Archaeological Evidence for a Horse-drawn Tramway at Bawley Point, NSW

Dr Michael MacLellan Tracey


The book 'Australia's Ever-changing Forests III', and this paper are copyright. Apart from any fair dealings for the purposes of study, research, criticism or review as permitted under the Copyright Act, no part may be reproduced by any means without permission. Enquires should be made to the Publisher or Author.

Publications Section

Centre for Resource and Environmental Science

The Australian National University

Canberra ACT 0200

Tel (02) 6249 277 Fax (02) 6249 0757

Bibliography Reference

Tracey, M. M., 1997, 'Archaeological Evidence for a Horse Drawn Tramway - A case study of the Bawley Point Sawmill System.' Australia's ever changing forests III Third National Conference on Australian Forest History, Centre for Resource and Environmental Studies, ANU, Canberra, pp 188-209.

Archaeological evidence for a Horse-drawn Tramway at Bawley Point, NSW


Bawley Point is a rocky headland located on the New South Wales south coast midway between Ulladulla and Batemans Bay. The surrounding terrain is sparsely vegetated and swampy with some eucalypts thriving in small pockets of elevated land. Dominant species in the area include spotted gum (Eucalyptus maculata), bloodwood (E. gummifera), silvertop ash (E. sieberi) and turpentine (Syncarpia glomulifera) (Costermans 1981: Evans 1978:5). These species were exploited during early forest harvesting and archival research supports that a sawmill operated there between 1892 and 1922 (Tracey 1994). The archaeological remains of the sawmill are located directly below the ridgeline on the northern shore of Bawley Point approximately 300 metres north-west of the Termeil Trigonometric Station (MS428C).

Imported tradition and technology

Industrial archaeological sites are complex by their nature and may comprise several independent, interactive areas where specific functions were performed. A timber extraction operation is where timber felling, transportation, milling and shipment functions combine in an industrial undertaking that has an impact on the landscape.

Identifiers of past industrial endeavours, discarded tools, including derelict machinery or parts thereof, building foundations, decaying hulks of sunken vessels and earthworks often remain in the archaeological record as testament to technological change. To fully comprehend the change in a specific location it is necessary to consider a wide sphere of often very remote, although influential, events of technological change. Constant changes in applied technology, evidenced in the archaeological record, have been prompted by existing human need in response to environmental conditions. The evolution of a tramway system or railway is a complex process of invention and development according to the 'spirit of the times' (Rapp 1981:7). It represents the culmination of the combined skills of the inventor, craftsperson, engineer and labourer according to the prevailing sciences and engineering practices. Tramways continued to evolve in Britain, eventually changing to steam traction and steel rails, although the wooden tramway survived in marginal timber operations throughout Australia long after the change in occurred in Britain. The tramway at Bawley Point remained a wooden, horse-drawn system from 1896 to 1916 and the archaeological remains provide a unique opportunity to study the applied technology of the era.

Tramways in Australia developed from the technology employed in Britain in the 18th and early 19th century which was imported by the early settlers. In the later half of the 16th century, crude railways employing horse-drawn wagons with wooden wheels and rails had been used in coal mining in England and western Europe (Daumas 1979:332-3). Tramways with rolling stock using both wooden and steel wheels on timber track had been in use in England c.1602 and an engraving, first published in 1752, illustrates a track system with carriages equipped with flanged wheels being pushed by people (Daumas 1979:332). Flanged cast iron wheels running on squared wooden runners fastened to sleepers with steel nails did not come into common use until 1790 (Snell 1973:2-5). The flanged wheel operates by employing a projecting rim or riser cast into the wheel to guide it along the tracks. The introduction and use of iron or steel rails enabled an economical, high-volume, reliable transportation method for both freight and passengers. In England, horse-drawn systems were gradually replaced with steam engines when Richard Trevithick, John Blenkinsop and William Hedley built and operated several steam locomotives from around 1797 to 1813. Trevithick built the first steam locomotive in 1804 for the Penydarren Iron Works in Wales. It was capable of hauling loads up to 25 tons, however, it was too heavy for the track and was converted to a stationary steam engine. Trevithick's engine inspired other engineers, such as Stephenson, to construct steam locomotives (Daumas 1979:338-339). Yet despite these revolutionary changes in motive power, horse-drawn tramways continued to be used and operated in industrial, mining, quarrying and even farming localities well into the nineteenth century. Despite their widespread use and importance in the industrial revolution in Britain, tramway systems have rarely been explored archaeologically (Jones 1987; Hughes 1990).

The early settlers of the Colony of New South Wales introduced a strong tradition of acquired technology, timber exploitation methods and knowledge of mechanised milling procedures and engineering practices (Pearson 1995). English railway design and innovation played a major role in early Australian tramway and railway history. The colony possessed a very limited manufacturing capacity, particularly for steel, and relied almost totally on England for imported raw material and manufactured goods. Cast iron railway wheels, axles and other line rolling stock infrastructure were imported from England. A horse-drawn tramway, based on the English inclined plane system, had been in service in the coal industry near Newcastle since the early 1830s. The tramway was used to transport coal from the pit face to the loading facility at the port (Pemberton 1986:31). Horse drawn tramways remained an efficient means of hauling heavy loads over short distances in the latter half the 1800s and early 1900s (Snell 1973:1-3). In the mid 1890s a wooden tramway operated at the wharf in Ulladulla. When first constructed, human labourers provided the motive power, and later, horses drew the loads (Longworth 1994:196-9, McAndrew 1995:70-71). This technology and transportation technique, once introduced into the timber industry in Australia, remained in widespread use until the advent of the internal combustion engine.

Figure 1. Maps comparing the historical record with the archaeological remains. (Click thumbnail to enlarge).

Tramway at Bawley Point.

The research discussed in this chapter was aimed at identifying and establishing the extent and significance of the archaeological remains that may illustrate the applied technology employed in a tramway system. According to historic sources the tramway had been used in conjunction with timber extraction operations in the Murramarang district on the south coast of New South Wales. Limited archaeological or historical research relating to the timber extraction industry in the area has been published. Local histories, such as They Came to Murramarang (Hamon 1994) a welcome effort in collating such history sympathetically include tales of the timber trade derived from local residents. However, the historical accuracy of such works is sometimes clouded by local folklore. Scarcity of archaeological research into historical methods of timber extraction is recognised, prompting the Australian Heritage Commission to comment that: 'To date there has been little systematic surveying of historic sites in forested areas. Forests have the potential to reveal a great deal more information about settlement in Australia and adaptation to the Australian forest environment' (Heritage News 1990:8).

