Cape Midland Branches (1) : From George to Knysna ©

Important Announcement:

When we set out to produce Soul of A Railway the idea was to remind our readers what an excellent organisation the SAR&H once was. It was always our intention to translate the website material into book format. A few years ago our initial venture into publication was unavoidably postponed due to the ill-health of David Payling. Sadly, David recently passed on, but thanks to the expert assistance of well-known regular contributors: John Hunt, Dick Manton and John Middleton, the first book, covering the narrow gauge at Port Elizabeth, is about to be published.

“A Railway to Adventure”, derived from our coverage of the Avontuur line in SoAR, will be in large format with 264 pages and upwards of 350 illlustrations (many in colour) plus historical data compiled by the best team of railway authors available anywhere. We hope that our readers will avail themselves of this once in a lifetime opportunity to acquire the first volume of what is to be a series based upon the SoAR website. These will be neither picture books nor potboilers but well-illustrated histories of what once was one of the world's great systems: the South African Railways & Harbours.

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 authors.

Acknowledgments

In terms of transport logistics the George - Knysna line was not at the forefront of SAR's priorities. But as a tourist railway it fitted the bill. Some far-seeing officials in head office took notice of SATOUR's surveys of visitors to South Africa in the 1970s and '80s which consistently showed that >15% of tourists came to ride our trains and, as time went on, to admire our steam locomotives. As a result, the line was chosen to be the museum's railway along which some of the priceless treasures in its care could be displayed at work in their natural environment. This brought about an unprecedented surge of visitors to the region which continued into the new millenium, only to be abruptly ended by the floods of 2006. Some of the visitors during that fabulous period have kindly sent examples of their work to be exhibited on "Soul of A Railway". As you will see, we have reason to be grateful. As usual our thanks are expressed in alphabetical order of surname (please note: we got so many contributions that the chapter has been split into two, so some of these acknowledgements only apply to the return journey).

Eugene Armer, Franciscus Boshoff, Bill Botkin, Bruce Brinkman, Alan Buttrum, John Carter, Eric Conradie, Andrew Deacon, Org de Bruin, DRISA, Allen Duff, Johannes Haarhoff, Geoff Hall, John Hunt, Chris Jeffery, Allen Jorgensen, Bob Koch, Justin Lewis, Melly Lewis, Marie Lilienveld and the George Museum, Alfred Luft (custodian of the late Harald Nave's photographs), Dick Manton, Bruno Martin, Yolanda Meyer and the THL collection, Peter Micenko, John Middleton, Leith Paxton, Ashley Peter, Les Pivnic, David Rodgers (via Julie), Peter Rogers, Peter Stow, Paul Stratford, Mike Tyack.

Introduction

In July 2006 the Knysna Branch was washed away in several places. The damage was severe but not so bad that it could not have been repaired by laying out some government money. However, at this time practically every SAR branch line had lost its traffic to road hauliers, so it was an obvious excuse for Transnet, the inheritor of SAR, to close it. But in the case of the Knysna branch there were sound reasons for keeping it open, not so much for freight (although this is still potentially worthwhile) but for a commodity that practically keeps the Swiss Railways in business: TOURISM. Trains and scenery are prerequisites for a rail-based tourist operation; as you will see, the branch traverses some of the finest land- and seascapes using some of the most spectacular contemporary engineering works on the SAR. Before the washaways the line was visited by tourists, holidaymakers and railway enthusiasts from all over the world. In most civilised countries restoration of this railway and building on its past successes would be a no-brainer, but it seems the powers in our present regime are not prepared to think outside the box......

So please sit back, relax and enjoy your journey to Knysna - it will not be dull. And we have a huge bonus. So many have sent us their photos that we have divided this chapter again: Part 1: George to Knysna; Part 2: the return journey from Knysna to George and the Calitzdorp branch; Part 3: the remaining Cape Midland branches.

As usual, Bruno's map is both a geography and a history lesson.

"Where Every Prospect Pleases" ought to be the clarion call for the whole railway to Knysna (although perhaps in the context of today, the next line ought to be included as well: "and only man is vile"). There was a time when tourism was an important part of SAR's business too, as illustrated by the charming poster. The Knysna branch in general, and Wilderness in particular, were considered sufficiently important to be included in the "Round-in-Nine" and "Round in Fourteen" tours that were a feature of the railway's passenger service from the 20s to the early 1950s - one of them is illustrated in the next picture.

This one came to us too late to fit into the numbering system but it is too good to leave out. Lit by the photographer himself, it is a pre-dawn study of class 24 No 3622 standing in platform 3 at George, ready to take 501-down T&P (the daily SuX mixed) to Knysna on 22 July 1980. This was when Bob made the arrangements for photo 53.

1. A quiet moment in well-kept George station. The ladies are likely waiting for the 11:55 mixed to Knysna.

2. While at George we'll duck down to the loco, and what more appropriate engine to find there than class 24 No 3669. We are fortunate to be able to show you Leith's portrait of driver Stephen's regular engine (for more than 20 years). A bit like the Oirish draught horse that stopped automatically outside every pub, 3669 knew all the stops along the road to Knysna - her driver just needed to crack open the regulator and let the engine do the rest.

3. Although it seems likely, I cannot confirm that this was a Saturday. If it were, 9-down Cape Town-PE mail shown pulling into platform 1 at George would have included a through coach to Knysna which would be shunted onto the branch train standing in Platform 3. This once-weekly service was withdrawn barely a month after the picture was made in May 1962.

If it was arriving on time it would be 12:22. The passengers on the other platform would be waiting for 9-down's opposite number, 8-up, bound for Cape Town, due in at 12:41. Meanwhile in platform 3 the Knysna train with the inevitable #3669 in charge would likely be collecting passengers off both trains and departing at 13:20.

4. The Mossel Bay express departed more or less simultaneously with the 08:35 Knysna mixed on a sunny morning in February 1968. "Mixed" was a euphemism for what the latter really was according to SAR's public timetable, which told us: "passenger accommodation is provided on the following goods trains when run". Note the throwaway "when run".

With apologies to Dale Carnegie, when it came to trains lower down the pecking order, SAR could have written a best-seller titled "How to Run Trains and Alienate People"

5. Probably with Driver Stephens in charge, this was class 24 No 3669 making a typical leisurely exit from George in July 1958. Not much energy was required for it was all downhill for the next 9 miles to Wilderness and the schedule allowed 2hrs 25 minutes for the 42 miles to Knysna. In case you're wondering whether 8-up was late, the answer is no, that GEA on the left was hauling its 2nd section, 888-up, on which I was travelling with my family in July 1958. Nowadays it seems incredible that such a slow train could be run in two, sometimes three sections, at school term beginnings and endings and during the Christmas holiday season. How SAR achieved this was with careful co-ordination and planning. Imagine all the extra linen and bedding, the extra dining saloons and kitchen cars, all scrupulously clean, and the extra staff - a feat of organisation inconceivable to those pretending to run our railway today.

6. After descending through the ever-steeper and wilder Meulrivier gorge the railway arrives at Victoria Bay then threads two tunnels, the other side of which is this seldom photographed view of the line from the Kaaimansrivier mouth.

7. Twenty minutes after leaving George the mixed plunges into those two rough-hewn tunnels through coastal cliff buttresses then creeps onto the Kaaimansrivier bridge, the dramatically situated structure portrayed in Org's photo.

8. At the end of Les's third North Coast chapter (see the Soul of A Railway index) we announced that the next one would be on the Knysna Branch. This resulted in a flood of photographs from all over the world, but the most amazing thing was that more than 100 were of the Kaaimansrivier bridge. We are fortunate to be able to introduce the story of its construction with Eugene Armer's magnificent photograph. As you can see there was not much traffic, but that's another story...... Let Eugene describe his photo for you:

"Soon after sunrise on a summer morning the bed of the Kaaimansrivier mouth at low tide was the place to see the 05:00 goods from George to Knysna crossing the famous bridge. On 11th December 2001, 19D 2683 was silhouetted by its own steam as it trundled across the bridge with just one empty type ST 'toast rack' wagon in tow. The wagon was dropped off at Duiwerivier for loading and 2683 continued light engine to Knysna, returning to George with the 09:45 Outeniqua Choo-Tjoe."

The undoubted hero of the bridge's construction was the Italian contractor, Vallero, whose last words upon completing the job were: “I frankly did not realise how awkward working in such a harmless looking seaway could be and proved to be.”

