Part 17: North Coast 3: Empangeni to Golela by Les Pivnic, Bruno Martin and Ashley Peter ©

SOAR PART 17 The Natal North Coast Line 3, compiled by Les Pivnic, Bruno Martin and Ashley Peter. ©

Please note: All photographs, maps and text in Soul of a Railway are protected by copyright and may not be copied or reproduced in any way for further use without prior permission in writing from the compilers of Soul of A Railway: Les Pivnic, Charlie Lewis and Bruno Martin.

NATAL SYSTEM

The request for photos taken on the Natal North Coast line generated so many contributions that it necessitated splitting the chapter into three parts. This is the third and covers the line from just south of Empangeni – 118 miles from Durban, to Golela – 246 miles and the original terminus of the North Coast line. The line was extended into Swaziland in 1977 (thereby making a direct connection with the CFM). This chapter does not include the extension into Swaziland. Included is the branch line to Nkwalini and the Locomotive Depot at Empangeni.

For convenience, the following statements are repeated because they apply in equal measure to this chapter.

In terms of steam motive power over the years, this line was worked by a variety of classes – including Garratts and large main line engines like the class 15CB/CA in later years. Electrification and dieselisation eventually took over from steam as will be revealed in the photographs.

My co-compiler Bruno Martin provides the historical background to this interesting line as well as details of the terrain covered as it winds its way up the North Coast. Bruno has also provided all the maps and gradient profiles – thanks Bruno! Charles Strydom's account of the rebuilding of the Umfolozi bridge after it was almost completely washed away in 1984 is fascinating and will be appreciated by the civil engineers among our readers.

Special mention must be made of the mammoth contribution by Ashley Peter in not only providing a comprehensive set of photographs by the Late Brian Couzens but also providing authoritative and informative captions to go with Brian’s photos.

My colleague Charlie Lewis also needs special mention in that he had to tackle the massive amount of Photoshopping of Brian’s images that over time have deteriorated very badly. Thanks, Charlie, for your sterling work in this regard (YAY from Charlie).

The following photographers and colleagues (in no particular order) contributed photos to this chapter:

Peter Stow; Late Brian Couzens via Ashley Peter – RSSA Durban; Charlie Lewis; John Carter; A.A.Jorgensen; John Middleton; Dick Manton; Robert Kingsford-Smith (Rags); A.E.Durrant; Leith Paxton; Glen Mills; Les Pivnic.

Thanks as usual are also due to Andrew Deacon for formatting the chapter and Charlie Lewis for posting it.

MAPS & GRADIENT PROFILE

THE NATAL NORTH COAST LINE:

CONSTRUCTION AND DEVELOPMENT OF THE

NINIANS - GOLELA SECTION

Compiled by Bruno Martin

In part 16, we followed the construction and the development of the North Coast Line from Tugela to Ninians. The extension from Tugela to Empangeni and Somkele was built by the Zululand Railway Company.

On 18 July 1902, the 44¾ mile section from Lower Tugela to Umhlatuzi (changed to Mhlatuze in 1953) was opened to traffic and a further 8⅛ miles was taken into operation on 16 January 1903 which brought the railhead to Empangeni. At the end of 1902, the formation was completed to Somkele and the rails laid up a point that was 76 miles from Lower Tugela. The line was opened to Kwa-Mbonambi on 17 April 1903 (18⅛ miles) and to Umfolozi (Matubatuba after 1919, Mtubatuba on the 1940 Standard Railway Map) (14¾ miles) on 26 June 1903 and Somkele (shown as Hlabisa on 1904 Standard Railway Map) on 17 September (12¼ miles).

In complying with the terms of the agreement, the line built by the Zululand Railway Company was purchased on 31 March 1905 by the Natal Colonial Government for £725 000 under Natal Act 7 of 1905. The coal deposits mined at the Somkele proved to be of mediocre quality and following the closure of the mine in the mid-1920s, the 12¼-mile section from Matubatuba to Somkele (names shown on the 1920s Standard Railway Map) was closed on 1 June 1928 and the track lifted.

THE EXTENSION FROM MTUBATUBA TO GOLELA

After the First World War, the continuation of the North Coast Line started in 1924 when the 81-mile extension was authorised from Matubatuba to Candover. During the 1926 session, parliament approved an extension of 12½ miles from Candover to Gollel, adjacent to the border of Swaziland Protectorate*. The Swaziland Administration contributed £10 000 towards the cost of extension which included building a 3580ft long bridge over the Pongola River.

The extension from Mtubatuba to Candover was intended to open up this remote region of northern Zululand to cattle ranching and afforestation as well as the cultivation under irrigation of cotton and sugar cane along the Pongola River. By taking the line up to the border of Swaziland, it benefitted the territory by providing an outlet from its southern border for railing agricultural products and consignments of cattle for export through Durban.

As each section was completed, the extension was progressively opened on the following dates in 1926: Hluhluwe on 1 July; Mkuzi (changed to Mkuze in 1953) on 1 August; Candover on 15 September and throughout to Gollel (renamed Golela in 1953) on 4 July 1927. The ruling grades on the line are 1 in 50 up and 1 in 66 down compensated for curvature – the sharpest curve has a radius of 528ft. The track was laid with 60lb/yd BSS rails on 1600 steel sleepers per mile. At the time of the line’s opening to Gollel, the steel girders for the Pongola bridge had been delayed from the overseas suppliers due to strike action. In the interim, timber trestles to support the rails were placed between the concrete piers of the rail bridge. Once the steelwork was received and placed in position, the timber trestles were removed.

The formal opening at Gollel on 4 July 1927 was performed by the Speaker of the House of Assembly, the Hon. E G Jansen, MLA for Vryheid. A bottle of champagne was broken as the inaugural train passed under the triumphal arch erected at the end of the platform.

Golela remained the railhead for 51 years until 1978 when the 94,5km connection from the junction at Phuzumoya, in Swaziland, to Lavumisa was opened. The direct link from the Eastern Line at Komatipoort was undertaken by SATS which constructed the 61,8km of the northern part to the Swaziland border from where Swaziland Railways provided the 59km long connection to a junction on the existing line north of Mpaka. The line was formally opened on 14 February 1986.

*Candover to Pongola River, 8 miles 51 chains in Natal, Pongola River to Gollel, 3 miles 56 chains in Transvaal.

The terminus at Gollel (Golela) was located in a corridor once part of Transvaal Province bounded by the Pongola River in the south, the Swaziland‡ border in the north and the Natal border in the east running along the ridge formed by the Ubombo Mountains. The reconfiguration of South Africa’s provincial boundaries carried after 1994 saw this corridor incorporated into KwaZulu-Natal Province.

