Reference

Date 1978

Event ID 640467

Category Documentary Reference

Type Reference

Permalink http://canmore.org.uk/event/640467

The investigation of this wreck has comprised 24 weeks of fieldwork by the Institute of Maritime Archaeology and the Bristol Undersea Archaeology Branch of BS-AC. Although some peripheral areas of the site remain unexcavated, it is considered that the full extent of the surviving structure has been defined.

The dimensions of the ship were documentarily recorded as cited by Adnams (above), and her established armament comprised 16 demi-culverin drakes (9-pdrs), 16 saker drakes (6-pdrs) and 4 minion forts (4-pdrs). Sir John Tippetts, who became Surveyor of the Navy in 1672, had studied ship design and construction in Denmark during the reign of Charles I, and the design of small British warships had developed rapidly in the 1640's, notably through the introduction of the revised hull form that was derived from the 'Dunkirk frigates' and was exemplified in the construction of the Constant Warwick, built in Pett's yard at Ratcliffe in 1646. The Dartmouth was built on the basis of these developments, and was reflected in Van de Velde's sketch of the mid-1670's, which well portrays the sleek light hull characteristic of this type. An itemised account of the work carried out in the major refit in 1678 and two succeeding surveys (of 27 October 1679 and 10 October 1680 respectively) form major documentary sources for ship construction of this period.

The wreck of the vessel has entered the folk-memory of the area, although often confused with the 'Tobermory wreck' [San Juan de Sicilia: NM55NW 8013]. Jacobite-inspired witchcraft features heavily (and improbably) in accounts which also mention the vessel drifting stern-first across the sound after her anchor-cables had given way and other anchors dragged on the third day of a storm. The vessel was probably at least partially capsized before she was driven by the wind for about 2 miles (3.2km), and the location on 'King's Point in Morvern' was remembered until at least the mid 19th century.

Although two modern wrecks lie nearby, they are separated from the archaeological site by a rock spur and most of the intrusive material recovered reflects island picnics and hunting or shooting parties, probably of the 19th century. A steady tidal stream runs across the N side of the islet from NE to SW, but the site is reasonably well protected from currents. It is also well protected from heavy seas, although exposed to the NW; the maximum fetch of 11 miles (17.7km) in this direction is too short to allow the development of a large swell. The stability of the seabed is increased by a luxuriant coverage of kelp (Laminaria digitata).

The wreck lies at the foot of a sloping rock against the side of the island. At the shoreward end, wreckage is contained within a wedge-shaped gully only 8-10ft (2.44-3.05m) deep at LWS; one gun (no. 20) lies in even shallower water. Another gun (no. 2) lies partly up-ended in a nearby cleft, and three more (nos. 1, 3 and 5) were found close to the base of the northern rock face. Excavation of the shoreward end of the wreckage has revealed artefacts clearly associated with the after end of the vessel: navigational and surgical instruments, a flintlock pistol, balance weights, and fine tableware in pottery and pewter. Leaded mica pieces found in this area probably came from the windows in this part of the ship, and may be identified with the 'Muscovy glass' specified at the time of refit. This serves to confirm the traditional view that the ship struck stern-first.

The main spread of coherent structure starts where the mouth of the gully starts to widen out and slopes gradually seawards, following the base of the rock down to a depth of 17ft (5.18m). About 18ft (5.49m) of the keel survives at the inshore end. Before excavation, the timber structure was covered by up to 2ft (0.61m) depth of shingle and boulders, and was further pinned by several large concretions which included guns, shot and an anchor 13ft (3.96m) long. Much of this material was evidently ballast as it contained flint chippings, several broken guns (nos. 4, 6, 7, 8 and 10) and fragments, and also shot too large to fit the armament carried.

Between the W end of the coherent structure and the inwards curve of the cliff towards the NW (a distance of some 9.14m), excavation revealed a gravely deposit up to 2ft (0.61m) deep which overlay bedrock near the shore and a firm clay/pebble conglomerate elsewhere. This deposit contained a quantity of scattered timber, most of it identifiable as elm planking or fir sheathing. Ballast flints were present throughout this area and in its western part there were found items of seamanship stores, consistent with the location of the bosun's store in the forward part of the ship. These comprised rigging fittings of various types (including cordage, parrel and shroud trucks, deadeyes, blocks of several sizes, and loose sheaves). Some of these items showed evidence of makeshift repair, while others were unused. Also located in this part of the ship was the armoury, represented by concreted boxes or kegs of hand grenades, with large quantities of lead shot of pistol and musket calibre. These identifications support the suggested general arrangement of the ship, lying ENE-WSW with the stern to the ENE and the bow to the WSW.

Some 30ft (9.14m) NW of the assumed centreline of the ship and roughly parallel with it, there was a linear scatter of guns (nos. 12-19) running for about 60ft (18.3m) from NE-SW at a depth of about 20ft (6.1m). Several Roman numerals cut from lead sheet and found near the SW end of this line are best identified as draught-marks from the stem. The line of the guns is also followed by a series of lead scupper-liners (close to guns 12, 14 and 16) which may be associated with the side of the ship at deck level.