They are quoted to encourage you to read the following account of its construction that first appeared in the November 2019 Journal of the South African Institution of Civil Engineers. We are grateful to its author, Dr Johannes Haarhoff, and the SAICE who have granted permission to publish the article here. Based on comprehensive researches by Johannes (you may remember a description of his hike through Toorwaterpoort in System 3; Part 19) and SoAR contributor and historian Allen Duff, the masterful narrative tells of a construction fraught with problems from the beginning, all of which were overcome thanks mainly to the larger-than-life engineer-in-charge: Nicholas Kingswell Prettejohn.

9. An Aerial view of the completed Kaaimansrivier Bridge, bracketed by the two railway tunnels on the George side and a third tunnel on the Knysna side. Also visible is the lookout on the N2 at Dolphin Point.

Preamble

The railway between George and Knysna traverses some of the most scenic terrain in South Africa, fully deserving labels such as the “Lake District”, “Garden Route” or “Eden”. The running of the highly popular “Outeniqua Choo-Tjoe” tourist train on this line provided a thrilling glimpse of the area — now sadly retired after heavy rain, catastrophic landslides and washaways blocked the line in July 2006. The Kaaimansrivier Bridge is the icon of the George-Knysna line. It is viewed on a daily basis by many travellers from the fine lookout at Dolphin Point on the N2 immediately above the bridge. Ninety years after being put into service on 30 November 1928, the story of its construction deserves to be told once more.

The construction of the George-Knysna line was preceded by about 30 years of planning and preparation. The benefits of such a line were first mooted in the 1890s. Shortly after the Anglo-Boer War, a flying survey was conducted in 1904, but Unification in 1910 and World War One during 1914-1918 stunted railway development for many years. In 1921 the SAR&H Resident Engineer based at George, Herbert G Dempster, made a more detailed survey of the route with a cost estimate. Parliament approved his proposal by promulgation of Act 30 of 1922, and budgetary approval followed in 1924. Act 30 of 1922 ambitiously authorised the construction of 18 new lines throughout South Africa, totalling 1 372 km, which obviously could not be tackled all at once. By October 1923, the first 171 km had been opened for traffic. By the end of October 1924, a further 426 km had been opened. The George-Knysna line had to wait until 1925 before construction could commence.

10. Resident Engineer Herbert Dempster, who did the first detailed survey of the route between George and Knysna.* A stalwart of the CCE's construction department, Mr Dempster went on pension upon reaching the age of 60 in 1924 and passed away in 1954 at the age of 90, having enjoyed 30 years of retirement. His widow survived him by several years.

* Mr Dempster was also responsible for much of the survey work on the Cape-Natal and Donnybrook-Underberg branches - see SoAR Chapter 6, Part 9.

Construction Commences

In January 1925, the South African Railways and Harbours (SAR&H) dispatched its engineers to start construction. At this time, the Chief Civil Engineer was Robert Catherwood Wallace, assisted by Theodore Heinrich Watermeyer as Assistant Chief Civil Engineer, both based in Johannesburg. They picked Nicholas Kingswell Prettejohn, who was credited at his death as “the Engineer who had built more railways in the Union than any Engineer then in service”, as Resident Engineer to oversee construction. He was a colourful character and had a brilliant construction record, with the completion of the Prieska-Kalkfontein railway line during World War One in record time possibly as the high point of his career. Prettejohn split his team to build the 68 km line from both ends. The eastern section from Knysna was put in charge of Engineer Townshend and the western section from George in charge of engineer TF Bromley. They were to be assisted by engineers Walton, WEA Milward, TV More, CJ Pybus (who resigned a few months later), AB Ferguson and SA Pethick, as well as surveyor Mickleburgh and clerks WR Townshend and G Keeling. Except for the first 8 km from George to the Kaaimansrivier Bridge, most of the work was done departmentally by the SAR&H, who also built all the tunnels after gaining considerable tunnelling experience on other projects throughout South Africa. The heavy earthworks and platelaying of the first 8 km were put out to tender. The platelaying contract went to the experienced contractor Cartwright who, 20 years earlier, had worked with the legendary railway engineer George Pauling on the Sir Lowry’s Pass and Riversdale-Mossel Bay lines.

11. N K Prettejohn is the large man in the white jacket. The other gentlemen are unidentified but since this photo was taken at Goukamma the one immediately to his right could be the clerk Townshend who was in charge of administration at the eastern end of the line.

Prettejohn’s first step was to peg out the precise route between George and the Kaaimansrivier — reporting that at some points it was “almost impossible to get a foothold along these krantzes” and “at present only a footpath has been made”. The survey was further slowed down by the extensive bush-cutting required to estimate the earthworks quantities, but was eventually completed on 19 February 1925. Prettejohn now proposed to move Dempster’s route to a lower level that would shorten the line by 540 m, reduce the earthworks and also move the railway line on the eastern side of the Kaaimansrivier further away from the motor road that was being built at the same time. A second survey team had also started at Swartvlei and progressed at twice the speed — by 19 February they had already completed 16 km up to Goukamma.

12. In its descent from George the route clings to the left bank of the Meulrivier, which gets increasingly precipitous before discharging into the Indian Ocean a mile before Victoria Bay.

With the new route between George and the Kaaimansrivier approved by the Chief Civil Engineer, its completion was a high priority. The existing wagon route from George to Wilderness allowed the transport of lighter material, but the heavier equipment and materials required for the Kaaimansrivier Bridge would have to wait for completion of the difficult section of railway from George. From George the line dropped about 215 m over a distance of 12.9 km along the left bank of the Meulrivier at an average slope of 1/60, all the way to Victoria Bay, requiring a river diversion through a deep, wide rock cutting 80 m long and numerous embankments [this diversion of the river would be a major problem during the 100-year flood of 2006]. From Victoria Bay, more heavy rock cuttings with two tunnels through treacherous cliffs high above the sea had to be completed before the Kaaimansrivier was reached. By 19 February, a contract for the first 3.8 km already had 250 men at work, which would soon rise to a total of 400 to 500 men working on the George-Wilderness section.

13. The east portal of tunnel 1. Tracklaying between the first two tunnels.

14. The track has been laid and the trains are running - the same viewpoint 50 years later!

16 (Figure 2). Mr Cairncross's drilling rig.

On the weekend of 20 September 1925 the area “was visited by one of the heaviest rainfalls that has occurred in this district for many years” and the Kaaimansrivier came down in full flood. A deep channel was washed out on one side of the river bed and rock slips on the Wilderness side of the bridge stopped all construction. The new road alongside the railway line was badly damaged and impassable for vehicle traffic due to large rock slips. It was anticipated that the Divisional Council would need a month or six weeks to repair the road. Most importantly, the experienced eye of Prettejohn alerted him to a danger previously unforeseen — an observation that had consequences for the later construction of the permanent bridge, as will be discussed shortly.

Figure 3. The geo-technical profile on the centreline of the permanent bridge as provided in the tender documents based on the Cairncross investigation.

Design of the Permanent Bridge

On 14 May 1925, Prettejohn collated all the available geotechnical information at the bridge site and submitted it to the Chief Civil Engineer, with the following comment in his covering letter: “It would appear on account of the foundations that the bridge will require to be founded on concrete cylinders.”

Prettejohn made an additional suggestion which is relevant today, as the survival of the bridge might very well depend on its tourism potential: “Whatever design is approved of, would you entertain the suggestion of making the structure of sufficient width to permit the bridge on completion being used as a halt, with a stairway on either side down to the sea shore?”

On 3 July 1925, the SAR&H Bridge Engineer in Johannesburg presented a cost comparison of three different bridge types. A bridge with 15 m long concrete arches (no price provided) or a bridge with 12 m concrete beams on concrete piers (£25 000) was found to be more expensive than a bridge with four 12 m spans and twelve 9 m spans, all spans for the latter

recovered from earlier steel bridge up-grades (£16 000). Throughout its lifetime, the SAR&H managed an ongoing main line upgrade programme to accommodate heavier and faster main line trains, which usually involved the replacement of the old steel spans. Putting these redundant spans to work on lighter branch lines had obvious benefits. The Kaaimansrivier Bridge deck would therefore employ used, recovered steel spans.