‡Swaziland was a British Protectorate from 1903 until 6 September 1968 when the territory attained independence. Renamed in 2018 as ‘The Kingdom of Eswatini’.

RECONSTRUCTION

North of Empangeni, a 10-mile deviation between Eteza (Teza) and Mtubatuba was handed over to traffic on 3 June 1929 which included a substantial bridge comprising eleven 150ft spans over the Umfolozi – the third-longest railway bridge in South Africa at the time. This section of the original line was frequently inundated by seasonal flooding. After crossing the Umfolozi bridge the line passes through Haig Tunnel (870ft). A short section of the old alignment at the Mtubatuba end was retained as a service line to the River View Sugar Mill.

With the start of construction in 1968 of the Pongolapoort (Jozini) Dam, a section of line between Kingholm and Golela was deviated to keep the track clear of the full supply level of the dam. A new alignment beginning at 237 miles 71 chains (from Durban) to 245 miles 60 chains was built which included a spectacular bridge comprising 10 segmental arches over the Pongola River.

With the introduction of more powerful motive power capable of hauling larger loads, interloops were lengthened; stations and yard layouts were remodelled at Nseleni, Kwa-Mbonambi, Teza and Mtubatuba to accommodate the increased lengths of trains. The upgrading, partial doubling and electrification of the North Coast line from Durban to Empangeni were completed during 1974.

Although the ‘Coal Line’ is not the central feature of this chapter of SoAR, this highly-important line nevertheless warrants a brief mention. The ‘Coal Line’ was the first heavy-haul line to be built in South Africa and ranks as a watershed in railway construction and operation as well as serving as the testbed for 25kV 50Hz AC electrification. The output of some 40 mines in the coal mining area around Witbank is railed for export to the Richards Bay Coal Terminal. It uses track standards far higher than had been used elsewhere on the SAR network and is engineered to handle 200-wagon compositions stretching 2,5 km with a gross load of 20 800 tons hauled by multiple electric locomotives controlled by Distributed Power Operation. The first section from Vryheid East Yard to Richards Bay Coal Terminal was opened on 4 May 1975. At the same time, the existing lightly-built line between Vryheid and Ermelo with its fierce gradients and tight curves was discarded and replaced with a heavy-haul alignment. To complete the access to the coalfields, a new line was constructed from Broodsnyersplaas to Ermelo South Yard. The ‘Coal Line’ from Broodsnyersplaas to the Richards Bay Coal Terminal was officially opened on 1 April 1976. Track duplication, strengthening of the structure and formation of the line was completed in 1989.

Concurrent with building of the Coal Line, a section of the North Coast Line was taken up in 1972 between Empangeni and Nseleni and re-routed alongside the Nzeze Marshalling Yard. More recently, the same new section was duplicated and the 25kV 50Hz AC overhead wires extended to Nseleni. The 1200-metre long bridge spanning the Nseleni floodplain, opened in December 2009, is reported to be now the longest rail bridge in South Africa.

1. With the overhead power already switched on as far as Mtunzini early in 1974, steam-haulage of the Empangeni - Durban passenger trains was soon limited to just the first 38km out of Empangeni – and seeing that the 15CAs were in the process of being transferred away to the Transvaal – unusual motive power like this Class GMA began to appear on 240-up, the 6.45am daily except Sundays Durban – Empangeni all-stations passenger. Severe drought in the area during 1973-4 meant that loco water became a serious problem and brought about a rapid acceleration of the electrification project. Within months electric units would be running right through to Empangeni, with diesels arriving just a year later to take over the steam duties beyond Empangeni as well as on the two branch lines, bringing steam operations on the North Coast largely to a premature end.

2. This 15CA was approaching its destination station of Empangeni around 4 o’clock in the afternoon. If it seems to be working uncharacteristically hard for an engine at the end of its trip, the opposite facing temporary speed restriction board alongside the first wagon suggests it may have been running at reduced speed for some time (and could possibly even have had a signal check as well!), so is actually accelerating into the station. This would no doubt have been one of the rare occasions where the “railway gods” deigned to smile upon an enthusiast like Brian, allowing him to obtain a second shot of the loco just 4km further along the line! (See photo 122 in the previous chapter. Ed.)

3. Empangeni Depot 7 January 1964 – a class GE (1st order) – note the extended coal bunker. Engine number not recorded.

4. According to Les's allocation lists, in January 1973 Empangeni had 9 GEs, 10 GEAs, 15 GMAs, 18 GOs (recently transferred from Lydenburg on the dieselised Steelpoort branch) and no class 1s or 1As. Since at least one class 1 was noted here several times/year from 1968 until 1973 one can only surmise they were on loan from Mason's Mill or Glencoe. Not counting the class 1 (it served as the shed pilot), the total official allocation of engines at Empangeni was 74 at this time, so Garratts (52 of them in four classes) were in the majority. Depicted here are a GO and a GMA being readied for main-line assignments.

5. A second-series GE photographed on 7 January 1964 with cleaners wiping off some of the grime.

6. 2nd series GE 2274 outside the Empangeni drop-pits in mid 1970.

7. Class leader GE 2260 awaiting assignment at Empangeni in July 1970. Classic Beyer Peacock machines the six examples of the first series gave SAR faithful service for more than 50 years (there were two subsequent orders each with detail differences).

8. One of those GMAs (No 4082) depicted in photo 4 above is moving off shed and a GO (number unknown) is preparing to follow. Meanwhile a GEA has moved in behind it while 1 cl 1276 on shed pilot duties is awaiting instructions from the shed foreman. It is impossible to convey to you the busyness of the place at this time, right at the crest of SAR's traffic boom.

9. Class 19A 684 on shed at Empangeni. 7 January 1964. They worked almost exclusively north of here on the main line but did have occasional banking stints on the Nkwaleni branch.

10. Class 1 No 1276 on shed pilot duties at Empangeni Loco on 10 January 1973. Rags added that it was a very gloomy day! To the left, a GO Garratt and behind the class 1, a GEA.

11. A class GE No.2274 (2nd series) having a drink and in the background – class 14R No.1579 at Empangeni Loco – also on 10 January 1973.

12. Rags final shot in the Loco Depot at Empangeni: Class GO 2583 and class 1 No 1276. Behind 1276 is a class GEA – quite a spread of engine power covering an in-service period from 1904 to 1954!

13. A 2nd series GE and a GEA getting ready for the road at Empangeni in January 1963. Ideally we should have had a 1st series GE parked on the right to show all the shapes of the GE/GEA series but not that day. If any of our readers should have, or find such a photo and would be prepared to publish it on SoAR, we will gladly dump this one.