Halfway along this line (around guns 15 and 16) lay a pile of bricks and tiles evidently connected with the galley structure. The bricks were of well fired red clay and typically measured 8.6 by 3.9 by 2.4 ins (218 by 99 by 61mm), dimensions significantly different from the 17th century English standard. The tiles were of softer and brighter orange fabric, and typically measured about 6 by 0.5 by at least 7 ins (152 by 13 by 178mm), and were pierced at the top by nail-holes. The brick-built floor of the structure apparently rested on the main deck of the forecastle, and supported a firebox, boiler shelf and flue of brick, while a tiling shroud on a wooden framework served to reduce the danger of fire. The slates previously noted by Adnams may have formed part of this structure. Among the remains of the galley, there were found considerable quantities of burnt debris, animal bones and coal, the latter being readily available in Greenock. A length of narrow-bore lead piping was found in this area. This was flanged at one end and may have formed part of the plumbing of the boiler. The bell would also have hung in this area.

Some 20ft (6.1m) W of the forward end of the line of guns lay a pair of medium-sized anchors measuring 7 and 8ft (2.13m and 2.44m) respectively in length and nestled crown-to-ring as if stowed and lashed when lost. A piece of lead sheeting and a lead pipe of large bore found nearby were possibly associated with the ship's sanitary arrangements; the latter may have been one of the 'pissdale pipes' of the 1678 refit.

A lateral trench of width 10ft (3.05m) was excavated from just forward of the coherent structure towards the line of guns and scupper-liners, so as to determine the relationship between the two main axes of the wreck. A deposit of loose stones and gravel some 18ins (0.46m) deep was found to overlie a firm n depth clay/pebble substratum of natural origin. The upper deposit was cleared down to this layer to reveal a scatter of artifacts. These present the impression of having been subject to a pronounced heeling motion towards the ship's side, where they were retained until the side rotted away.

The sectional relationship between the coherent structure and the seabed strata aft indicates that the ship came to rest heeled at an angle of 23 degrees to starboard. The port side of the ship would then have projected above the surface, and so has vanished above the inboard edge of its garboard strake. Much of this side is thought to have been salvaged [by persons unknown and presumably without diving]. The section also shows that the lower part of the keel and the entirety of the false keel were removed by abrasion. These two timbers are recorded as measuring 13ins (0.33m) and 8ins (0.2m) in depth respectively, and these measurements are reflected in the depth of the trench that was evidently driven by the wreck into the substratum while the ship was substantially intact, allowing water movement to rock the hull and cause the keel to act as a moving fulcrum. The starboard planking would then have fanned away to form the hollow into which it settled with minimal damage. A thick matte of wood splinters from the abraded keel and other organic components then settled beneath and around the stabilised hull remains while several substantial portions of framing (presumably from the disintegrated port side) settled into the trench. This trench also became a trap for about half the artifacts subsequently recovered, suggesting that the ship broke up fairly rapidly and in a continuous sequence before the natural seabed deposits became re-established.

The situation forward is rather different forward in that the coherent structure ends around a rock spur which projects beyond the keel axis almost as far as the large anchor amidships, strongly suggesting that the ship broke her back in this area (close to the mainmast step). The steep cliff and the sloping seabed would have acted together to roll the forward portion to starboard, into a position now represented by the line of guns. In this process, much of the planking would have sprung, releasing the ballast and many artifacts to fall onto the seabed. The forward part of the hull then settled onto its starboard side, forming a trap for a variety of objects, while the brick and tile galley fell onto the guns, with the belfry above it. The timbers themselves then rotted away (leaving the lead scupper-liners and draught numbers as markers) while the heavier objects worked their way through the seabed sediment, in some cases reaching the harder underlying substratum. The two westernmost anchors and the postulated 'heads' waste pipe indicate the location of the bow itself.

In general terms, analysis of the site formation processes (and also the demonstrable non-deployment of the spare anchors) supports the traditional account of the ship having drifted across the Sound on her beam ends before striking stern-first and going to pieces 'to the great rejoicing of the MacLeans'.

Individual structural components were recorded in detail. The keel was of elm and much abraded on the underside although clear toolmarks and traces of paying stuff (apparently white lead and oil) survived on the upper surface. The surviving portion measured 18ft (5.49m) in length and included the full length of a scarf joint; the deadwood bolt-holes and garboard recesses were clearly identified. The dimensions cited accord with those recorded at the time of refit, and this keel deadwood bolt-holes and garboard recesses were clearly identified and this keel must be that fitted in 1678. The overlying rising deadwood was also of elm, and served to retain the (non-continuous) frames in an unconventional manner, which may reflect the constraints of timber supply or the re-keeling operation.

C J M Martin 1978.