While acknowledging the historical record, primary importance was accorded to identifying the impact and archaeological evidence of the construction of the tramway. Initial surveys also indicated that there may have been a contradiction between the archaeological evidence and the documented material. Several photographs and their inaccurate and ambiguous descriptions in published histories and the collections of the National Library of Australia led to confusion in the field (Hamon 1994:36).

Photographs are an excellent resource when identifying applied technology as 'the archaeologist of Australian industry cannot in any circumstances ignore the historical context of his site and the relevant written, oral or pictorial evidence' (Jack 1983). Although historians treat documents with 'selectivity and interpretation' and know well that the 'documents lie" McGowan (1995:5), exceptional care is required when using historical records, particularly photographs, to identify and describe archaeological sites and artefacts. For example, Jack (1987) describes the photograph of a dredge as 'Gold mining on the Turon, No date' This and several other photographs of the same dredge were well described by the original photographer (Tilba Tilba Collection, National Library of Australia). It was built by William Clow, a timber miller of Wagonga for the Punkalla Gold Dredging Company in 1902 and its archaeological remains are in Simpson's Swamp on the south coast of New South Wales where it remained for the duration of its working life. The Turon Goldfield or the Turon River is over 450 kilometres north east of Simpson's Swamp. It is difficult to comprehend how the history of the two areas could be confused.

A second example is provided by a photograph of a bridge-crossing which is described variously in several publications as the tramway crossing at 'Willinga Lake', at 'Milton' and at 'Dermal Lake'. Hamon published an non-annotated version of the photograph and stated that it depicts: 'Horse team led by 'Old Prince' crossing the tramline bridge over Willinga Lake, c.1914' (Hamon 1994). Backhouse (1991:24) also published the photograph with the annotation on the print 'Log Hauling, Milton, New South Wales' and the caption 'Log Hauling about 1926.' These publications, written by local historians, state two conflicting dates twelve years apart, describe the location as Willinga Lake and Milton thirty kilometres apart and accredit the photograph to these respective areas. It was standard photographic practice to annotate the title, negative number or date of the photograph on the negative in Indian ink which then appears white on the developed print. In the case of this photograph it is possible that copying or retouching of the original has occurred (Tracey 1994:62). It is reasonable to assume that the person who marked it knew where it was taken. It is annotated as 'Milton' not Willinga Lake. A copy of the photograph archived in the National Library of Australia, has the description:

Mr McKie Collection, National Library of Australia Milton / Timber Log Hauling (lent for copying August 1982). Date: Approx 1924-1925. Photo taken on Dermal Lake, Near Ballie Point near Milton on the South Coast of NSW.

The horse teams [sic] on its way to the sawmill to Ballie Point. The fellow sitting on the first log was possibly Jack Backhouse. The leader of the team of horses had to be an outstanding horse with intelligence, good temper, able to stand a little bit of whip, cunning, not too free, one that will come to and gee off promptly, good hearing, placid and lovable. The bullock teams hauled the logs out of the steep hills to a dump. The mill laid a narrow gauge rail line into the bush (in this case about 5-7 miles). Then three bogey sets of train wheels were fixed up and the horse team would haul the logs from the dump to the mill at Ballie Point. The horses had to be hooked one ahead of the other to walk between the rail lines. Not used very widely as a method of hauling logs according to Mr. George Backhouse. There are no reins on the horses but they are driven by talking to them, lightly hitting them, at times cracking the whip. A good driver loves his horses and only hits them when they really deserve it.

The following issues raised in the description of this photograph are contentious and diminish its credibility:


1 1924-1925 - Bawley Point Mill was burnt down in 1922 and was not rebuilt;


2 Ballie Point - The spelling may vary (Arnheim's Map uses Ball-y Point). Irrespective, it is not near Milton;  

3 Dermal Lake - Meant to be 'Termeil Lake'. A family by the surname Dermal operated a sawmill at Dermal Lake or as it was later to be known Termeil Lake. Termeil Lake is not near Milton but many kilometres south;  

4 a narrow gauge -The gauge was to be established from the archaeological evidence was 1.2 m (48") which is not considered narrow for a tramway;


5 train wheels - bogie wheels were specially cast for forestry purposes;

6 not used very widely as a method of hauling logs - This was a very common method of haulage which had been imported with the early settlers.


A second published photograph of horses on a tramway caused confusion when relied upon for site identification and direction for log haulage. Hamon (1994: 36) states it is at Kioloa and shows the photo with the horse team moving left to right, whereas in Ewin (1991:254) states it is at Termeil and has the team moving right to left. Ewin also states that jinkers are being hauled on the tramway whereas, although horses did haul under-slung jinkers, they never did so on a tramway; trolleys are hauled on tramways. This photograph has also been published in Heritage News (1990:8) showing the horses moving right to left, with caption 'Termeil c.1900'. In Ewin's (1991:256) Meet the Pioneers a wooden framed poppet head constructed at a mining site at Nerrigundah in the Nerrigundah Division of the Southern Mining District NSW, is described as the Kioloa sawmill located at O'Hara Head (AR 1899:36). Further authority is inferred in the photograph by the identification of two workers on the site and specific dates are quoted that cannot be applied to the sawmill. Archaeological site descriptions derived from historians 'writing knowingly about places that they have never actually seen' must be accepted by the archaeologist and historian with caution (McGowan 1995:5).

Theories generated by archaeologists often stop at the high water mark ignoring the myriad of terrestrial artefacts lying on the seabed (Nayton 1992:17). '... to look landward and not seaward is a fundamental mistake, and that the criteria which allows such sites to be best understood can be either wet, or dry, or both in various proportions' (Hunter 1994:261).

Figure 2. The author measuring the remains of one of 9 bogies located on the seabed at Bawley Point. Note the heavy crustacean and marine growth. Photo: Steve Harding 1995.