14. A rare quiet moment at the shed on a Sunday in July 1968. Note the dilapidated state of the roof. Although by this date the decision to electrify throughout to Empangeni had already been taken it was repaired (probably at the instigation of 'Pikkie' de Wet, the very short but very competent Locomotive Superintendent at Durban).

15. 2nd series GE 2267 – coaled, watered and ready to leave the shed for her next turn of duty on 12 January 1964.

16. Empangeni Shed was packed with various classes on 12 January 1964. In this image we see class 19 No 1367 with a GE behind her and a 14R getting wet under the waterspout.

17. The shed’s mascot found a good place to take a nap on the cab steps of a class 19A. This reminds me of Alec Watson’s “Felix” in De Aar – Felix the cat would accompany Alec to the Depot and patrol around the 25s and 25NCs during the day – often watching the Koi fish in a storage tank with longing eyes for a good meal.

18. Lots of steam power in Empangeni Loco on 12 January 1964. From left to right: a class GF, 19A and a class GEA.

19. After their introduction in service on the Kimberley - Vryburg section in 1928, the 4 class 19 locomotives worked the rest of their years out at Empangeni, mainly on the Golela branch. Here No 1367 poses briefly for her portrait on 19 September 1966. Shortly after this date she was fitted with a Standard No 1A boiler and thus became the only class 19R.

20. Leith arranged to have 1367 placed in the Empangeni Shed for another portrait while she was in the Depot for a boiler washout – this time fitted with the Std boiler which made her the only 19R in service.

21. Class 19 No 1368 arrived on shed while I was walking around. The crew are seen unloading their kit before the Shedman takes her for fire-cleaning and watering. 12 January 1964.

22. Class 14R 1747 ready to depart Empangeni for Durban with 292-up at 07:10. The driver is taking advantage of the few minutes left to enjoy a mug of coffee as there will be little time for that once they get started. Some members of the Natal Branch of the Railway Society of Southern Africa had taken the 22:00 overnight train from Durban, photographed it departing Empangeni for Golela behind 19R 1367 at 06:30 and then joined this train for the return trip to Durban. Date Saturday 5 April 1969.

23. The private siding to Empangeni's Sugar Mill did duty as the yard head shunt at its southern end. In April 1969 14R 1584 was kicking out wagons at the north end of the yard. On the left is the Nkwaleni branch and the main line to Golela takes off into the middle background. On the right, construction of the link to the new harbour at Richards Bay has started.

24. A Class GE stands ready to depart in Empangeni yard in 1967, probably destined for the Nkwalini branch line. Outside of the sugar season traffic on the Nkwalini branch was generally quite light, as can be deduced from the short wagons directly behind the loco - these probably made up the bulk of this train. However, traffic increased dramatically during the peak cane harvesting season – usually in the drier period from about July to September, when the leaves and other waste material could be burnt off easily before delivery of the cane to the Felixton Sugar Mill. Oh yes, and that is a really old GZ bogie cattle wagon on the next line, dating back to the late 1920’s.

EMPANGENI – NKWALINI BRANCH

Authorised under Union Act No.33 of 1925, this branch, 39 miles 67 chains long, was officially opened on 19 September 1930 by Mr H Graham MacKeurtan, KC. It was built to open up the agricultural area in the valley of the Umhlatuze River east of Empangeni to sugar cane, cotton, wattle and dairy farming. At its peak during the early 1970s it was an extremely busy, and presumably profitable line but in common with practically every SAR branch line railwaymen somehow managed to lose ALL the business by 1990.

We hope you've booked your hotel in Empangeni because we're about to take you on a side trip up the Nkwaleni branch - incredibly busy in the sugar season.

25. During the peak sugar cane harvesting season early in 1973, this GEA threads its way across the main road at Mandozi, the last siding on the Nkwalini branch line before the junction at Empangeni. During these intensive harvesting periods, as many as five additional trains per day (seven days a week) would be rostered to clear the relentless flow of stalk cane from this branch to either Empangeni or Felixton sugar mills. Steam would only see two more full seasons of cane traffic on this line, with diesels taking over completely in September 1975.

26. The same train as in photo 40 rumbles downgrade towards Empangeni with a block load of stalk cane, probably destined for the Felixton Sugar Mill. Perhaps this is the appropriate place to make mention of an unusual operation reported (in the RSSA's SA Rail magazine) to be taking place at Felixton during the sugar season of 1973/4. During the peak milling season, Operating would send a locomotive (usually a GEA or GO) from Empangeni to Felixton, where it would spend the day shunting loaded wagons into the mill siding as well as clearing empties in order to expedite their return to the canefields. At the end of the shift, the loco would return to Empangeni by simply attaching to the front of the first available northbound goods train which, in many cases, turned out to be a double-headed Garratt combination. This meant that the train would complete the last 10km of its trip as triple-headed Garratts! * Unfortunately no photos of this remarkable operation (which evidently went on for several months) are known to exist, but word about it must eventually have reached the attention of Pikkie de Wet, the Locomotive Superintendant at Durban, who would have promptly sent out an instruction expressly prohibiting this method of working!

*In case you're wondering why all the fuss about what would be a perfectly logical answer to the congestion problem that was chronic on the North Coast at the time, triple-headed Garratts would have been OK if they were all GMAs or GOs for these were the only SAR Garratts that had self-adjusting pivots. The GEAs, inspite of recommendations by Beyer-Peacock, were supplied without this essential maintenance reducing improvement.

27. The Nkwaleni branch was the last stamping ground for 1st-series GEs, which were also the last of the traditional Beyer-Peacock no frills 'everything square' design in service on SAR.

Class leader #2260 was heading the daily T&P, 215-down with mostly empty DZs for loading with pulpwood in July 1971 (i.e. before SAR tumbled to the idea of ribbed timber wagons!).

28. The Nkwalini branch had some beautiful scenery, as captured in this going-away shot of one of the 10 second series of class GE with raised coal bunker, No 2273. September 1973.

29. In September 1973 we found GE 2263 working the cane season, approaching Dukaneni with a full consignment for Felixton Mill.

30. GE 2260 on the same day with the same train depicted in photo 27 paused for passengers at Dukaneni. Note the primitive means of ensuring that loaded stalk cane stayed put (it didn't always). The ES 4-wheeler in the foreground had probably brought a load of fertilizer for the local farmer. These short wagons were extremely popular with SAR's customers - that in itself was sufficient for the accountants to look upon them with suspicion. Eventually they managed to kill off this business by charging the same daily rate for demurrage as for a bogie DZ which had double the capacity.

31. GE 2275 makes an impressive silhouette as it crosses the Mfuli River on its way to Nkwalini. This bridge was washed away by floods in February 1977. Following a concentrated effort, involving both day and night shifts, the line was reopened in April of that year. Date Friday 8 September 1972.