The twelve 9 m steel spans used at the Kaaimansrivier were previously in use on the Natal South Coast, completed in 1901. The remaining four 12 m spans came from the bridge over the Baakens River in Port Elizabeth, originally built in 1892. All the spans were sent to Assistant Bridge Engineer B Antliff at the Bridge Depot at De Aar, where they were refurbished for use at the Kaaimansrivier. The spans had to be adapted to allow for a footway and hand railings — the latter of expensive boiler tubing, because wire would perish quickly in the harsh marine environment. The unusual layout of the bridge with two curves separated by a straight section complicated the allowance to be made for the cant of the line on the bridge. The cant was to be taken up by 260 wooden sleepers, each sleeper cut with a precise, unique taper which necessitated the numbering of the sleepers. Moreover, every third sleeper had to be 0.9 m longer

to support the walkway; the sleepers between had to have the standard length of 2.1 m. The sleepers, with bolts of varying lengths, were cut to their exact dimensions at the Bridge Depot in De Aar. The design team next turned to the bridge foundations.

After considering the cost implications of five different foundation types, the Chief Civil Engineer proposed on 22 September 1925 a complicated combination of concrete caissons down to three metres below sand level, which would then rest on piles below. He qualified this proposal by asking Prettejohn whether such a complicated proposal would be practical to build. Fortuitously, this letter was written at exactly the same time when Prettejohn was watching the Kaaimansrivier coming down in an angry mood, referred to earlier. He made the key observation that the river bed was easily washed out to a depth of about 7 m, with the point of washout moving from side to side at will. Prettejohn responded promptly, “strongly recommending” concrete caissons all the way down to bedrock as he was “of an opinion [that] the nature of the bed of this river was too treacherous for any type of floating foundation”. Although the all-caisson option would be more expensive, the Chief Civil Engineer withdrew his proposal and accepted Prettejohn’s suggestion, shown in Figure 3.

The next question was whether to put the permanent bridge out to tender, or to do the work departmentally. The Chief Civil Engineer leaned towards departmental construction (probably

because of the unpredictability of the foundation conditions), but Prettejohn argued successfully for a private contractor. Contractors could procure special equipment quicker than the

SAR&H. Furthermore, the payment of staff working under exceptional conditions, subject to SAR&H regulations, would be “almost impossible to carry out.” By now it was December 1925 and the Bridge Engineer in Johannesburg proceeded to prepare the tender documents. On 15 January 1926, Prettejohn received a draft set of documents for his review and possible

amendments. On 15 February 1926, five tenders were received. The tenderers were EG Dowse & Co from Johannesburg; R Giletti from Redhill, Durban; D Mackenzie from Johannesburg;

Roberts & Olivier from Oudtshoorn; and MD Vallero from Durban. Vallero’s tender, the cheapest at £18 228, was approved by Sir William Hoy, SAR&H General Manager, on 10 March 1926. The contract starting date was 9 April 1926 with completion scheduled for 15 June 1927 — a construction period of less than 15 months for the completion of both temporary and permanent bridges. The contract was structured such that the SAR&H supplied all materials, while Vallero had to supply all labour and equipment. The SAR&H budget for the entire Kaaimansrivier Bridge project was about £38 000, of which Vallero finally got about £20 000. The rest went towards Vallero’s materials, the refurbishing of the 16 reused bridge spans and the erection of the bridge superstructure.

The Temporary Bridge

During April 1926, the bridge contractor arrived on site from Durban. The company was owned by Michele Domenico Vallero, born at Valperga north of Turin, Italy, on 28 May 1887. As a young man, Vallero worked in the USA and the Cape Colony before returning to Italy to marry. He was back in South Africa by 1913 and worked on contracts in the Cape Province and Natal before arriving at Kaaimansrivier. Vallero’s foreman on site was another Italian, Bapiste Giletti, who had previous experience of concreting. Soon after Vallero’s arrival, Prettejohn made a few changes to the temporary bridge as specified in the contract, which did not affect the contractual completion date. The temporary bridge had to be moved from the upstream to the downstream side of the permanent bridge and the level of the track on the temporary bridge was raised from 4.6 m to 7.9 m above the sand level. The last point was queried by head office, as the tidal variation allowed for seemed to be excessive compared to measurements made at Knysna harbour. Prettejohn explained that he based his proposed level after close observation of where driftwood and other ocean debris were deposited on the riverbanks. Moreover, he took the wave action into account which he observed at the site during the previous 15 months, which was much more severe than anticipated. Fortunately, Prettejohn’s practical experience won the day — Figure 7 shows the severe wave action during storm conditions.

17 (Figure 4). Start of construction at the George end of the bridge showing concrete footings of the first wooden supports for the temporary bridge and foundations of piers 1 and 2 of the permanent bridge around May 1926.

18. (Figure 5). Work on the temporary and permanent bridges continuing in tandem in June 1926 with piers 1 and 2 in the foreground. Note the pile-driving rig for the temporary bridge, cantilevered out from the previously-driven supports.

19. View from the south side showing pile-driving for the temporary bridge in progress.

20 (Figure 6). Construction activity on the temporary bridge during September 1926 while caissons for the permanent bridge were under way up to pier 8.

21 (Figure 7). The temporary bridge nearing completion during September or October 1926 with work suspended due to rough seas.

The construction of the temporary bridge, designed to carry a 150 tonne train, is illustrated in Figures 4, 5 and 6. The bridge had 36 timber spans of 4.6 m. The first eight trestles on the

George side and the first four on the Knysna side were supported by regular concrete foundations; the rest rested on wooden piles, each 15 m long. The piles were driven from the George end by a 3½ tonne movable steam-driven rig which cantilevered out from the previous support, through the sand to bedrock or refusal.

In June 1926, as the temporary bridge construction got under way, Engineer Bromley who was in charge of the western section, was transferred to act as Resident Engineer of the Klerksdorp-Maquassi-Vermaas line then under construction. Bromley’s position on the George-Knysna line was filled during July 1925 by Engineer RA Glenday, relieved from working on the foundation problem at the Durban grain elevator. Glenday was to supervise the bridge construction, temporary and permanent, from his arrival up to the end of the contract.

The work on both the temporary and permanent bridges did not go as planned. Instead of having the temporary bridge completed by the end of August, Vallero was still working on it during September, when work had to be stopped from 16 to 23 September due to heavy rains and flooding. In October, all work was halted again for a full week by foul weather.

Prettejohn was not pleased with the slow progress. By the start of October 1925, he formally voiced his concern in a report to the Chief Civil Engineer, but still held out hope that, once

problems with equipment breakdowns and a boiler inspection were ironed out, “there should be no reason why a satisfactory monthly progress could not be obtained in future”, also adding that he regrets to state that “Contractor Vallero’s general organisation on this bridge leaves much to be desired.” A month later, after admitting to the lost week due to foul weather, Prettejohn’s next report to the Chief Civil Engineer was worded in no uncertain terms:

“In connection with the work generally, I desire to bring to your notice that in my opinion Contractor Vallero will not have half of this bridge completed by the date of expiry of his contract, viz 15 June 1927. The cause I firstly attribute to lack of organisation, constant changes of the method of carrying out the work, understaffed in the way of artisans, and a disinclination to lay out money on plant. Further, in my opinion, the job has proved a bigger piece of work than Contractor Vallero first anticipated. He has been given every assistance possible from the staff on the works, but unfortunately he changes his mind and method of procedure almost daily.”

Vallero, in turn, blamed the poor progress on adverse weather conditions. He pointed to his good progress on driving six 15.8 m piles in a week when the weather was good. He described the weather conditions in a vivid way to Prettejohn: “You may imagine what can be done when the force of the wind is such as to be able to shift the piling frame bodily over 3’6” and at a later date, in spite of all precautions taken, the same thing happened. As the wind in this particular spot is a daily nuisance, with very sudden variations of direction and strength, men on this structure cannot effectively do any work with facility and confidence ...”

Work on the temporary bridge, which was scheduled for completion by the end of September 1926, dragged on until the middle of February 1927. Prettejohn also had second thoughts about the design of the temporary bridge and brought his own SAR&H gang onto the structure to add additional bracing not originally allowed for. As soon as the temporary bridge was

completed, it was put to work — first by construction trains, and from July 1928 for carrying regular trains operating between George and Knysna.

22 (Figure 8). Construction train on the newly completed temporary bridge in November 1927. The locomotive is an ex Natal Government Railways class A 4-8-2T.