Initially archaeological evidence located for the tramway at Bawley Point was very limited. However, a survey of the seabed near Bawley Point was conducted and artefacts were identified and recorded. These artefacts may represent elements of the earliest applied technologies and trade of the Colony of New South Wales (Jeffery 1987:17-19, 1993:1-6; Hunter 1994:262; Westerdhal 1994:265). The romance associated with shipwrecks attracts maritime research, and although wharves, jetties and lighthouses have attracted heritage conservation studies, there has been minimal intensive archaeological analysis of these remains. Discarded industrial components are often ignored (McCann 1994:92-99). Until the development of road systems and subsequent rail transport, the Colony of New South Wales was heavily dependent upon transport by sea, the maritime industries providing vital links for economic and cultural exchange (Bach 1976:3; Hannah 1986:29; MUDHS 1988:5). Milled timber was dispatched by ship and the Bawley Point sawmill had been located where moorings for ships were accessible and minimum handling of the finished sawn product was necessary (Hannah 1986:15).

Establishment of the timber industry at Murramarang

The first European settlement on the south coast of New South Wales was in the Milton district at Croobyar Creek in 1827 (Antill 1982:38; Bayley 1975:29; MUDHS 1988:5). Sydney Stephen, a barrister, applied to the Land Board in February 1828 and was granted land in March (Dixon [map] 1837). He gained possession in 1829 and named it 'Edgecumbe' (Brown: 1987; MUDHS 1988:5). William Carr, a fellow barrister, formed an association with Stephen, enabling the convicts assigned to Carr to begin to oversee the property now known as 'Murramarang.' Stephen eventually left the district for New Zealand after a prolonged dispute with Carr. Carr planned an expansion of his interests and in 1842 purchased 860 acres (348 hectares) at Kioloa. Despite its poor harbour he had argued, albeit unsuccessfully, for a village to be surveyed at Kioloa and by the time of his death in 1854 he owned 5340 acres (2161 hectares) from Kioloa to Bawley Point (Hamon 1994:21). He bequeathed his estate including Murramarang, the main property, to his widow Charlotte. Eventually the property was sold to William Yates and Evan Evans of Dapto and later it passed to John Evans who worked the area as a dairy farm (Tracey 1994).

Bawley Point, part of Evans' land holdings, was surveyed 16 October 1892 by Fred Arnheim. Noted on his plan was 'Sawmill in course of erection' (Spl91.4). This notation is the earliest official documentation relating to the Bawley Point sawmill. The mill and its tramway were built simultaneously and were in use by 1893. The editor of The Ulladulla and Milton Times (1893) described 'a crude light line of timber railway servicing Guy' Mill at Bawley Point.' Goodlet and Smith had built the Kioloa sawmill at O'Hara Head five kilometres south of Bawley Point in 1891 and about this time, Francis Guy Jnr., son of Francis Guy, who had sawmilling, mining and shipbuilding interests around Batemans Bay since 1870 commenced milling at Bawley Point (MUDHS 1979:4, 9-10).

The 1890s depression detrimentally affected the south coast timber industry and operations ceased at the Kioloa sawmill after a boiler explosion in 1893. Guy's Bawley Point sawmill continued to function under duress. Ships which were vital for transportating the timber and delivering supplies were lost in 1895, 1896 and 1897 causing further shut downs. Despite the loss of his last schooner, Gleaner, Guy maintained operations until 1912 when he finally sold out to A. & E. Ellis of Sydney. Goodlet and Smith had withdrawn and sold their equipment from the Kioloa sawmill in 1898 (Brown 1987; Hamon 1994).

Sawmill handling process

A sawmill represented the final stage in the timber harvesting process. Rough cut logs were broken down into usable sizes and converted to sawn product before transportation or delivery to markets. The specific location of the Bawley Point sawmill on the coastal extreme made it reliant upon a tramway system to transport logs from the forests which were up to eight kilometres inland (Hannah 1986:15).

Although the land adjacent to Bawley Point is relatively flat and sparse, its timber resources around Termeil, Ingolds Knob and Cockwhy Ridge, southwest of the sawmill site, had major gradients. When the trees were felled, they were snigged usually by bullocks to loading areas where they were loaded onto tramway trolleys. Horse teams or individual horses hauled the trolleys directly to the sawmill, or to a point where freewheeling was possible (Hamon 1994:37-38; Tracey 1994:56).

Timber tramways generally required minimal capital investment and were constructed of wooden runners secured to wooden sleepers placed on the ground. Limited earthworks, such as cuttings and causeways, were required to maintain the required gradient of the tramway bed and few other expensive constructions, such as bridges, were necessary in the Murramarang area. By contrast, similar tramway systems in the ash forests of Victoria and the Waitakere Ranges and Mokai of New Zealand often required extensive engineering features, such as bridges and tunnels (Evans 1993; Griffiths 1992; Mahoney 1991:79; Lowe n.d:8-11).

Figure 3. A four wheeled under-slung jinker reconstructed according to the archaeological remains located, oral descriptions and historical photographs.

Limiting factors of a tramway system were the steepness of up-gradients, on which a team or individual horse could pull loads of between 1.2 and 8 tons, and the steepness and length of down-gradients which horses and trolleys, possibly with a brakeman, could negotiate. A gentle terrain was ideal to minimise these limitations and to enable the tramway bed to be constructed with relatively few curves. However, curves were utilised as a method of slowing trolleys. Tight curves lowered the speed of trolleys as they were freewheeled along the track. Steep inclines were negotiated in a tramway system by using a zig-zag configuration, thereby reducing the gradient. Long and steep downgrades could be negotiated by unhitching the horses and allowing the loaded trolley to freewheel down the slope with a brakeman controlling its descent (as had been practised in England's coal industry (Snell 1973:13). Other requirements for a tramway system were a fleet of bogies, trolleys, horses, their tack, shelter and feed. Bullock teams snigged the logs in rugged country with steep inclines, valleys and gullies usually using under-slung jinkers or snigging skids (Tracey 1994:39-51), although horses were also used sometimes on relatively level terrain. It appears that tramways ran where gradients did not exceed approximately 10 degrees. Ridges, narrow valleys and gullies with gradients of approximately 20 degrees were worked by bullocks whose greater pulling power could sustain the loads on these inclines (Cannon 1985:9-15). The location of the sawmill at the lowest point in the cutting area and almost at sea level was used to take advantage of gravity for freewheeling trolleys on the tramway, for at least part of the way to the mill (Hamon 1994:37-38; Griffiths 1992:39; Evans 1994 pers. comm.; Lowe n.d:8-11).