32. An unknown GE plying its way towards Empangeni with pulpwood destined either for Mandini (more likely) or Sappi-Saiccor at Umkomaas in July 1971. At this time the labour-intensive method of loading pulpwood was still the norm - the new 'ribbetjiestrokke' (= ribbed trucks) classified ST had only just been introduced.

The Nkwaleni branch could only have been profitable. Apart from the intensely busy six-month cane harvesting season there was year-round pulpwood, general freight and fruit in season.

33. A first series class GE climbs out of the Mfuli River valley. Unlike their Rhodesian counterparts, South African drivers often preferred to drive their Garratt locomotives backwards, especially in the summer months. It was cooler but you had the disadvantage of coal dust blowing into the cab. This could, to some extent, be reduced by continually wetting the coal in the bunker and putting up canvass screens as seen in the picture.

The hand-me-down 96lb rails alongside the track shows preparations are already in progress for the arrival of class 34 diesels some two years hence. With their 19-ton axleload the 34s needed heavier rail than the Nkwaleni branch's 60lb so SAR took advantage of released rail from the main line where 116lb rail was being installed. A careful examination of the top right-hand corner of Peter's photo will show you the relaying gang at work. September 1973.

34. The same train as in the previous photo passing a field of sugar cane en-route to Nkwalini. The line clearly had a water shortage given that most trains had auxiliary water tanks. The difference between 96lb and 60lb rail easily discernable.

By the 1950s steam was a swearword among railway executives, especially accountants. On SAR this bad language only arrived a bit later - in the sixties. But lovers of steam found it hard to understand how good locomotives could be scrapped and track ripped out to make way for modern diesels. An example is here where the trusty GE is working a much lighter load (albeit on lighter but perfect track) than the new class 34s could handle but whether economics justified replacing a labour-intensive old steam engine in good working order with a multi-million Rand diesel, along with upgraded track, is beyond this prejudiced writer (Charlie!) to calculate or comprehend.

35. Banking of trains on the 64km Nkwalini branch was quite common, with various combinations of locomotive possible. Here a westbound class GEA is banked by a domeless class 19D out of the Mfuli River valley. September 1973

36. First series GE 2263 returning to Empangeni in the afternoon with a load of stalk cane for the mills. September 1973

37. A special arrangement was made with the shed staff at Empangeni to run the morning train out with the GE chimney first. The driver was initially not impressed but a few ice-cold cokes seem to help a little. September 1973.

38. An Empangeni bound GE has a full load as it makes its way back home. The bunker is clearly becoming depleted as it passes various forms of interesting vegetation. September 1973.

39. Another view of the series 1 class GE returning to Empangeni with its load of sugar cane. September 1973

40. Class 19A 676 brings a load of empties for Nkwalini up and out of the Mfuli River valley on Friday 8 September 1972.

It is interesting to note that Timetable 157 of 30 November 1959 still shows a passenger service on the line. The train left Empangeni at 09:00 Mo, We & Fri and only arrived at Nkwalini at 13:30, 4½hours for only 64km. The return journey was equally slow, leaving Nkwalini at 14:30, arriving at Empangeni at 19:00.

41. A class GE No 2270 leans into the curve near Heatonville for Nkwalini on 10 January 1973.

42. Heatonville, about a third of the way up the branch, was a major cane and pulpwood loading point as well as a good destination for local general freight in July 1971. GE 2262 was picking up an empty fruit van and some loaded stalk-cane bogies. Contrary to regulations the whole ensemble headed off to Empangeni with the ultra-light 4-wheeled fruit van sandwiched between the engine and > 1000 tons of cane.

43. So far we haven't mentioned that the Nkwaleni branch roughly follows the Umhlatuze river. A mile west of Dukaneni the railway suddenly swings away from the river to navigate around some cliffs and at this point the pulpwood train we were following decided to follow suit. The picture was made in June 1970.

44. GE 2263 bringing quite a mixed load into the quaintly named Nkwenkwe siding in September 1973. Included in the consist are two B bogies of milo, a foul-smelling waste product of the sugar-refining process that makes good fertilizer.

45. In the afternoon of the day we took photo 27 we followed GE 2260 on her return trip with a full load of stalk cane which she worked through to Felixton.

46. It's GE 2260 again, a bit further along on the same turn to Felixton depicted in the previous photo.

47. At Dukaneni some three years later, cane loading was in full swing with GE 2263 restarting after stopping to offload a passenger. She is on the same train depicted in photo 44 minus one truckload of milo, shunted off at Nkwenkwe. Those 4-wheeled vans at the rear are for loading with pineapples at Nkwaleni.

48. There were some hefty gradients where the line bends northwards away from the Umhlatuzi valley towards Heatonville thus restricting the loads of down block cane traffic to 15 bogies as far as Heatonville from where the load could be increased to 25 bogies. In July 1971 GE 2262 was working up the bank with a light load of 9-bogies of stalk cane in preparation for picking up more loads at Heatonville.

49. Approaching the confluence of the Mfule and Umhlatuze rivers at Quneba the line hugs the Umhlatuze river on its way down from Nkwaleni. In September 1973 GE 2260 was plying this scenic stretch on her way to the Mfule river bridge.

50. GE 2260 doing a spot of shunting at Nkwaleni having arrived earlier that day in September 1973 with 215-down T&P. After placing the wagons she would pick up the pulpwood loads shown in the previous photo. At this time Nkwaleni was almost overwhelmed with traffic of many descriptions, ranging from stalk cane and pulpwood to livestock and fruit.

51. As fast as they could, three cranes were loading stalk cane at Nkwaleni while #2260 patiently looks on.

51. With many sidings to shunt in both directions on the branch there was always a danger of running short of coal so there was a rudimentary coal stage at Nkwaleni. 2nd-series GE 2268 was being refueled for her return trip in June 1970.

52. GE 2260 on its way to Felixton in the loop at Manyela for a 2nd-series sister bringing more empty bogies for loading stalk cane, the crude poles already in position. Note how all the tracks are shining - this was high season. September 1973

53. 2nd-series GE 2274 bringing in yet another load of harvested cane stalks down the hill from Heatonville. The fireman can relax, practically all the way to Felixton. July 1968.