23. The completed temporary bridge looked much sounder than in those earlier views.

24. Taken some 50 years after it was constructed, the late Harald Navé's study of the 'Outeniqua Tjoe Choo' (wish they'd thought of a more dignified name) emerging from Tunnel 3 is the nearest we could get to one close to the time of its construction. Who's complaining?

25. An extremely rare picture of a normal service train stepping gingerly onto the frail-looking structure some time after July 1928 - one can almost hear the timbers creaking. By the late twenties the 7th class would already be unfashionable but the lady passenger in her cloche hat was clearly up with the times. Incredibly, the 'temporary bridge' had been called into public service some months before the new bridge was opened - nowadays such enterprise would most likely be vetoed by the Rail Safety Regulator.

Construction of the Permanent Bridge

The Kaaimansrivier Bridge has an overall length of 158 m, divided over 12 spans of 9.1 m and four spans of 12.2 m. From the George end, there are seven short spans, then four longer spans, and finally another five short spans, supported on piers numbered from Pier 1 to Pier 15. The track enters the bridge, again from the George side, on a curve with radius 100 m up to Pier 4, followed by a transition curve to Pier 8, a straight section to Pier 10, a second transition curve, and finally leaving the bridge on a curve with radius 100 m starting at Pier 12 (Figure 9). The archives do not elucidate why different span lengths and a combination of straight and curved sections were selected — it was probably dictated by the availability of the reusable steel spans. The level of the track is 12.5 m above high water ordinary spring tide, compared to the 7.9 m of the temporary bridge.

Figure 9. The layout of the Kaaimansrivier railway bridge.

The foundations of the 15 bridge piers were at the heart of the construction. Piers 1, 2 and 15, being adjacent to the abutments where the bedrock was close to the surface, were founded on conventional foundations cast into excavations down to bedrock. The remaining 12 piers had to be supported by caissons which were to be sunk to bedrock. Figure 10 shows how much of the bridge construction had to do with foundations invisible to the present day observer.

Figure 10. Diagrammatic representation of the four central piers to show the size of the caissons in proportion to the piers.

Work on the permanent bridge started as soon as Vallero arrived on site in April 1926. The construction programme of the permanent bridge slipped from the beginning. The progress on the permanent bridge predictably suffered while Vallero was simultaneously working on the temporary bridge, as he directed more resources to the temporary bridge. The completion of the temporary bridge, however, did not improve the rate of progress on the permanent bridge. We find the following scribbled comment by the Bridge Engineer on Prettejohn’s progress report for November 1926:

“When the weather for a period has been less difficult, failing a commensurate improvement in his progress it looks as though it would be necessary to take the work away from Vallero and do it departmentally.”

By February 1927, Prettejohn provided a detailed report on the lack of progress to the Chief Civil Engineer. He admitted to difficult soil conditions that slowed down caisson penetration and was satisfied with the quality of Vallero’s work, but not happy about the speed at which it is done which he blamed on Vallero’s inexperience with the type of work. However, he did not think it would be possible to dismiss Valero in terms of the contract before the completion date. The Chief Civil Engineer concurred, stating that “.. all we can do is to warn Vallero that we

will take over the work if the contract is not complete by 15 June [1927].”

Vallero’s slow progress was not only due to inexperience and poor organisation, as claimed by the SAR&H officials. He had at least four real problems to contend with:

> Labour problems: The black labourers (with fixed contracts to Vallero) went on strike early on to claim the same pay as the white “poor relief workers” (paid by SAR&H on a piecework

basis on other parts of the project). The police had to be called in to remove the ringleaders before those remaining returned to work. Vallero also noted the frequent absenteeism on Mondays due to alcohol abuse throughout the contract.

> Bad seas. When the sea tide rose during periods of strong river flow heavy unforeseen swells up the river were encountered. This slowed down the work in many ways. The 15.8 m long wooden piles could not be floated to the temporary bridge, having to wait up to three days at a time for the waves to subside. Workers had to be withdrawn from the structures with workers swept off the structures (without injury) on two occasions.

On 4 February 1927, all temporary timber and shuttering were stripped and swept up the river, which took five days to be recovered and reassembled. Caissons freshly cast were damaged

while the concrete was green. Caissons 8 and 11 had to be demolished and constructed from scratch. The waves suspended so much sand in the water below the bridge that pumps and jetting equipment were clogged and rendered unusable. Vallero estimated that bad seas alone cost him three months of working days.

> Bad weather. The labourers used the slightest excuse of rain during winter to remain in their huts. On occasion only about ten workers turned up for work. To this was added severe cold and rain on a number of days when the work had to be stopped completely. Vallero wryly added that “peculiarly enough the seas were always at their quietest when it was raining or just after rain.”

> Hard clay. This slowed down the sinking of the caissons and was a serious hindrance to progress — a problem discussed in more detail under the next heading.

Three weeks before the completion date of 15 June 1927, Vallero officially asked for six extra months to complete the contract. He based the extension on the first three points listed in the previous paragraph. After visiting the site personally on 13 September 1927, the Chief Civil Engineer agreed to extend the completion date by nine months to 16 March 1928, with the proviso that his progress would be reviewed on 16 January 1928 and terminated if “good and rapid progress was not being made”. When the extended completion date in March 1928 was reached, the Chief Civil Engineer concurred with the Resident Engineer that “under the circumstances the best working arrangement will be to extend the contract date of completion from month to month”. By the start of 1928, the construction activity on the George-Knysna line started to wind down. Prettejohn went on leave from February for one or two months and engineer CW Ballenden was temporarily called up from the Wintersrush-Koopmansfontein line (between Kimberley and Postmasburg) to act on his behalf.

By 16 June 1928, three months after the extended completion date, it appears that all the caissons had been sunk to bedrock, except for Caissons 9 and 10. Vallero, however, remained on site to complete the entire contract. With everything complete, except only two remaining caissons already partially sunk, there probably would not be much to gain by ordering him off site as the SAR&H was legally entitled to do.

The Caisson Problem

The difficulty in sinking the concrete caissons of the permanent bridge to bedrock was undoubtedly the major cause of Vallero’s slow progress (and later financial difficulties, not addressed in this article). To Prettejohn, the integrity of the caissons was a key concern, aroused during the flood of September 1925 when a channel was scoured in the river deeper than anyone had anticipated. The base of the foundations had to be able to withstand scour right down to the base of the caissons. Prettejohn thus proposed to the Chief Civil Engineer that all caissons, once in position, should be filled with 1.5 m of concrete to provide a “concrete seal” before filling the rest of the caisson with sand. The contract did not specify the concrete seal, and this would entail extra cost. The Chief Civil Engineer concurred, adding additional precautions:

> Where the caisson is unevenly supported by rock or hard strata around its rim, the gaps should be packed with bagged concrete by a diver before the concrete seal is provided.

> If the caisson is evenly supported all round, then the concrete seal can be omitted to save money.

Figure 11. Cross-sections at the base of the cutting edge of the Caissons shown in plan in figure 12; the bevels start 1.3 m above the tip of the cutting edge (from a THL drawing).

Figure 12. A cross-section of the caissons showing the "dead spots" that contributed to the difficulty of sinking them (from a THL drawing).

Caisson construction, in its simplest form, entails the casting of a fairly thin 'ring' of concrete on the surface, then removing the material on the inside of the caisson. As the caisson settles, successive 'rings' are added as the caisson goes down. The process continues until the caisson encounters bedrock or when settlement ceases. Once the caisson reaches the desired position, excess concrete protruding at the top is trimmed off and discarded. To allow a caisson to cut its way more easily into underlying material the bottom of the caisson underside is bevelled towards the inside to produce a sharp cutting edge capped with heavy steel. The caisson design adopted for the Kaaimansrivier Bridge was more complicated than just a simple ring, shown in plan in Figure 12 and the bottom part in section in Figure 11. The problem at the Kaaimansrivier Bridge was that the caissons settled very slowly, or not at all, long before bedrock was reached. This problem was reported by Prettejohn as early as October 1926 when Caisson 4 was started (Caisson 3 was too shallow to cause the problem). After the “considerable trouble” encountered with Caisson 4, which was only 6 m deep, he correctly attributed the problem to two root causes. First, the cross-section of the caisson was not a conventional thin annular ring, but a rather complicated shape with limited clear space to operate the grab and, more importantly, with large “dead areas” under the caisson where material could not be reached by a grab. The angle of repose of the soil was such that it had to be excavated by grab to almost 2 m below the cutting edge before the “dead areas” would be cleared. Second, the material to be removed was a very hard mixture of sand, clay and rocks — not the “black mud and vegetable matter” as described in the tender document.