Evidence for the tramway

The historical evidence for the Bawley Point tramway system comprises: an oral report and a 1:25,000 scale map annotated by Peter Scheele of Kioloa (EJLFA 1994), a conservation plan for the Edith and Joy London Foundation property at Kioloa (Titchen 1986), public record cadastral maps for the Parish of Kioloa (Map 44682) and the Parish of Termeil, a map published by Hamon (1994), and photographs in private and archival collections. Given the limitations of these sources, archaeological evidence was sought.

To establish the existence of the tramway system, it was important to assess the applied engineering features such as earthworks and bridges at the time of operation of the sawmills and isolate that period from later forest operations. Archaeological evidence for the tramway system was difficult to locate, given the ephemeral nature of the earthworks and that sleepers and runners may have been laid directly on the ground in some sections. Runners and sleepers had been removed when farms had been established along Reedy Creek immediately behind Bawley Point in the 1940s (Gillespie 1995 pers. comm.). The linear clearings were subsequently used for roads, and the tramway timbers were used elsewhere or in other tramways.

An estuarine lagoon, Willinga Lake, had to be crossed to gain access to Bawley Point. During a period of low water level, several runners and remains of bridge timbers were recorded. Many large lengths of timber, damaged by teredo worm, lay scattered about the shore of the lake. In the mud on the eastern bank lay lengths of timber 10 x 10 cm x 2.7 m [4" x 4" x 9"] with 9.5 mm [3/8"] diameter holes drilled at 46 cm [18"] centres along their lengths and 7.6 mm [3"] from the ends. Most holes were blocked with rust indicting the use of nails or spikes for the fixing of the runners and sleepers. However, no nails or spikes were found in the area. These lengths of drilled timber displayed characteristics of turpentine. They were in a better preserved condition than the larger timbers and had minimal teredo worm damage.

The bridge construction methods were similar to those interpreted from the remains of a tramway bridge approach at Johnsons Creek in the Kioloa sawmill system further to the south. Earthworks on the eastern bank next to the lake and in alignment with the timbers on the lake's edge provide evidence that this was the bridge crossing. Earth removed to a depth of 2.5 metres from a small quarry may have been used to infill an embankment for the track on the bridge approaches. The tramway bed, approximately 3.6 metres [12"] wide, curves away at this point and heads through a small cutting to a depth of 1.5 to 2 metres towards Bawley Point until it is truncated by the bitumen road at Shearwater Crescent (Tracey 1994:60). The tramway then ran along the ridge down to the mill. This area is now a housing development and no evidence of this section of the tramway remains.

Five kilometres from Ingolds Knob on the southern and eastern flanks near Boardinghouse Road remains of the tramway bed have been located running parallel to the present graded road for half a kilometre. This section of the tramway branched away from the Stephens Creek spur line and ran west to the foot of the Cockwhy Ridge, continuing on to cross the Old Princes Highway, and on to Fletchers Road. Although there is no further evidence for the tramway beyond this point, it is considered that it did continue for approximately one kilometre along Fletchers Road towards Bridge Creek following the ridgeline.

The gradient of the tramway bed on the approach to Ingolds Knob varies from 10 to 15 degrees where freewheeling of trolleys is known to have been practised (Hamon 1994:37-38; MUDHS 198:33). Loaded trolleys rolled down the Boardinghouse to 18 Mile Peg Road until they reached the flatter land via Gillespies Creek behind Willinga Lake. Evidence exists for the tramway bed on the eastern side of the Princes Highway in the vicinity of Gillespies Creek. A linear clearing with associated earthworks is evident in the forest. Sections of turpentine runner complete with nails were reported to have been removed by the owner of the property to enable ploughing of the land and for safety purposes (Gillespie 1994; pers. comm.).

The bed for the tramway has been cut into the hillside to about 1 metre and up to 2 metres at various points to form a terrace and is approximately 3.5 to 4 metres in width. The earthworks on this section of tramway display similar construction methods to those at the zig-zag, Willinga Lake and Johnsons Creek. Roots of several spotted gums up to 1 metre in girth have grown over the terraces of the earthworks, thereby supporting the estimated date of construction for the tramway bed. This line provided access to the timber resource, mainly spotted gum, east of the Cockwhy Ridge and the turpentine at Gillespies Creek. Some timber was felled here for the construction of the SS Douglas Mawson, indicating that this section of tramway was in operation c.1914 (Antill 1982:323; Tracey 1996, Settree 1994 pers. comm.). This could also indicate that any quality timber closer to Bawley Point and Termeil had been previously cut out, supporting a statement in the historical record that there was a 'log shortage' (Hamon 1994:33).

Limited and mostly ephemeral archaeological evidence for the physical dimensions of the tramway was available in the form of earthworks and associated disused tramway beds. Contradictory remarks existed in the historical record as to the gauge, construction materials and types of equipment used for the tramway and its specific location in the field. It was important to the understanding of the operation to establish the gauge of the system, however, efforts to locate physical evidence in the terrestrial environment had not been successful. Once a tramway had served its purpose it was disposed of or disassembled for re-use. If the description 'though crude'is accepted, it may have been constructed with inferior quality material, such as spotted gum, which is prone to rapid decay in the damp environment of the forest. These processes are common to timber tramways (Evans 1994 pers. comm.) which 'are often very short lived affairs built to serve a purpose and then abandoned. Consequently, despite all efforts of historians, some of them will never be traced.' (LRRSA 1974:5-6). Moreover, the snigging tracks and tramway beds were subject to forest regrowth and natural weathering, while some were destroyed when they were used much later by bulldozers and log trucks. During the course of the research described here, the archaeological integrity of sites, along with many artefacts, was lost when a section of the bed of the tramway at the Dangerboard on the Kioloa tramway system was destroyed by logging and the area surrounding the Bawley Point sawmill was graded twice for use as a car park. Sites in this area must be considered under the threat of destruction from grading, logging, bushfire, flooding, storm surge and advanced rates of decay due to environmental conditions and infestation by teredo worm.

To establish the gauge of the tramway, it was necessary to locate a sleeper and measure the nail holes for fixing the runners. Sleepers could not be located and a decision was made to undertake a survey of the marine environment adjacent to the sawmill site at Bawley Point. There was a possibility that components of the tramway may have been dumped in the sea when milling operations ceased or if the equipment had been damaged beyond repair. Prior discussions with scuba divers had indicated that 'rusty lumps of metal' did indeed lie on the seabed (Harding 1994: pers. comm.).