54. More than fifteen years after the last regular steam had disappeared from the North Coast, and seemingly against all odds, steam returned to the 64km Nkwalini branch line in 1991 – albeit just for two days! To put this amazing development into context, one needs to take note that virtually every regulation in the SAR Natal Local Appendix concerning routes and facilities available to steam locomotives in Natal state categorically “All classes except 25/25NC”. Now, thanks largely to the untiring efforts of retired Empangeni Goods & Passenger Superintendent Gerald Allner (and no doubt some others) to organise a “Great Train Race” competition in the Empangeni area, approval was granted by Spoornet’s Head Office for a steam locomotive to be used for this once-off event on the Nkwalini branch line – and not just any steam locomotive, but a Class 25NC!! So it came about that Class 25NC no. 3476 “Griet”* was worked under its own steam from Braamfontein to Empangeni in April 1991. A test trip was run with 14 coaches on a weekday before the Saturday event, and despite the fact that neither the loco nor her driver were acquainted with the line, 3476 ran faultlessly on the 1:40 gradients and sharp curves of the branch. In this photo we see the actual passenger train climbing the grade away from the Mfuli River crossing, about 20km from Nkwalini, on Saturday, 13 April 1991.

*"Griet" was the shortened name of Transnet Chief Executive Braam le Roux's wife Margriet (or Marguerite in English). Braam, along with his friend and colleague Helmuth Hagen were both closet steam enthusiasts ('closet' because you couldn't broadcast it those days) and would most certainly have given this show their blessing.

55. No. 3476’s safety valves have lifted to show how effortlessly she was managing the 14 coach train, seen here at Bedlane, between Dukaneni and Nkwalini. Nkwalini was the terminus station and the finishing point of the race, where all the day’s festivities were scheduled to take place. The train was held back at several places along the way to give the runners a fair chance, so I suspect the train did not win on the day! The yellow coach was a service coach for the loco and technical crew, with the passengers travelling in 13 third class sitter coaches. It is hard to imagine that this magnificent machine was almost 40 years old but still looking as if it has just emerged from the works.

56. While the awards and speeches were being made at Nkwalini, 3476 turned on the triangle (she made it with literally millimetres to spare!), had her fire cleaned and attached a waiting water feeder tank wagon. Sometime later in the afternoon we see her trundling back downhill through Nkwenkwe siding, about 13km along the line from Nkwalini, sporting the “Spoornet” emblem on a banner attached to her fireman’s side smoke deflector.

57. As if the 1991 event was not pushing things far enough, the organisers unbelievably convinced the Spoornet authorities to repeat the whole exercise in 1992, and so Saturday, 12 April 1992 saw 25NC 3476 “Griet” once again making a visit to the Natal North Coast. In this photo we see her blasting into Mevamhlope with the now-standard 14 coach train, a little over thirty minutes after taking off from the starter’s gun at Empangeni. There was no prior test run done this time.

58. Majestic is the word....... 3476 stretching her legs near Heatonville, 28km from Empangeni. Thanks to some “artificial hold-ups” it had taken almost an hour and a half get here.

59. As can be seen from the steam effects, 12 April 1991 was a relatively cool day, which must have been great news for the runners competing against the train. With her safety valves lifting and adding to the steam effects, 3476 was clearly not taking any strain as she climbs steadily towards Nkwenkwe siding – at this point the train is only 15km from the finishing post.

60. In a scene reminiscent of the tight curves and steep gradients of the Bethlehem – Ficksburg line, 3476 leans into a right-hand curve as she climbs away from Dukaneni.

61. The enthusiastic Zulu dancers are in full flight as 3476 and her train draw up to the finish line at Nkwalini station. Once again, we have no idea who won this epic battle between man and machine, but a sad sign of things to come may be seen in the background. Note the BC sugar cane wagons standing out of service in the siding – their white-painted axle boxes confirming their sad “not-to-go” status. Hopefully some of them may have been converted to roller bearings and seen further service, but by the 1990s the Railways were rapidly withdrawing themselves from various cargo markets.

62. With the day’s activities done and dusted, 25NC 3476 departs Nwalini for Empangeni. This event was unfortunately not repeated although surprisingly another Class 25NC (No 3407) did visit Empangeni a year later, in April 1993, being the motive power of a Union Limited special tour train that ran down the Coal Line (Yes, you heard right!) before heading north to Golela and Swaziland en route to Komatipoort.

63. At the instigation of the Coal Line Resident Engineer, Helmuth Hagen (later, Assistant General Manager, Technical) steam was used during the construction of the Richards Bay Coal line and this shot was taken of 12AR 2121 in August 1974 near Ulundi where a water gantry was specially provided. Her condition suggests a regular driver and fireman during this work.

64. From this angle the concrete arches of the White Umfolozi bridge look positively graceful. A carefully composed picture by Allen Jorgensen of the 12AR crossing the bridge with a ballast train.

65. From this angle the bridge can be seen for what it actually is, a no-frills structure designed to cross the White Umfolozi on grade with provision for future doubling.

The 'Coal Line' runs from the centre of the Transvaal coalfields at Broodsnyersplaas to Richards Bay. It was designed to carry 25 million net tons/year over a single line with 1/66 against loaded trains. The line was opened with diesel traction in 1975 and the target tonnage was soon met and exceeded. Plans for doubling and electrification were initiated right from the start; no sooner completed than deviations and extra tunnels to ease the gradient against loaded trains to 1/132 had to be made. The target grew to 90 million tons/year but was never reached, stalling somewhere in the 70-80 million ton mark. What with the world-wide move against coal-fired power stations it is now unlikely that these tonnages will be exceeded.

All the same, this 'Heavy Haul' line is an amazing achievement, considering that it runs from an altitude of > 6000 ft down to the coast through mountainous terrain almost all of the way. With electric haulage trains eventually reached 200 wagons in length and 20,800 tons in weight. The aforementioned facts justify a chapter in their own right and if we live long enough we would like to bring one to you. More likely it will have to be tackled by one or more of our 'Soul of A Railway' successors.

66. This 25 Kv AC electric – class 7E3 E7220 in Spoornet livery was photographed at Richard’s Bay on 10 March 1993.

67. John Middleton called in again at Richards Bay on 11 June 1994 and photographed these class 7E1 AC electrics in their original Gulf Red livery.

With those (we trust) pleasant interludes up the Nkwaleni branch followed by the Richards Bay Coal line behind us we return to the north end of Empangeni to complete our original goal of travelling up to Golela on the North Coast main line.

68. A general view of Empangeni yard absolutely jam-packed with traffic in July 1968. HOW did SAR manage to lose it all? There was quite a grade against northbound trains starting out of the yard as is evident from this GEA working hard to get its short but heavy load of locomotive and power-station coal out of town. The loco shed on the right is full of activity, with a 15CA/B moving off shed and careful study reveals several other movements in progress. Also of interest are the inspection trolleys on the left and at least five cabooses on the right.

69. A rare sighting in July 1968 was double GEAs both chimney first, having fetched a big load of stalk cane from the numerous cane-loading sidings on the main line north of Empangeni, in this case bound for Felixton Mill.