Vallero used the same arguments when he asked during May 1927 for extension of time. He considered the design inadequate — the steel cutting edge should have been deeper and the dead areas should have been reduced. About the nature of the soil, he wrote: “In tendering for the work I looked upon this, I venture to suggest, as any contractor or official of the Administration would have regarded it; something comparatively easy to sink through, a material such that when the caisson being say four feet undercut by grab, would fall into the open excavation under its own weight, ready for removal by the grab. Upon this point I was personally reassured by the District Engineer and the Inspector respectively who had seen the borings made and had seen the nature of the material, that sinking through this should prove the easiest portion of the work.”

Three months later, Vallero did not mince his words: “I did not and could not fully realise that ... the information given to me at the time of tendering by Administration officials could have been so misleading and erroneous.”

Prettejohn acknowledged that the geo-technical information available at tender stage was misleading to both the designers and the contractor: “I do not think that there can be any doubt that the whole of his [Vallero’s] trouble has been due to the design of the caissons being unsuited to the nature of the material he has been sinking through; or, going back a step further, to the erroneous information as to the nature of the material shown by the borings.”

The SAR&H engineers, at this point still smarting after the severe geo-technical problems encountered at the grain silos in Durban harbour shortly before (a story begging to be researched and told by a history-minded geo-technical engineer!), were particularly sensitive to more criticism about poor foundation design and shifted the blame to inadequate interpretation of the drilling and probing results: “This is another case in which the reports of borings have proved misleading. It would appear that men lent by the Irrigation Department, presumably accustomed to boring for water, do not sufficiently appreciate the necessity for accurately describing the strata pierced through, especially as regards degrees of hardness when the borings are for foundations.”

The hard soil was already noticed when many of the wooden piles for the temporary bridge refused at shallow depths. Started from the George end, Piers 1 and 2 did not require caissons

with concrete foundations cast directly on the shallow bedrock. Pier 3 had a short caisson which was taken to bedrock without a problem. From here the caissons increased in length up to Caisson 7. At about 9 m below the surface, their settlement stopped. Vallero hoped that the obstacle would be a layer only a few feet in depth and he installed a jetting plant to loosen the soil below the caisson. This helped him to get the caissons for Piers 4 and 5 to bedrock and those for Piers 6 and 7 partly down, but then no further. Vallero next tried three other methods — firstly by increasing the undercut with the grab to as much as 9 m below the cutting edge (no improvement), secondly by pumping the caissons dry and excavating below the caissons by hand (gaining only a few feet), and thirdly by increased loading by casting more concrete on the caisson until the top protruded 7.3 m above the surface (no improvement). Even when using all these methods in combination the caissons would not budge. In desperation Vallero tried a method, proposed by Prettejohn, which could roughly be described as a vertical rake which was maneuvered up and down with the crane with a set of bent-back teeth such that they could dislodge the material directly below the caisson, to no avail.

By the end of 1927, already some months into the extended contract, numerous caissons were still stuck at 9 m while bedrock was at an average of roughly 20 m. Vallero had no option but to turn to the costly and controversial method of using divers and explosives. The divers placed small detonating charges to the material below the caissons and the explosions would then fracture the hard soil to allow the caissons to settle for a short distance and the process would be repeated over and over. This was a tedious process. The caissons, moreover, were not designed to withstand the shock of the exploding charges, so to prevent damage to the concrete the charges were reduced and they were also fired one at a time. Despite this care the concrete at the base of the caissons was seriously shattered, partially ascribed to insufficient reinforcement in the bottom 1.2m of the caissons. After reaching the desired depth the caissons had to be properly seated by hand-chiselling to match the rock with the damaged bottom edges of the caissons. This was a particularly difficult task, described by Vallero: “I may quote you as an example Caisson 5 where a diver has worked 290 hours and it is not ready for sealing yet this caisson has been on rock since April last and had it been possible to seal on the shell layer as shown on the plan, a week would have sufficed to complete it. The same remarks apply to Caissons 6 and 7, the former has been on rock since May and the latter since the middle of July, both of these could have been completed and the piers finished off by now...”

The fortunes of Caisson 11 epitomised the problems caused by the hard rock. Caisson 11, with its total length of 22.3 m of which 19.2 m were below the surface, and despite the fact that Vallero had tried all the tricks described above, still had its cutting edge about 2.1 m from bedrock. A close inspection of the caisson base revealed that the concrete was shattered on all the sides of the caisson for a distance of 2.7 m from the base with numerous cracks everywhere. On the George side, a large chunk of concrete broke away over the bottom 1.5 m, fell and blocked the inside of the caisson and could not be moved without blasting. Furthermore, the concrete parted at the construction joint 2.7 m from the bottom, and the bottom part of the caisson slipped down the reinforcement, leaving a horizontal crack along the joint of about 75 mm. It was clearly impractical to try to sink the caisson further and it was agreed that a

load test of 600 tonnes was the only way out. If there were no movement after six days, the load could be removed and the caisson sealed with concrete, which was what was done.

Vallero soldiered on and by end July 1928 all the piers were completed to bridge girder level, except Piers 9 and 10. During the entire month of August, the total depth gained by both caissons was a paltry 3.0 m. Caisson 10 was sealed on 5 September, and Caisson 9 a month later on 8 October 1928. The construction of Pier 9 and the installation of the bridge deck by the SAR&H followed — finally bringing a successful end to a remarkable series of engineering difficulties (Figures 13 and 14). Moreover, despite the time delays, the entire Kaaimansrivier Bridge project came to £33 000, £5 000 less than budgeted for.

26 (Figure 13). The almost complete bridge presents a rather forlorn sight by about September 1928 with pier 10 completed and still struggling to sink pier 9 to bedrock.

27. Westbound out of Wilderness (i.e. heading for George) the line climbs on an easy gradient to tunnel 3, the western end of which is shown in photo 23.

Epilogue

The Kaaimansrivier Bridge entered official service on 30 November 1928 when the first scheduled train crossed the bridge. Up to the very end, the fickle weather played a part. A week before the opening, heavy rains were once more responsible for washaways on the line which prompted the Station Master at George to send the following telegraph to Johannesburg:

“Anticipate serious delay completion Kaaimans Bridge”. The washaways were presumably repaired in time, for the opening went ahead as planned.

The temporary bridge was broken up in early 1929, and here the tenacious soil conditions had the final say, as the extraction of the timber piles of the temporary bridge turned out to be more difficult than anticipated. A special pile extractor had to be supplied from Durban for a period of four to five weeks for this purpose.

Nicholas Kingswell Prettejohn left the Kaaimansrivier Bridge when he took long leave from August 1928 to January 1929. He retired on 1 August 1931 at the age of 59 and died in 1935.

Michele Domenico Vallero struggled to find contracts after the completion of the Kaaimansrivier Bridge (this was at the time of the worldwide Great Depression). He returned with his wife, two daughters and three sons to Italy in 1931 only to move to South Africa a final time in 1935 with two of his sons. He stayed in Pretoria where he started a brick-making business. Vallero died in Pretoria on 31 December 1950.

Chief Civil Engineer Robert Catherwood Wallace retired in 1928 shortly before the Kaaimansrivier Bridge was completed in 1928. In the following year, 1929, he served a term as President of SAICE. His erstwhile young Assistant Chief Civil Engineer Watermeyer, who was also closely involved with the bridge design during 1925, was SAICE President in 1926, went on to become SAR&H General Manager from 1933 to 1941, and died on 3 December 1948 in Johannesburg.

The Kaaimansrivier Bridge, today merely a fleeting glimpse to most motorists in a breathtakingly beautiful setting, deserves to be remembered as a monument to a previous generation of

never-say-die engineers. In the words of a letter by Vallero to Prettejohn dated 28 May 1927: “I frankly did not realise how awkward work in such a harmless-looking seaway could be and proved to be.”

28. What looks like a goods train, in which case it could have come from Knysna, about to cross the new Kaaimansrivier Bridge in either December 1928 or early 1929. The footwalk and handrails were yet to be fitted in this illustration from the 1929 General Manager's report.