A detailed marine survey revealed seven trolley bogies complete with wheels and axles, the wheels having seven spokes, a wide tyre, deep flanges and risers. The number of spokes in bogie wheels in general use varied from five to nine, depending on the manufacturer or the casting process (Houghton 1975:57). From the measurements of the artefacts on the seabed, the tramway was demonstrated to be of 4 ft gauge (measured between the inner wheel flanges - Figure 4).

Figure 4. Reconstruction of bogie and axle derived from artefacts on the seabed intricate measurements were not possible to obtain owing to the concentration of marine growth.

Measurements taken were wheel diameter 61 cm [24"], tyre width 20cm [8"] riser height 6 cm [2"], axle aperture 10.1 cm [4"]. Sleepers probably extended up to 30 cm [12"] on either side of the runners making each sleeper 1.8 metres [6"] in length. This measurement corrects the historical account from an unquoted source: 'The turpentine tracks were less than a metre wide' (MUDHS 1988:33).

The bogies are heavily encrusted with soft coral, plants and marine crustaceans. Most are partly buried in sand, which in this area is subject constant movement and re-deposition, adding to the difficulty of recording the artefacts. One bogie set displays a square raised feature at the centre point along the axle shaft. It is thought that this may have been a swivel box to allow the negotiation of tighter curves by the trolleys on the tramway. Justifiable doubt remained that this discarded equipment may have come from another or later operation and the measurements of the bogies on the seabed had to be checked against physical evidence on the tramway bed to confirm the gauge of the system. The wheels were cast from chilled, possibly grey cast iron, using the sand casting process (Arnold 1989). The purpose of the curvature in the spokes in the wheel was to allow for expansion and contraction within the mould during casting thereby ensuring an even tyre and flange.

After the gauge of the bogies and the dimensions of the wheels had been established, further evidence on land in the form of a fractured segment of grey cast iron 15 x 10 x 2.5 cm (6 x 4 x 1") was recovered. This artefact was exposed on the lower eastern end of the sawmill site when this was graded for use as a car park (Tracey 1994). It has a curved outer face and was initially considered to be part of a cast container or cauldron. However, by estimating the radius of the curved section, the position of the spokes and the width of the object, and comparing these measurements to the wheels of the bogies on the seabed, it was concluded that it was a part of a bogie wheel. The curvature of the rim matched the 61 cm [24"] diameter, the material was the same and the raised section was similar to the inner radius that held the cast spokes.

During a severe storm, eight additional segments of cast iron were exposed at the edge of grading scree where the surface soil had been washed away. These were re-assembled into the radius of the wheel diameter of 61 cm [24"] established from the wheels on the seabed. The initial artefact located fitted to another section of the casting. The wheel had been fractured in many places and the amount of rusting on the edge of the artefacts indicated that the fracturing did not occur during grading. It had probably failed during operation and been discarded. Although no evidence of the hub or bearing could be located, further indication of sub-surface ferrous metal was recorded electronically. The recovery of the hub and examination of the type of bearing cast or fitted into the hub would assist in identifying the period in which the casting was done. When the segments were assembled, they matched the measurements of the wheels on the seabed. Subsequently this area has been regraded and the remaining artefacts lost in the grader scree, including the suspected hub section.

The zig-zag

An interesting engineering feature of the tramway, the zig-zag, exists on the slopes of an extended ridgeline at Stephens Creek. It was constructed to enable horses to haul logs up a gentler slope than was possible in a direct ascent. The tramway bed continues to the top of 18 Mile Peg Road and on to join the Boardinghouse on the ridgeline. In initial surveys, attempts to locate this feature failed, owing to the dense forest cover, leaf litter and extensive disturbance of surrounding areas by later logging operations.

At 150 metre square section of the forest where the tramway was reported to have traversed was selected for survey and a search was instigated to locate any nails, sleepers, runners or parts thereof. Several extensive surface surveys had failed to locate any evidence of the tramway owing to the dense vegetation cover and forest litter. It was probable that sections of the tramway were disassembled when the timber cut out and were reused at other locations (Hamon 1994:109-110). Such actions would account for the minimal extant archaeological evidence for the tramway. Excavation was beyond the scope of the research at the time and any surface disturbance had to be kept to a minimum. Despite extensive effort, the field survey failed to reveal any indication of nails used to secure the tramway runners to the sleepers along the bed of the tramway, although thousands would have been used. It is likely that the nails for the tramway were manufactured in the blacksmith's shop at the sawmill evidenced by the salvage of a nail heading tool (Bealer 1976:96-97). The aperture of the heading tool fitted all nails found on the tramway.

Figure 5. The zig-zag tramway at Stephens Creek.

This historical engineering feature must be considered under threat from further forestry operations. It is unique on the southeast coast of NSW and worthy of heritage protection.

The fact that turpentine is difficult to burn makes it unlikely the runners were collected for firewood, or that the lack of remains could be attributed to the district's susceptibility to forest fires. Also, turpentine is resistant to insects and dry rot, characteristics which should have assisted its preservation in the archaeological record (Maiden 1889:89; Warren 1892).

Adopting the hypothesis that any metallic artefacts in these areas represented post-contact deposition, a metal detector was selected as the most non-destructive means by which to test the archaeological viability. A Compass Relic Magnum 6 Metal Detector fitted with a 200-mm coil was selected for its accurate metallic discrimination functions, depth of ground penetration, and adaptability to be altered for use in the marine environment. Only artefacts on the ground surface were recorded. If a sub-surface object was located, the detector was used to establish its approximate size, shape and metallic composition (ferrous or non-ferrous). No further investigation or disturbance of artefacts was undertaken. After scanning in a random pattern with this instrument for several hours a ferrous metal detection was recorded. A one meter square grid was set up in the area and systematically swept to reveal a similar reading approximately every 38 to 45 cm [15 to 18"]. According to the documentary sources the gauge of the tramway was less than 1 metre and it was thought that remains of nails would be found according to this description. A second parallel bearing was established by alignments of the detector's readings 1.2 m (48") to the west of the string line. Removal of leaf litter revealed timber runners with steel nails driven into them lying on the ground surface. The gauge of the tramway runners matched the measurements of the bogies on the seabed establishing 4 feet gauge tramway.

Figure 6. Nails and nail header tool from the tramway and blacksmith's shop.

Figure 7. The rusting process evidenced on the nails provided an identifiable artefact that was used to locate the path of the tramway.