70. Just before 6am on a Saturday morning early in 1974 this 19D departed northbound from Empangeni with 293-down, the overnight passenger which had left Durban at 10:10pm the previous evening. Running daily except Sundays, it was officially classified as a mixed from Empangeni to Golela, but on Saturdays it ran only as far as Mtubatuba and, being booked to convey its entire load of 12 coaches all the way, there was not much room for goods traffic (even if there was any offering), as a 19D mixed-train load was set at 500 tons.

71. Less than fifteen minutes after the arrival of 293-down at Empangeni on 11 December 1973, class 19 No 1369 departed northwards with a goods train towards Golela while 14R 1753 did a spot of shunting in the background.

72. Moments after class 19 No 1369 had departed northwards, 14R 1753 (partially obscured), GO 2584 and No 2595 were at the north end of Empangeni yard on 10 December 1973.

Glen adds: GO’s along with GEA’s were seen on goods trains between Gingindlovu and Empangeni. Also, GEA 4044 was seen on a northbound goods at Sihlepu.

73. By 1970 construction of the link to Richards Bay mentioned in the caption to photo 23 above was far advanced. With no turning facilities available the GMAM was returning from end of completed track with an empty perway train.

74. Following hard on the heels of the Mtubatuba passenger train was this GEA heading a departmental material train destined for the still under construction Vryheid – Richard’s Bay coal line. The lack of watering facilities along the line meant that steam locos had to take their own water supply along with them – hence the two auxiliary water tanks in tow. Note too that the ballast wagons are empty and would no doubt either be loaded, or exchanged for loaded ones en route. As an aside, have a look at the interesting variety of private cabooses parked in the sidings adjacent to the main line – on the right hand side are a pair of old trainmen’s lantern-end cabooses; in the centre a much newer 1960s era Union Carriage & Wagon workmen’s steel caboose with an elliptical roof, whilst a 1940s Durban SAR Workshops-built officer’s wooden caboose with bullnose roof ends is on the extreme left. Also, something of a rarity are the short four-wheeled XE domestic water tank wagons standing on each side of, and partly obscured by, the X bogie tank wagon.

75. For almost a decade construction trains on the Richards Bay heavy-haul coal line were steam-worked, thus taxing to the utmost the mental powers of operating staff - watering, coaling and servicing of the locomotives having to be planned in the minutest detail.

Just north of Empangeni Brian came across yet another pair of Coal Line works trains, running combined, with a Class GO bringing up the rear. This GO, having only a van as its load, was most probably being used to assist the other loco with its heavy cargo of ballast, and was no doubt en route to collect its own set of empties from the section. Interestingly enough, during the construction of the Coal Line, the distance of more than 200km between Vryheid and Richard’s Bay was considered too far for the efficient operation of works trains from each end, so a Class 12AR was temporarily stationed at Ulundi and would work material trains (delivering rails, ballast and/or sleepers) as required in each direction from there.

76. Beyond Empangeni the North Coast route generally became a lot flatter, and smaller locos were the norm, making it look much more like the branch line it really was. Here for example, a Class 19 locomotive heads a goods train past Kulu siding on the original 1903 alignment between Empangeni and Nseleni, which was abandoned and lifted in 1973. The current line runs on a much more easterly course via Richard’s Bay Junction, on the seaward side of Nseleni Dam. The detritus in the foreground is a sure sign that this siding has recently been used for the loading of sugar cane.

77. A few miles out from Empangeni GO 2586 was in action during the 1973 season with empty cane wagons for distribution among the loading points along the main line to Mtubatuba. The siding is thought to be old Nseleni which soon after was radically changed when it became an interchange point with the coal line.

78. A 14R working hard past plantations as it heads south toward Empangeni with a mixed load, the first six wagons of which could have been a consignment of processed sugar products: bagged sugar in the OZs (tarpaulined to prevent ingress of water), bulk sugar in the tarpaulined C wagons and molasses in the tankers; either from the Umfolozi Mill at Mtubatuba or the Pongola Mill at Golela. The rest is empty AY ballast wagons from the relaying north to Mtubatuba on their way back to the quarry at Mvemhlope* on the Nkwalini branch . September 1973.

* Thanks to Martin Carrick, a former engineman who worked the North Coast, we can tell you that the empty ballast wagons would probably have been destined for Mvemhlope from where a quarry provided much of the ballast for the lower sections of the Coal Line as well as the Golela line.

79. A Class GO heads a block load consisting mostly of B bogies with timber, thought to be just south of kwaMbonambi. Although by 1974 upgrading of the track north of Mtubatuba was already well under way in anticipation of the imminent arrival of diesels, trains originating beyond Mtubatuba were often worked all the way to Empangeni by the lighter Classes 19, 19A, 19D or GO locos.

80. A track worker pauses to watch the passing show as a GO hauls a splendidly mixed freight southbound in the kwaMbonambi area circa 1973. To the extreme right is the line of gum trees planted as a windbreak to prevent locomotive sparks from blowing into the adjacent canefields with potentially disastrous results, as many claims for fire damages were laid against the SAR Administration by farmers. The Class GOs were reportedly somewhat notorious for throwing sparks and, judging by the direction the smoke is being carried, the prevailing wind was in fact blowing on this day!

81. A class GEA with a block load of 16 wagons of pulpwood logs from Kwa-Mbonambi destined for the Mondi mill at Tugela or the Sappi/Saiccor factory at Umkomaas heads south toward Empangeni. September 1973.

82. The sole 19R 1367 piloting a 19D through Kwa-Mbonambi with a combined train in mid 1967. By the late sixties even north of Empangeni was so busy that operating had to resort to all sorts of tricks to reduce single-line occupation. This combined freight-and-mixed had empty molasses tankers for the Umfolozi Sugar Co-op, some general freight and passengers all heading north.

83. The four class 19’s were allocated to Empangeni for many years. Here a very grubby class 19 is working hard southbound toward Empangeni from Mtubatuba. By 1975 these locomotives would be withdrawn. September 1973.

84. A class 19 making a gentle start away from the water stop at Mposa in July 1968. On the right is universally well-known and best-loved railwayman and railway enthusiast, AE 'Dusty' Durrant. I have to confess to my colleague Les, who disapproved of such things, that both the smoke and the restart were arranged during the water stop.

85. Charlie reporting: on a Saturday in August 1973 I encountered the strangest combination I ever saw on the North Coast (a route noted for surprises). That 253-down mixed had joined forces with the goods that had been running ahead of it wasn't so unusual but the make-up of the passenger portion was weird: tri-compo day coach followed by three UCW tin coaches then a tarpaulined OZ followed by a pair of posh-looking clerestory saloons* then three ordinary clerestory coaches. At the time there was so much going on that it didn't seem at all extraordinary but with hindsight I should have been more curious.