Acknowledgments (Bridge article only)

The photographs came from two sources — some were made available by the DRISA volunteer project with the permission of the Transnet Heritage Library (THL) in Johannesburg, while others came from an album now held by the George Museum (GM), donated by Marie Lilienveld. Allen Duff of George had the privilege of interviewing MD Vallero’s fourth and last surviving child, Valeria Maria Ciocatto, in 2005, who provided his biographical details. Having studied the local history with emphasis on railway history for many years, Allen made all his research notes and publications available for this article, which he was kind enough to review before publication.

29. Judging by the articulated saloons this could be the same "Round in Fourteen" tour illustrated in photo 1.

30. If contractor Vallero had had weather like this for the duration of the work it would probably have been completed on time, or maybe not, considering the problems encountered while sinking the caissons (foundations for the piers) to bedrock. Seventh class haulage in this fine THL photograph probably dates it to before 1948 when the class 24s were introduced.

31. The photo was sent by Leith Paxton via DRISA and the details are from SoAR's resident historian, Les Pivnic: "[This] was one of the standard railcars built in Durban Mechanical Shops between 1929 and 1932, designed by A G Watson. They were powered by two American Hudson (motor car) Super 6 petrol engines and the driver could cut out one engine on the flatter sections. There were actually 15 cars built between 1929 and 1932 and they proved to be entirely successful. The last car to survive was a hybrid version with half bus/half goods wagon operating on the Kootjieskolk – Sakrivier section in the Cape. She was taken out of service in 1967. An additional two cars to the same design were built in Durban Shops for the CFM* in 1935 and delivered to Lourenço Marques [some of Les's information came from an article in the SAR & H Magazine for March 1931]. "

* the Mocambique Railways

Les also provided this extract from the 1936 timetable. The columns headed 'G' denote the railcar service. We don't have the date when the railcar service was terminated but it doesn't appear in the 1939 timetable. Instead, a much truncated and not nearly as convenient Road Motor Service ran once/day; three times/week between George and Knysna via Thomas Bain's Three Passes route, the original road between the two towns which avoided all the stations along the railway, including Wilderness.

32. Charter re-creations by the museum working with the Publicity and Travel Department could be almost authentic - as with 7A 1007 trailing three restored coaches in Imperial Brown. The second and third coaches are both ex CSAR; the third, SAR #6006 is described in full by Peter Stow in the caption to photo 24 in Part 1 of our history of the Port Elizabeth Suburban Service, and the middle one, even more historical, in his caption to photo 23 of the same chapter. For a close idea of how these coaches looked in brown please look up John Middleton's photo 34 in Part 2 of this chapter (soon to come). Here follows Peter Stow's pen history of the first coach:

"The coach nearest the engine represents an interesting phase in the history of SAR passenger stock. The SAR was concerned that the use of old main line coaches on branch lines was not a satisfactory solution and decided to build new vehicles for that purpose. The first 10 such 1st- and 2nd-class composites of type M-24 were built in Bloemfontein in 1927/8 for the Free State branch lines and were such a success that many more were built; 20 type M-33 in Bloemfontein in 1931/2 and 26 type M-36 in Pretoria in 1932/3. Although these vehicles had compartments, there were no sleeping facilities and they became known as 'Day' saloons. The emphasis in design of the last 2 batches was for low mass for branch lines and the underframes were of a lightweight design, resulting in them often being referred to as “Depression” underframes by railway technical staff given the period in which they were built but this was not the reason for their lightweight design. Some of these coaches were allocated to Port Elizabeth for use on the George – Knysna and Port Elizabeth – Kirkwood branches but they could be found on such trains as those operating from Durban to Port Shepstone and Mooi River; Cape Town to Worcester; Dover to Parys; Johannesburg to Klerksdorp and Pretoria to Pietersburg, among others. The example illustrated is one of the type M-36 which spent much of its life in Cape Town before being withdrawn from scheduled services."

33. The intricate geometry of the bridge as described by Johannes is evident in this view from Dolphin Point which shows the 100 m radius super-elevated curves at both ends, each approached by transitions. Each individual sleeper had to be cut with the appropriate inclination to provide the correct cant at that point. The photo, taken on Christmas Eve 1972 depicts the longest and heaviest-looking train I ever saw with one locomotive on the Knysna branch. The six main-line saloons were for a party of children on holiday, I think at Sedgefield. No Briggs, those six saloons didn't go through to Knysna. The 24s were good strong locomotives but not strong enough to get out of Goukamma with this load.

34. A David Rodgers special with preserved 7A 1007 in June 2006, barely a month before the great flood. But for the livery of the coaches this photo could have been made in the thirties.

35. Just looking gives one vertigo. Luckily for us, Driver Geoff chose to ride the cushions one November morning in 1988 giving us this memorable view from his window seat on the 06:50 George-Knysna mixed.

36. Before the Road Transportation Act of 1977 the Knysna branch carried plenty of traffic, frequently requiring combined working which in those rule-abiding days called for locomotives to be separated by wagons or coaches to spread the load on the bridges. This 'dupla' was surprisingly late in the day - February 1984. By the end of the decade nobody cared anymore.

37. During 1980 SAR Travel organised several round South Africa Safari Tours using locos still in service at that time. After an early ascent of Montagu pass, three immaculate Class 24's 3670/22/27 approaching Wilderness with a return trip on the Knysna line from George in October 1980. That's tunnel 2 in the background, on the far side of the Kaaimans estuary.

38. "Where Every Prospect Pleases". The fledgling seaside resort of Wilderness was beginning to flourish when this class 8 approached with an interesting mix of general freight, c 1948.

39. The "Sunset Limited" making a slow and dignified entrance into Wilderness, home dorp of "die Groot Krokodil" [the big crocodile] South Africa's Prime Minister at the time: P W Botha. Note the welcoming band. This was in April 1979 when SAR ran the first of several "farewell to steam" tours during the ensuing 25 years (yes, twenty-five!). They catered for steam lovers from all over the world but none was as slickly organised as the first one for the simple reason that locomotives and servicing facilities were still available in all corners of the SAR.

Planning and organisation was in the hands of the late Alan Clarke and Graham du Plessis of the RSSA and Mike Kruger of SAR handled the co-ordination with Operating.

40. The spray off the sea provides a background to the 08.45 George - Knysna Mixed, running as a passenger train on this occasion at Wilderness in July 1995. By this time the goods traffic on the branch had become seriously reduced. It would be interesting to speculate on when last the Wilderness SM had had a customer use his goods shed loop, not to mention the shed itself. Note the tidiness of the station, still immaculate despite 17 years of direct competition with subsidised road hauliers that have caused the goods siding to be bare and rusting.

41. Forty years earlier and everything on the SAR was rosy. The equivalent mixed, 501 down, was 08:30 off George and well-loaded with coaches full of passengers (at least this stayed the same until the end) and a healthy array of goods vehicles. The 3669 was but six years old and that could be driver Stephens. As usual, Peter Stow has come up with some reliable information about the rolling stock: "The vehicle with the chimney is an ex-NZASM 4-wheel coach converted for use as a non-revenue vehicle for such purposes as Railway Missions etc. The third to last vehicle is probably the last ex-NGR 36’6” passenger coach being used as a third class vehicle, having somehow escaped scrapping like most of its peers in the late ‘20’s and early 30’s and being an ideal light vehicle for this branch line."

42. View of the newly-constructed line looking from east to west towards Wilderness probably late in 1928. All three tunnels are visible as is the scar on the hillside that eventually would become the N2.

43. About as authentic a recreation as was possible in July 1998 was this charter mixed organised by Steam and Safaris crossing the Touw River at Wilderness with classes 7A 1007 and 7BS 1056. The same train is seen again in John Hunt's photo 58 and it returned from Knysna at sunrise the following morning, providing some memorable photos that will be seen in the next part of this chapter.

44. Class 24 No 3622 crossing the bridge at Fairy Knowe with 961-down T&P in July 1980.

45. Early on the morning of 15 December 2000, 19D 2649 reflected the first rays of the sun while crossing the Touw River at Wilderness with the 05:00 George to Knysna goods. Sad also to reflect that as the new millennium dawned the only goods traffic remaining on the branch was timber and that would soon be lost to subsidised road hauliers as well.

46. Wilderness lagoon, formed by the Touw river, is a favourite holiday boating water. It is crossed in two places by the railway, this one at Fairy Knowe taken from the Pirates Creek resort. The train was 501-down T&P; the engine was class 24 No 3658. Note the houseboat parked off on the left.