Depressions, 1.8 metres [6"] in length and at 46 cm [18"] spacings were present under the runners. Directly above each depression and in place in the runners, were two nails. Nail holes in the turpentine may have been pre-drilled to 9.5 mm [3/8"] because the nails may have bent when attempts were made to drive them into this very hard dense timber. The depressions were semi-circular, suggesting that the sleepers were half rounded, not squared, as had been reported. They were cut from spotted gum. Although spotted gum rots quickly on the ground, it is in abundant supply in the immediate vicinity and is soft enough when green, for a nail to be driven without pre-drilling. The repeated angle of the nails driven into the sleepers also indicates that the sleepers were half round.

Figure 8. Details of the construction method of the tramway and drainage practices.

Further evidence to support that the sleepers had rotted away may be seen in the condition of the nails. All nails inspected had points and the upper section 10 cm [4"] was heavily rusted, whereas the lower section remained in good condition. Rusting resulted from moisture being retained by the turpentine timber accelerating corrosion on the head-end of the nail. When the sleeper had rotted, the steel would have been suspended in the cavity left by the decayed spotted gum. Exposed to the relatively dry air, this section of the nail would have been less susceptible to corrosion. This pattern was repeated at a bridge and approach at Johnsons Creek and Gillespies Creek. The evidence of this process could then be used as an indicator in the field to identify nails that had been used to affix tramway runners. In several sections, where earthworks cut into the slope, shallow depressions in addition to the sleeper depressions were located. These hollows, up to 30 cm [12"] deep, may have been drainage culverts, however, positive identification was not possible without considerable disturbance or excavation of the site.

Further use of the metal detector and marking of nail positions confirmed the direction of the runners and enabled a compass bearing to be taken. The direction of the switchback of the tramway was estimated by comparison with the angle of slope. Switch backs 9 to 12 metres [30-40"] long had been constructed at either end of each arm of the zig-zag. Point systems, drainage works, embankment earthworks and a small cutting were also identified.

The switch-backs were worked by a pivot runner or switch runner held into the sleeper by an 18 x 46 cm [11/2" x 18"] steel pin that was manually pulled across the line. The horse teams would then have been unhitched and walked to the opposite end of the trolley, where it was then re-hitched. The length of the switchback would only allow for the length of the trolley, complete with log and horse, to be approximately twelve metres [40"]. Allowing three metres [10"] for the horse and chain this leaves the average maximum length of the log to be approximately nine metres [30 feet]. This provided one limiting factor for this section of the line. A log of a maximum length of approximately nine metres [30"] only could be hauled on this section of tramway. The gradient of the slope in this location varies between 18 and 25 degrees although the maximum grade on the tramway bed of the zig-zag does not exceed 10 degrees.

Tramway line infrastructure

According to historical sources, the Bawley Point/Termeil tramway ran as a complete system for approximately 22 kilometres and was constructed of turpentine and spotted gum (MUDHS 1988; Hamon 1994:36-39). The timber runners were reported to be turpentine, Syncarpia laurifolia, which was obtained only north of Termeil and did not grow in the area traversed by the tramway (Titchen 1986). Approximately 44 kilometres of turpentine runners were required for the Bawley Point/Termeil system, representing a major capital investment including its haulage into the area.

The hardy properties of turpentine were known by the time of the construction of the tramway. It was being used in jetty and bridge pylons, as well as for tramway runners (Warren 1892:10). The field survey confirmed that turpentine grows south of Termeil and within the area served by the tramway. At 18 Mile Peg Road and Gillespies Creek mature turpentine trees are evident, however, they are S. glomulifera and not S. laurifolia (Costermans 1981; Blombery 1967:67). One tramway spur line may have been constructed primarily to exploit the turpentine and transport it to the remainder of the Bawley Point tramway under construction. Oral history maintained that sleepers measured 30.5 x 30.5 cm [12 x 12"] possibly split to 30.5 x 15.3 cm [12 x 6"] and were mounted with turpentine runners affixed with nails. Although this suggests the use of squared timber for the sleepers, it was contradicted by field investigations at the zig-zag which revealed that split round sleepers were used.


Early rolling stock used on the tramway consisted of four-wheeled trolleys with two sets of bogies per jinker and two jinkers per trolley (Evans 1994:34; Griffiths 1992:39; Winzenreid 1986:51). The trolleys were used in pairs and were spaced apart to suit the length of the logs, enabling them to be transported full length. A typical set of loaded bogies weighed 8 tons (Houghton 1980, LRRSA 1974; 5-6) The trolley unit had a short wheelbase, they were not sprung, and had deeply flanged wheels fitted to avoid derailments on uneven tracks (Mahoney 1991:79). Trolleys, with the logs chained securely, were connected by chain to an eyelet welded to a plate that was bolted to the wooden frame above the bogie axle. This enabled the trolleys when attached to the horse to negotiate bends on the tramway. Several plates and eyelets were located on the seabed. The nature of the bogie set, with wide 20 cm [8"] tyres were such that they would be very stable on timber runners by allowing a skewing movement over the surface of the runner to maintain stability (Griffiths 1992:150).

 Figure 9: The wide bogie wheel allowed skewing of the bogie set over the wooden runners enabling freewheeling of trolleys to be practised.

Stability of the trolleys on the tramways would have been of paramount importance requiring a wide gauge line particularly during freewheeling. The horse team was detached, brakes applied and the jinkers pushed to freewheel downhill. Freewheeling of loaded wagons was employed with logs for the Bawley Point sawmill at Boardinghouse Road and down the Dangerboard at Kioloa. Early settlers would have been aware of the inherent dangers associated with log transport, particularly where horses could be inadvertently killed while engaged in hauling loads down wooden tracks. The runners became very slippery in wet conditions (Evans 1994:15-17). Snell (1973:3) comments about events in 1754 in England: '... accidents were frequent. Steel and iron rails when wet get slippery enough, but wooden ones were far worse. Wagons  'running a main were a common site...' As stated by Hamon (1994:37) accidents were apparently common on the Dangerboard on the Kioloa tramway system.

Earthworks on a section of the tramway leading from Willinga Lake Crossing averaging 3.6 metre [12"] in width, allowed for a track of at least 1.2 m [4"] as well as a turning clearance for the log on the curved track. Some timber cutting operations in other areas may have employed horses hauling logs on steel rails (Houghton 1975:16, 30). No evidence for a steel rail tramway exists in the study area or in the district's historical record. However, a bogie located on the seabed, having only a 5 cm [2"] flange and an axle of 61 cm [2"] is consistent with the type used on steel to transport saw dust from the mill to the sawdust dump thirty metres east of the main building.