*Our resident coach expert, Peter Stow advises: "it appears to be the GM’s articulated coach 78 and not 2 separate coaches. One can only speculate as to why he would be visiting Golela and why the trains composition was so strange."

Here may be a clue: the Pongola dam, completed in 1973, involved a major deviation of the railway. It is possible that the SAR GM was on his way to attend the official opening of the dam.

Those 19Ds were working up the bank away from the Umfolozi River, having crossed the bridge destined to be washed away in the great flood of 1984 (see Charles Strydom's account of its replacement). Ahead of the train is the short Mtubatuba tunnel and then a half-mile gallop into the station.

UMFOLOZI RIVER BRIDGE WASH-AWAY BY CYCLONE DOMOINA

28 JANUARY – 2 FEBRUARY 1984

Contributed by Charles Strydom

Storm centres, with precipitation exceeding 700 mm (28in) over 28 January to 2 February 1984 occurred over the upper Umfolozi, Mkuze, Pongola, Middle Usutu and Komati Rivers. The Ingwavuma tributary had 700 mm (28in) of rainfall in both the upper and lower reaches. Discharge on the Umfolozi River reached 16 000 cubic metres per second (570 000 cu ft/s) which is about 3 times the 100-year return flood while flow velocities of 2,6 metres per second (8·5 ft/s) were measured.

On 31 January Cyclone ‘Domoina’ dumped a massive amount of water in the White and Black Umfolozi catchment. The bridge with the most and longest spans in South Africa, and that had stood there since 1929, was almost obliterated. It was an important link from landlocked Swaziland to the seaports down south. Ralph Travis, Dave Cowie and I flew by helicopter from Johannesburg to inspect the damage. The sight from the helicopter was unreal: 10 of the 11 spans gone! Only one span left in the distance and piers snapped off like matchsticks; nine of the 10 x 150ft, 160 ton each, steel truss spans were swept downstream. Some as far as 5km.

Dave Cowie, the track engineer was more interested in building a low-level causeway to get the trains running as soon as possible than in the bridge. Meanwhile I wandered about in the river bed trying to fathom a solution. He was so intent on doing HIS thing he wanted to leave me there because I had wandered off on my own. Had he done that, it would have changed the course of events. We flew back that Saturday night. The next day, Sunday, I was in the office preparing a possible scheme using the Callender-Hamilton emergency equipment that we had only recently acquired. However, their system did not cater for skew spans. I prepared a drawing showing how we could achieve this by adding some extra structural members.

On Monday morning, I put my suggestion on the table. The Chief Civil Engineer, Mike Myberg looked at it and said: “When can you fly to England to discuss with Balfour Beatty?” I said "Tonight". He said, "A bit soon but how about tomorrow?" I met Ron Breach and Doug Wells in London and devised the adaption of their system. Modus operandi: assemble 3 spans on the bank and launch 450 tons of steel 600m over the river using centre-hole hydraulic rams and ‘Diwidag' bars. 3 spans are optimum to provide counterweight during launching. The nose of the leading span would deflect 950mm on full-extension, then lifted with jacks onto specially-designed skates, advanced just far enough for the tail of the last span to be lowered in the same way and the process is repeated from one pier to the next. To launch the first 45m span took almost a day. With some innovations, the rate was eventually reduced to 45 minutes. I was ably assisted by Andre Botes, Albie Vos and Daan Strydom. The first 3 spans were assembled and ready for the first launch when I detected a serious deficiency of one of the crucial structural members that would have resulted in a spectacular collapse. I halted proceedings and told Ralph I would not launch without confirmation from the designers. They agreed and arranged for immediate strengthening. Their young design engineer Mark Grant admitted to an oversight and thanked me for saving his skin. (Somebody up there liked me again).

The bridge was destroyed on 31 January 1984, 9 spans replaced and one recovered from the river bed and the line reopened to traffic on 4 October 1984. Almost 9 months of excitement. It was my swan song.

Some background information:

The day after our helicopter reconnaissance I had a proposal on the table to use the Callender Hamilton Emergency Bridging System that we already had in stock in the Bridge Yard. There were three complete spans of 45 m each. There was also material for three trestles of 45 m each which could be used for trusses but these trestles needed cross girders and stringers. This meant that reconstruction work could start immediately with the balance of the components to be ordered from Cape Steel. A second proposal to replace the bridge with prestressed reinforced concrete spans was also tabled as an alternative to steel. The steel option was favoured as a result of cost consideration and the fact that the Main Railway Line from Swaziland could be reopened to traffic at a sooner date.

A railway deviation over a temporary causeway on a series of pipes was constructed post-haste and the line opened for traffic within a few weeks. Strangely, this causeway, as well as the temporary road bridge, was destroyed by a flash flood on the day that the new bridge was officially opened. Remarkably, the Callender Hamilton Modular Bridging System had only been in the process of development and acquisition by Balfour Beatty in England (to our specification) since 1980, ostensibly to cope with sabotage scenarios. The material was supplied in 1983 and was ready for use in 1984.

86. Shortly after 8am and following a remarkably rewarding couple of hours of trainspotting, Brian photographed the 19D with train 293-down reaching its final destination at Mtubatuba. As mentioned above, on Saturdays the train would terminate here, with the coaches being shunted out of the way, to be worked back to Durban as overnight passenger train no. 294 at 7.5pm on Sunday evening.

87. As described by Bruno in his introduction to this chapter, the original terminus of the North Coast route from Durban was at the coal-mining town known as Somkele (initially Hlabisa). It is shown here in NGR days, i.e. between 1905 and 1910. Note the clerestoried engine shed in the background and we thought it appropriate to repeat photo 29 from the previous segment of our North Coast history showing one of the Zululand Railway Company's locomotives. The section from Mtubatuba to Somkele closed in 1928 as shown on Bruno's map above.

The saw-tooth profile of the section from Port Durnford to Somkele was a consequence of the many river crossings; some of them like the Umhlatuzi and Umfolozi quite substantial.

88. CORONATION DAY 1902: An unidentified Baldwin tender locomotive of the Zululand Railway Company elaborately decked out in palm fronds, shields and flags for the occasion of the Coronation of Edward VII and his wife Alexandra of Denmark as king and queen of the United Kingdom and the British Dominions which took place at Westminster Abbey in London on 9 August 1902. As regards who is posing for the photograph, my guess is the driver is standing next to the locomotive while the fireman is standing in the opening of the cab. The identity of the well-dressed gentleman in his Stetson and the other two men is open to conjecture.