47. Traffic jam at the well-known combined road and rail bridge at Ebb-and-Flow. For once the vehicle on the road is older than the locomotives. A 1934 Chev Roadster belonging to Jack Boshoff of George waits impatiently for the "Kei Explorer" rail-tour of April 1989 to pass. Note the sparkling state of the Chevvy and of the locomotives; 24 class No's 3684 & 3627.

48. A "Round in Fourteen" tour train (not the same one as in the introductory photo) stopped at Duiwerivier on its outward journey to Knysna. c 1933 is my guess.

49. The same dupla as in photo 29 coming around the corner from Duiwerivier in February 1984. The passenger carriages look well filled.

50. Dick describes #3622 "Rosie" as the cleanest engine he ever saw, but I think driver Stephen's #3669 would have given her good competition - see photos 76 and 77.

51. The 08:35 George-Knysna mixed, 501-down (officially a T&P), departing from Rondevlei with class 24 No 3670 in December 1971.

52. Org was new into the game when he took this one at Swartvlei. When your editor asked him about it he said "Well I thought that was what was meant by steam photography....."

53. We asked Bob to send his famous Swartvlei photograph. Here it is and this is what he had to say:

"This is one of my better pictures from South Africa and it took four trips [all the way from Detroit Michigan] to get it right. Usually the weather wasn't co-operative, but on that morning of 21 July 1980 all the stars were lined up. We went to the station at George and talked to the driver of 24 class 3622 J.P. 'Tamatie' van Rensburg, a very friendly fellow, telling him what our plans were. I asked him to make a little smoke (but not too much!); get set up about half a mile before the bridge; hold a steady speed and don't touch controls until after the whole train has passed over the vlei (we were also going to take a movie of the run by).

We went straight to the spot and waited. As we were waiting mist began to slowly rise off the water - and I thought Oh No! not another botch up; but without any wind the mist just hung there. At 07:45 here he comes - everything goes well, and the images come out pretty good - a beautiful sight. In later visits I never got it as good as on that magic morning in 1980.

A lucky shot."

You can say that again Bob. It musta been Gail that brought you the luck.....

54. After a night of incessant rain in July 1981 the following morning dawned clear and still as the 24-hauled 08.30 from George crossed the Swartvlei en route to Knysna. Dick regrets not going for a wider reflection shot on this morning which would have involved a tree climb and considerable vegetation clearance, but your editor is content with an exquisite picture.

54. By the turn of the 20th Century all the 24s had been withdrawn and their duties entrusted to the ubiquitous 19Ds. On 15 December 2000 Eugene caught blue 19D 2649 in early morning light trundling across the Swartvlei with the 05:00 George - Kynsna goods and its by then usual consist of ST timber wagons.

55. Ideally the 24 should have had the coaches in photo 56 and the 7th class the coaches above.

56. The same train as in photo 34, this one by Allen. June 2006. The very next month the great flood took away most of the causeway over Swartvlei and a couple of spans of the bridge.

57. The 08:30 off George again, passing the aptly-named Lake Pleasant as it heads for the saddle between Ruigtevlei and Goukamma.

58. For information about this photo, made c 1938 (sorry - 1998) here is John's description: "One of the attractions of the latter-day special organisers was their ability to recreate fairly accurate scenes of yesteryear. Here 7A 1007 and 7BS 1056 catch the late afternoon light near Ruigtevlei, en route from George to Knysna, with a [charter] mixed train on 29 July 1998."

59. In case our readers are wondering where all these historical images lent by Allen Duff came from, here is his description of how he came upon them:

"A few years ago I was approached by a George resident, Marie Lilienveld, with a photo album which contained mainly photographs of the construction of the George/Knysna railway. In the 1970s Marie was appointed George's first Tourist Information Officer. She had an office on George station. One day she was approached by a railway clerk with the photo album. He

explained that it had been in the drawer of a desk 'for years and years', but thought that Marie should take over its safe-keeping. Concerned about its ongoing safe-keeping, Marie approached me. The album is now Album A38 in the archives of the George Museum. The photographs in the album demonstrate a professional and engineering eye. Their composition is balanced with interesting perspectives. They are certainly not snap-shots. The vantage points from which they are taken, show that the photographer had access to the railway's route no matter the then isolation of many of the localities. The clarity of the photographs indicates that they were taken with an expensive camera. My supposition is that they were taken by Bromley who was the engineer in charge of the western section of the George - Knysna railway's construction.

Included in the album are photographs of a narrow gauge locomotive/tractor hauling cocopans (see the next three photographs). From one of the photos, and from other photos in the archives of the Knysna Museum, the location can be identified as the Goukamma River valley. On the western ridge of the Goukamma Valley, a massive cutting was made through the shale with pick and shovel. The earth from this cutting was taken in cocopans down the valley to construct an embankment on either side of the road bridge and to the concrete abutment of the girder bridge over the Goukamma River. It would appear that gravity was used to run the cocopans down the valley and that the locomotive hauled them back up to the cutting, where there were loops for loading. In this way the cutting was excavated and the embankments constructed. Coming out of the eastern end of the cutting the line curved to the north-east on an embankment before following the contour downwards towards the river. Photographs show that when the tarred N2 was built in 1951 the track alignment was altered to accommodate the road through the same cutting."

60. Cocopans of earth removed from the cutting being tipped onto the formation of the railway on its twisting approach to the bridge over the Goukamma River. This and the next two photos were sent to John Middleton, who commented as follows:

"Two Baguley tractors were delivered to the SAR in 1925 : Baguley 1503 and 1504. The word 'tractor' was commonly used for small, narrow gauge diesels in the 1920s and 1930s : particularly in British spheres of influence. Baguley had a construction arrangement with Drewry which name probably appears on the locomotive's worksplate. Hence it was known as a Drewry Tractor."

61. The Drewry (Baguley) about to drag the empty cocopans back up to the cutting excavation. This is the embankment leading up to the right-hand aboutment of the Goukamma Bridge. The concrete bridge was over the original dirt road to the farms along the river but later it was used for many years by the N2. Note the flimsy-looking wooden trestle on the other side of the road bridge. This was the incredibly crude method of constructing railway embankments in those days. The cocopans were simply tipped over the void beneath them until it was filled, with no compaction whatsoever - thus railway embankments would continue to settle even after the line was opened to traffic, sometimes for months or even years.

62. A nice close-up of the Drewry (Baguley) and her driver. Again we are indepted to John Middleton for the information on the plates: "the top rectangular plate is a Hubert Davies plate (they were the importing agents), the lower oval plate reads 'The Drewry Car Co Ltd, London & Burton-on Trent. Makers. No. 1503 of 1925' ".

63. The same curve shown in photo 61 with the embankment leading up to the Goukamma bridge completed, as well as its concrete abutments. The steel through-truss span still to be installed. This was the only non deck-span bridge on the line. The Model T conversion was presumably Mr Prettejohn's inspection trolley.

64. A lengthy 503-down goods awaiting crossing at Goukamma in July 1973. From Ruigtevlei to Keytersnek single 24s were restricted to 270 tons on 44 axles, but this 64-axle train was comprised mostly of empty DZs which nevertheless would have induced a healthy grunt up to Keytersnek. The photo was taken from a higher level of the line, see the topo extract below.

From Goukamma to Keytersnek the railway climbs 255 feet in 4 miles. A straight-edge laid south-westward from Goukamma cuts the line five times before it heads away up the valley to Keytersnek (extract from the Sedgefield 1: 50000 topo map courtesy of the Chief Directorate: Surveys & Mapping, Mowbray, Cape Province).

65. Eugene liked this engine. Early on the morning of 12 December 2000, blue 19D 2649, surrounded by natural beauty in the heart of the Garden Route as it charged up the grade from Goukamma to Keytersnek with the 05:00 George - Knysna goods. [Note from Charlie: that 'natural beauty' didn't always encroach so closely; it was a sign that money was no longer being spent to keep the railway reserve clear of vegetation].

66. Dick's caption: "Class 24 No 3670 rounds the curve into the horseshoe above Goukamma hauling the 06.50 George to Knysna in July 1978. This section could be very slippery on a winter morning after rain and I once saw a similar working making four attempts to get past this spot. Little had I expected to get 4 photos of the same train at the same spot!"