Figure 10. A typical trolley reconstructed from the archaeological evidence .


No mention of the breed of horses at the Bawley Point sawmill is made in documented history. Photographs indicate that they were Shire horses or Clydesdales, or at least that these breeds were genetically dominant. The Clydesdale is an even-tempered, intelligent draught horse originally bred in Lanarkshire, Scotland, for drawing heavy loads. It had been imported as early as 1795 undertaking the arduous journey of over 16,000 miles in sailing vessels (Denholm 1979:102-106). A horse of 18 hands is capable of drawing up to 3 tons (Cannon 1985:9-13; Kennedy 1992; Wilcock 1976) and some argue that a single horse can haul up to 12 tons on a tramway (Snell 1973:7). Many articles or segments of horse harnesses and tack were found on a blacksmith's shop site adjacent to the main mill building. Large horseshoes of hand forged steel found on this site have calkins steps on the ends of the shoe to afford the draught horse sure footage. This artefactual evidence convincingly indicates that draft horses were employed in transporting logs to the sawmills.

Stables for the Bawley Point sawmill horses were situated on the north west bank of Willinga Lake on the western side of the tramway crossing. No relevant surface features could be observed as this area had been graded in the late 1960s. The site was examined with a metal detector however it revealed no evidence that could be attributed to the stables or their function. The site most likely to have been the stables is visible in the 1944 photograph although it had been graded for a small tree plantation by the time the 1972 photograph was taken (Collins 1994: pers. comm.; AP Run 13 NSW 2016\5117; J55\464 Map 277).


The terrestrial and maritime archaeological evidence supports the idea that a wooden tramway system operated between Bawley Point and Termeil. It was constructed at the same time as the sawmill and was not converted to steel rails. The bogies used were not altered or changed and no form of traction other than horses was employed. The dimensions and physical characteristics of the tramway were clearly established from the evidence as previously stated. Using this data and field measurements it was possible to produce computer reconstructions of rolling stock, bridging designs, loading areas and associated applied construction techniques described in historical texts. The archaeological evidence does correlate at certain common points on the composite map based on the historical documentation, however it does not fully support all the routes suggested by the various historical sources. Bawley Point to Termeil line carried logs and supplies through the Bawley Point hinterland, where gradients do not exceed 5-10 degrees. On the northern side of Willinga Lake the tramway crossed the lake via a wooden bridge then continued to the sawmill.

Boardinghouse branch line ran from the loading area and possible assembly point at Fletchers Road down the Boardinghouse to 18 Mile Peg Road with an initial steep (20 degree gradient descent) into Gillespies Creek to join the Bawley Point/Termeil line approximately 3.5 kilometres from Bawley Point. The historical record suggests 3 kilometres for this junction. The archaeological evidence supports the length of the Bawley Point tramway system to be approximately 13 kilometres.

These sites are under threat of destruction and, if the comment by the Australian Heritage Commission in respect to forest histories and their importance to the understanding of settlement patterns within Australia is accepted, there is an urgent need for further research and recording of such tramway systems. A similar tramway operated at Kioloa mill at O'Hara Head south of Bawley Point, another at Pebbly Beach further south from O'Hara Head and at Bannisters Point near Ulladulla. At Jones Beach on Bannisters Point silica deposits were mined and transported for processing. Scant evidence, if any, exists for the operation of these tramways in the historical record and the ephemeral archaeological remains are under constant threats, similar to those at Bawley Point.


Steve Harding Geographer and scuba Instructor and my colleagues Jennifer Tracey, Cultural Heritage Research Centre, University of Canberra, Wilfred Shawcross and Ian Farrington, Department of Archaeology and Anthropology, Australian National University. My thanks also to Kate Tracey for her valued assistance in the field.


Antill, R. G. 1982. Settlement in the south: a record of the discovery, exploration and settlement of the Shoalhaven river basin 1803-1982. Kiama: Weston & Co.


Arnold, K. 1989. Cast iron: valuation guide. Crown Castleton: Maiden Gully.


Bach, J. 1976. A maritime history of Australia, Sydney: Pan Books.


Backhouse, G. 1991. Where the big wheels scream and other stories. George Backhouse.


Bayley, W. A. 1975. Shoalhaven: history of the Shire of Shoalhaven NSW. Nowra: Shoalhaven Shire Council.


Bealer, A. W. 1976. The art of blacksmithing, New York: Funk and Wagnall.


Birmingham, J., Jack, I. & Jeans, D. 1983. Industrial archaeology in Australia rural industry. Richmond: Heinemann.


Blombery, A. M. 1967. A guide to native Australian plants. Sydney: Angus and Robertson.


Cannon, A. R. 1985. The bullock driver's handbook. Shepparton: Night Owl Publications.


Costermans, L. 1981. Native trees and shrubs of south-eastern Australia. Adelaide: Rigby.


Daumas, M. 1979. A history of technology and invention; progress through the ages, the expansion of mechanisation 1725 -1860. New York: Crown Publishing.


Denholm, D. 1979. The colonial Australians. Ringwood: Allen Lane.


Evans, M. 1978. 'Forest Types.' Forest & Timber 14(2): 5.


Evans, P. 1993. A report to the Australian Heritage Commission on saw milling and tramway sites in the central highlands forests of Victoria. Melbourne: Department of Conservation and Natural Resources, v.2.


Evans, P. 1994. Rails to Rubicon; a history of the Rubicon Forest. Melbourne: Light Railway Research Society of Australia.


Ewin, J. 1991. Meet the pioneers, early families of the Milton-Ulladulla District with photographs. Milton: Joanne Ewin.


Griffiths, T. 1992. Secrets of the forest; discovering history in Melbourne's Ash Range. St Leonards: Allen & Unwin.


Hamon, B. V. 1994. They came to Murramarang; A history of Murramarang, Kioloa and Bawley Point. Canberra: Australian National University, Centre for Resource and Environmental Studies and The Edith and Joy London Foundation.


Hannah, H. 1986. Forest giants, timbergetting in the New South Wales forests 1800-1950. Sydney: Forestry Commission of New South Wales.


Henderson, G. and Baker, P. 1993. 'The wreck of the Sirius'. Australian Geographic. 29:103-7.


Houghton, N. 1975. Sawdust and steam: a history of the railways and tramways of the eastern Otway Ranges. Melbourne: Light Railway Research Society of Australia.


