were introduced on the Uganda Railway in 1913. A.E. Durrant notes that they consisted of “a batch of eighteen 0-6-6-0 compound Mallets to what was the North British Locomotive Co’s standard metre-gauge design, as supplied also to India, Burma, and Spain. They had wide Belpaire fireboxes, inside frames and piston valves for the high pressure cylinders only. Built at Queens Park works in 1912-1913, these locomotives entered service in 1913-14 and remained at work until 1929-30, when they were replaced by the EC2 and EC2 Garratts.” [1: p66]

North British Class ‘MT’ Mallets arrived in Kenya just before the start of WW1. [2]

These locomotives were given the classification ‘MT’ within the Uganda Railway fleet. Disappointing performance and high maintenance costs led to them being relegated to secondary duties and eventually being scrapped in the late 1920s as the Beyer Garratt locomotives began to arrive. [2] Their presence on the system was heralded by, “Railway Wonders of the World,” with the picture shown below. [3]

An ‘MT’ class Uganda Railway locomotive as illustrated in ‘Railway Wonders of the World’. [3]

References

  1. A.E. Durrant; The Mallet Locomotive; David & Charles, Newton Abbot, Devon, 1974.
  2. Kevin Patience; Steam in East Africa; Heinemann Educational Books (E.A.) Ltd., Nairobi, 1976.
  3. http://www.railwaywondersoftheworld.com/uganda_railway2.html, accessed on 1st June 2018.
  4. https://en.m.wikipedia.org/wiki/Eritrean_Railway, accessed on 22nd March 2024.
  5. https://en.m.wikipedia.org/wiki/Addis_Ababa%E2%80%93Djibouti_Railway, accessed on 22nd March 2024.
  6. https://en.m.wikipedia.org/wiki/0-4-4-0, accessed on 22nd March 2024.
  7. https://www.farrail.net/pages/touren-engl/eritrea-mallets-asmara-2010.php, accessed on 24th March 2024.
  8. https://ecc–studienreisen-de.translate.goog/historische-eisenbahn-und-strassenbahnreisen-mit-peter-1/8-tage-eritrea-mallets-in-den-bergen-afrikas?_x_tr_sl=de&_x_tr_tl=en&_x_tr_hl=en&_x_tr_pto=sc, accessed on 24th March 2024.
  9. R. Ramaer; Steam Locomotives of the East African Railways; David & Charles, Newton Abbot, 1974.
  10. https://www.facebook.com/urithitanga.museum/photos/pb.100063540805743.-2207520000/2336640756358366/?type=3, accessed on 24th March 2024.
  11. https://www.facebook.com/photo/?fbid=2169593963301193&set=pcb.2169594269967829, accessed on 24th March 2024.
  12. https://de.m.wikipedia.org/wiki/Liste_schmalspuriger_Lokomotiven_von_Henschel, accessed on 24th March 2024.
  13. https://commons.m.wikimedia.org/wiki/File:Eritrean_Railway_-_2008-11-04-edit1.jpg, accessed on 24th March 2024.