67. John's description: "With trips to South Africa in 1972, 1976, 1978 and 1979, a visit to the George - Knysna line was a must with its super coastal scenery, regular service trains, delightfully photogenic stations, not to mention the pie shop at Sedgefield. Most of all, was the super shine condition in which the locomotives were kept at George, often with brass and copper embellishments and names. Here one such example, 3669 Noeline, makes a splendid sight climbing out of the valley from Goukamma on 27th June 1979."

68. The afternoon mixed from George is clearly doubling as a work train, given the number of workers riding the DZ wagons. The train probably stopped en-route to unload material before proceeding to Knysna. Here it is seen climbing up from Goukamma with long winter shadows already providing a dark background. July 1978.

69. As the Brinkman Special approached Keytersnek in July 1983 two days of unadulterated enjoyment were nearing the turn around at Knysna.

70. The seedlings have turned into saplings and by the time that banked mixed reaches Keytersnek they'll probably be full-grown pines.

71. The Sunset Limited was so long and the Knysna branch so twisty that not many successful photos show 24 class No's 3677+3669+3670 and the entire train. This April 1979 photo was made before newly-planted pines had obscured the line and just after the new Bridge Engineer had decided he didn't really worry too much if all three engines were coupled together.

72. Class 19B 1412 must have been thinking that she was back on Lootsberg Pass as she pounded up the grade from Goukamma with the 05:00 George - Knysna goods on the morning of 18 December 2001. That's the N2 highway in the background and the line is just visible on the opposite side of the valley, above the road and below the tree line on the right of the picture.

73. The same train seen in photo 55 crossing the last bridge before entering Knysna.

74. Dupla on the last bridge before Knysna on 24 August 1978, engines 3627 & 3670.

75. The following account of the arrival of the first train at Knysna was extracted from the Bulletin of the Historical Society of Port Elizabeth. It was written by the late Hans Huisman, erstwhile Harbour Engineer of Port Elizabeth:

"Upon arrival at Knysna promptly at noon on 17 October 1928, drivers Dowson and White brought their 7th classes under the flower-bedecked 'ALLES VAN DIE BESTE' banner. On board the train were the Minister of Railways, Charl Malan and Mr G. Wilcocks of the Railway Board. As the dignitaries and their wives stepped off the train they were greeted with loud cheers; flowers were presented to the ladies and buttonholes to the men. Buglers of the Knysna Secondary School cadets (under the Principal, Captain Hart) sounded a welcome and the troops presented arms. After the inspection of the guard of honour the scene shifted to the specially erected platform at the end of the station. On the dais were Mayor Foakes, the MPC Mr H van Huysteen, Mr C Thesen (son of Cpt Thesen of the Albatross who built the Knysna Forest Railway), Mr Brink MLA, Mr Mileham (Town Clerk), Mr Hill (System Manager) and Mr Prettejohn.

After the welcoming speeches the Minister replied and made special mention of the scenic beauty of the district. He exhorted the local industries and farmers to do their best to make the railway payable, suggesting the establishment of a publicity association and a high standard of hotel accommodation. He then declared the line open and hearty cheers from the crowd of about 700 terminated the proceedings at the station. The official party were the guests of honour at a luncheon given at the Royal Hotel."

76. Some 34 years later we found Driver Stephens giving #3669 the walk round in preparation for taking 504-up, the daily except Sundays 09:15 mixed, to George. May 1962.

77. Raking 3669's ashpan clean and the rest of the engine is already spotless.

78. Among the last remaining artifacts of the Knysna Forest Railway: the 2ft-gauge tracks on Thesen's Pier in Knysna. The railway operated from 1907 until 1949 and is well documented by Soul of A Railway's consultant historian Allen Duff. But the photographer is a complete ignoramus when it comes to ships and shipping (Briggs: that was not funny!) so he didn't realise he was recording a sad and historic event when this photo was made in December 1983. The tug, still in SAR&H livery, was the Alwyn Vincent. She had recently retired after of more than 20 years of service at Mossel Bay. Brian Ingpen, who has assisted us with all the chapters dealing with the H part of SAR&H, takes her story on from here:

"After operating in Mossel Bay for most of her career the Alwyn Vintcent – one of five Italian-built pilot tugs to be towed to South Africa in 1956 – came to Cape Town, whence, in a rather bizarre attempt to steam her to Australia, she entered Port Elizabeth where she remained for a long time. Via another protracted stay in Knysna, she returned to Cape Town to serve as a leisure vessel, carrying tourists around the harbour and into the roadstead. However, when her boiler did not pass the mandatory survey, her steam plant was decommissioned and replaced with a diesel engine. A national museum organisation took her over, but before long, she had become a sadly neglected hulk in Cape Town’s Waterfront. That a tug with more than 50 winters astern and the last steam vessel on the South African register should have been neglected in such a manner was scandalous.....A group of steam enthusiasts dismantled the old tug and took her inland to Villiersdorp where she was to have been restored, but as such projects are extremely costly, especially in a country where the true worth of shipping and its history are hardly recognised, funds simply were not forthcoming and the project stalled."

"In a rather ambitious project after powerful launches replaced the pilot tugs in South African harbours, R.A. Leigh, S.J. Harrison and Alwyn Vintcent were to steam to Australia to ferry trippers in one of the river mouths. Without much planning surrounding the wild sea conditions to which the three small tugs would be exposed on the long voyage to Australia, and crewed by a diverse group of people, the three tugs left Cape Town, encountered rough weather off the southern coast, and put into Port Elizabeth where the Australian odyssey was abandoned. Eventually, Alwyn Vintcent returned to Cape Town after a lengthy time in Knysna; S.J. Harrison went to scrap, and, under a new owner, R.A. Leigh went to Durban where she lay at various berths – including a spell at the Port Natal Maritime Museum, before she was dismantled in the port."

The above has been extracted from Brian Ingpen's forthcoming book that promises to be a best-seller: "Cape Town’s Dockland". Anyone interested in buying a copy should please get in touch with him via email: brian@lawhill.org

79. Almost ready for departure: Knysna and Plettenberg Bay mails being loaded into the van of 960-up T&P in July 1982. It is now almost 40 years since the mails last went by rail in South Africa. However, it can truthfully be said that the switch to air and road has not improved the mail service one bit.

80. The mail has been loaded, the passengers are all on board, so why are we waiting??

81. Were it not 15 years too late these two gracious ladies could have been at the front.....

82. Bob is American, from Detroit, Michigan. This is what he wrote about his picture: "Class 24 No 3622 'Rosie', the pride of driver J.P. 'Tamatie' van Rensburg, shunting the yard at Knysna after completing the run from George which included [my] classic Swartvlei shot [see photo 53]. This loco was in immaculate condition, as were all the locos on the Knysna line. It is with a sad heart to hear of the current state of this branch. A National treasure lost."

83. And here he is, the hero of photo 53, J P (Tamatie) van Rensburg himself (could his nickname perhaps be a reference to his complexion?).

84. Thanks to Bruce we can bring you the low-down on probably the last major development on the Knysna branch:

"Following the transfer of the Knysna Branch to the THF in the late 1990s and the reduction in freight traffic Transnet decided to redevelop their land in Knysna – this land was the whole area between Waterfront Drive and the lagoon. A new company was formed to develop the area and the Marina. The area between the station building and the lagoon containing the yard and turning triangle was alienated and developed into the Knysna Quays Centre (fancy name for yet another mall), Hotel, Holiday apartments and the new Yacht Harbour.

The platform of the old station was on the lagoon side of the building. With the redevelopment a new platform was constructed on the town side on what was the carpark. To compensate for the loss of the triangle a turntable was installed at the eastern end of the station adjacent to Waterfront Drive. This area and the extended running line that ran eastwards to the new transfer depot / waste depot was now part of the new carpark for the Knysna Quays Centre.

Unfortunately with the new layout and high level platforms the platform was about a metre higher than the floor level of the old station building resulting in a contorted layout – however the plus factor is that the old building was retained and a new extension added. Access from the road was now on the lagoon side with the trains on the town side.

A new yard was constructed at the western end between the station and the curve leading onto the bridge over the lagoon. A passing siding was also constructed at the eastern end between the station and the Transfer/ waste depot."

Having arrived in Knysna we'll check into the Royal Hotel and rise at cracka tomorrow for the return working.

The next segment of the Midland Branches will cover the return workings to George and the Calitzdorp branch. We are short of material on the latter, a sleepy but attractive branch line so there's still time to contribute should you feel so inclined. We would welcome it.