Houghton, N. 1980. Timber and gold: a history of sawmill and tramways of the Wombat Forest 1855 to1940. Light Railway Research Society of Australia.


Hughes, S. 1990. 'The Brecon Forest Railway: The archaeology of an early tramway. Aberystwyth': Royal Commission on Ancient and Historical Monuments in Wales.


Hunter, J. R. 1994. 'Maritime culture: notes from the land.' The International Journal of Nautical Archaeology 24(4): 261-4.


Jeffery, B. 1987. 'Some views on the historic shipwrecks legislation in South Australia,' Bulletin of the Australian Institute for Maritime Archaeology 11(1):17-19.


Jeffrey, B. 1993. 'Maritime archaeology: what's in it for Australians.' Bulletin of the Australian Institute for Maritime Archaeology. 17(2):1-6.


Jones, N. W. 1987. 'A wooden wagonway at Bedlam Furnace, Ironbridge'. Post-Medieval Archaeology, 21:259-62.


Kennedy, M. J. 1992. Hauling the loads, a history of Australia's working horses and bullocks. Melbourne University Press.


Light Railway Research Society of Australia, 1974. Tall Timber and Tramlines: An Introduction to Victoria's Timber Tramway. Surrey Hills: Light Railway Research Society of Australia.


Longworth, J. 1994. 'The Ulladulla harbour tramway'. ARHS Bulletin, July: 196-9.


Lowe, D. (no date). Rails across the ranges; an introduction to the bush tramways of the Waitakere Ranges Auckland: The Lodestar Press.


MacAndrews, A, 1995. The Narrawallee story retold, Epping: Alex McAndrews.


Mahoney, P. J. 1991. 'Bush tramways in New Zealand: an unrecognised resource.' Australian Historical Archaeology, 9:79.


Maiden, J. H. 1889. The useful natives plants of Australia (Including Tasmania). Sydney: The Technological Museum of New South Wales, Turner and Henderson.


McCann, A. M. 1994. 'The Roman port of Cosa. 'Ancient Cities. New York: Scientific American.


MUDHS, Milton Ulladulla and Districts Historical Society, 1988. Nulladolla 1988. Ulladulla,


Nayton, G. 1992. 'The importance of the maritime perspectives to the understanding of Australian historical archaeological sites'. Bulletin of the Australian Institute for Maritime Archaeology. 16(2): 17-24.


Nef, J. U. 1957. Coal mining and utilisation, a history of technology. in C. Singer, E. J. Holmyard, A. R. Hall and T. I. Williams (eds), Volume III, Oxford: Clarendon Press.


Pacey, L. 1990. The story of Wagonga Inlet. Narooma: Laurelle Pacey.


Pearson, W. 1995. (in press), Water power in a dry continent; the transfer of water wheel technology from Britain to Australian in the 19th century.


Pemberton, P. A. 1986. Pure merinos and others; the Ashipping lists' of the Australian Agricultural Company, Canberra: Archives of Business and Labour, Australian National University.


Rapp, F. 1981. Analytical philosophy of technology. Boston: Kluwer.


Snell, J. B. 1973. Railways: mechanical engineering. Manchester: Arrow Books.


Titchen, S. M. 1986. Towards a conservation plan for the Edith and Joy London Foundation property at Kioloa on the south coast of New South Wales. Canberra: Unpublished Honours Thesis, Australian National University.


Tracey, M. M. 1996. A launching and shipwreck - the saga of the Douglas Mawson. Canberra: Archaeology Australia Publishing (in press).


Tracey, M. M. 1994. When the timber cut out, archaeological aspects of timber extraction procedures and shipbuilding in the Murramarang district, NSW. Canberra: Honours dissertation, Australian National University.


Warren, W. H. 1892. Australian timber. Sydney: Government Printer.


Westerdahl, C. 1994. 'Maritime cultures and ship types: brief comments on the significance of maritime archaeology'. The International Journal of Nautical Archaeology 23(3): 265-70.


Wilcock, R. 1976. Gentle giants, the past, present and future of the heavy horse. London: Lutterworth Press.


Winzenreid, A. P. 1986. Britannia Creek: an essay in wood distillation. Melbourne: APW Productions.



Arnheim, Fred. Extract from Lands Department NSW REF: 5-3022 Sp. L 5 CA map. Surveyed by Fred Arnheim 16 October 1892. Lands District of Milton, Land Board of Cooma


AUSLIG, Canberra, ACT. Ulladulla area 16,000 feet 12 February 1944, Run J55\464 Map 277; NSW Coastline 20,000 A.S.L. 6 June 1972, Run 13, NSW 2016\5117.


Australian Geographic, 1993


Brown, L. J. 1987. John Evans family history, Epping.


Department of Mines and Agriculture New South Wales. Annual Report for the year 1899.


Dixon, R. Map of the Colony of New South Wales, exhibiting the situation and extent of appropriated lands including the counties, towns, villages and reserves. 20th July, 1837.


Forest and Timber, 1978- 1982, Forestry Commission of NSW.


Heritage News, 1990.


Jack, I. 1987. Heritage Study Report to the Evans Shire Council, Volume 1.


Kioloa Sheet 1:25,000 8926-I-N, Second Edition, Department of Lands, Sydney.


McGowan, B. 1995. Historic mining sites survey of the Shoalhaven and south west slopes districts of New South Wales. Canberra: The New South Wales Department of Urban Affairs and Planning and the Australian Heritage Commission.


Parish of Kioloa cadastral map (AO Map 44682, Archives Authority of NSW. Cancelled 1907).


Parish of Termeil cadastral map. (No reference available, Cancelled 1916) (Appendix 6).


The Scheele Map, Edith and Joy London Foundation Archives, Department of Sociology, Australian National University, Canberra.


Ulladulla and Milton Times, 1897.


Interviews with:

Mr. Innis Collins, Kioloa NSW, 1994;

Mr. Peter Evans, Melbourne Vic, 1994;

Mr. Bill Gillespie, Bawley Point NSW, 1995.;

Mr. Steve Harding, Canberra ACT, 1994;

Mr. Alf Settree, Huskisson NSW, 1994;

Mr. Peter Scheele, Kioloa NSW, 1994.


Heritage Archaeology ~ Historical, Archaeological & Genealogical Research Consultants

 (Copyright 2013)

Saturday, 09 May 2015