For the construction of the line from Tugela to Somkele, Messrs Middleton Bros., the contractor for the Zululand Railway Company, imported 3 construction locomotives, one side tank 2-6-2 and two 2‑6‑0 with tenders, from the Baldwin Locomotive Company in the USA. The six-wheeled tenders of these locomotives were unique in that they were carried on a single pair of wheels and a trailing bogie. Under the Zululand Railway Company, the locomotives were numbered 1 to 3; when they were absorbed into the NGR locomotive stock, the tank locomotive was assigned number 512 and 513-514 for the tender locomotives.

What is of further significance is that the Baldwin tender locomotives were the first true tender locomotives to operate in Natal, in 1902. NGR’s first tender locomotive, Hendrie ‘B’ No 275 entered service on 25 October 1904.

89. About 1½ hours after leaving Golela, 296-up has pulled into its first loco servicing stop at the village of Mkuze, where the brooding peak of Ghost Mountain seems to preside with a moody presence over the proceedings.

90. A Class 19AR, one of the highly successful and ubiquitous series of Classes 19 to 19D locos, heads a mixed goods load towards Golela (pronounced “Gollell”) in the latter half of 1972. This section of line appears to be the realignment between Kingholm and Golela, necessitated by the construction of the Pongolapoort (also known as the Jozini) Dam, which would flood a section of the original formation once completed in 1973 – it also added 17km to the 371km “branch line” from Durban to Golela, effectively wiping out the 15km saved by upgrade deviations at the southern end of the route. When officially opened in 1973, Pongolapoort was reported to be the largest dam in South Africa. The realignment of the railway brought it adjacent to the oldest game reserve in the country – the Pongola Game Reserve, officially proclaimed by the President of the Transvaal Republic, Oom Paul Kruger, in 1894 – four years earlier than the Sabie Game Reserve which would later be expanded to become the Kruger National Park.

Actual construction work on the dam began as early as 1963, and negotiations had been concluded with Great Britain (who governed Swaziland at that time) for part of that territory to be flooded by the dam. The construction was also delayed by unstable rock and concrete that wouldn’t set properly in the oppressive heat of the Pongola River valley – the former resolved by the use of extensive rock anchors and the latter by using copious amounts of ice to achieve acceptable curing temperatures! So, when Swaziland gained independence in 1968, the dam was still far from complete, and the agreement reached with GB was declared null and void by the new Swazi government, sending the whole scheme back to the drawing board. I am not sure if any compromise was ever reached, but it has meant that the dam could never achieve its full capacity – it seems to be maintained at about 40 to 60% capacity at most. All these issues meant that it took nearly ten years to complete the work and the dam was thus only officially opened in 1973.

Note the “new” twin block-type concrete sleepers used to build the deviation in the late 1960s. These would be used all over South Africa, but rapidly fell out of favour, apparently because the connection pipe, rod or bar tended to work loose and then failed to maintain the gauge. Interestingly, they are still much in evidence elsewhere in the world, particularly in European countries. The auxiliary water tank behind the loco is evidence of the frequent droughts experienced in this part of the country, a factor that would eventually accelerate the demise of steam on these routes.

91. On 10 December 1973, domeless 19D No.2680, having arrived at Golela on 297-down, proceeds around the south leg of the triangle prior to taking water (geographically it is the north leg!). This locomotive had worked 295-down from Empangeni to Mtubatuba where 297-down Mtubatuba to Golela originated, with only the two parcels-carrying vehicles going through to Golela. A canvas covered DZ wagon No.51-132 745 consigned Durban to Golela and the water tank were attached at Mtubatuba. This 'train' had originated at Durban as 295-down through to Mtubatuba with 5E1 1099 and 5E 718 being replaced at Gingindlovu by GMA 4072.

A similar arrangement occurred on the return journey with 298-up Golela to Empangeni with the engine losing its auxiliary water tank at Mtubatuba and 240-up originating at Empangeni with GMA 4086 running chimney-first being replaced by 5E1 E796 and E607 at Gingindlovu through to Durban.

92. Golela station pilot on this day, 19A 686 also proceeds around the triangle at Golela into the apex to better position itself for shunting the arrived Train 297 from Durban.

93. Two 19Ds Nos. 2680 and 2749 were positioned for me at the apex of the triangle at Golela on 10 December 1973. The friendly driver of 19D 2680 obligingly spotted the locomotives for the photograph as both locomotives took water with a “garden” hose. Presumably 19D 2749 had followed 297-down with a goods to Golela. The apex of the triangle behind the photographer is now connected to the extended mainline just south of the Swaziland border before entering the new yard on the Swaziland side of the border (see Ashley's caption below).

Peter Stow provided this interesting snippet: "19D 2749 was later to become famous for a significant role she played in the history of steam in SA. She was transferred to Capital Park after dieselisation and later painted blue to shunt the Blue Train at Pretoria station and work the Cullinan train on occasion, especially during the jacaranda season. These activities were recorded by railway enthusiasts from all over the world."

At this juncture it also seems appropriate to give the reason for the haulage of locowater feeder tanks on this last leg of the North Coast line. Ashley has referred to the fact that the countryside north of Mtubatuba gets increasingly arid and this is indeed the reason for the wasteful haulage of locowater feeder tanks. It was ever so; the first reliable supply south of Golela was at Mkuze 42 miles away, but the 62 miles onward from there to Hluhluwe would have tested the extreme range of those MP1 tenders.

94. Strangely enough, the Class 19AR and train awaiting departure from the railhead and international border station of Golela on a Sunday afternoon in 1972, are actually in the Transvaal and will only cross into Natal once the new bridge over the Pongola River, about 13km away, is reached. Any local cars parked outside the station would thus have had a TAG registration (Piet Retief district). I think this was one of only three “inter-provincial” branch lines in the old South Africa – the others being the Cape – Natal Line south of Pietermaritzburg, and the Standerton – Vrede line which ran between the Transvaal and the Orange Free State. The train is no. 296, (1.25pm Sundays only, Golela to Durban), the only complete passenger train to run from Golela each week – it would reach Durban at 5.4am on Monday morning. Forming a solid backdrop to the East, the Lebombo Mountain range shows that the train is about to leave in a westerly direction, a full 180° swing from the original alignment. Interestingly, the abandoned alignment would once again come into use – partially – about five years later, when the line was extended into Swaziland and linked up to the existing route to Maputo (and later on to Komatipoort), effectively changing its branch line status after roughly fifty years.

This concludes our coverage of the North Coast Line from the Umgeni Bridge just north of Durban, all the way up to Golela in the Transvaal and close to the Swaziland border.

In the next chapter we go south from Rossburgh near Durban – all the way to Port Shepstone.