Memoirs of the Queensland Museum (ISSN 1440-4788)/media/Documents/QM/About+Us/Publications/... · Papers published in this volume and in all previous volumes of the Memoirs of the

MEMOIRS OF THE

QUEENSLAND MUSEUM BRISBANE

© Queensland Museum PO Box 3300, South Brisbane 4101, Australia

Phone 06 7 3840 7555 Fax 06 7 3846 1226

Email [email protected] Website www.qm.qld.gov.au

National Library of Australia card number

ISSN 0079-8835

NOTE Papers published in this volume and in all previous volumes of the Memoirs of the

Queensland Museum maybe reproduced for scientific research, individual study or other educational purposes. Properly acknowledged quotations may be made but queries regarding the republication of any papers should be addressed to the Editor in Chief. Copies of the journal can be purchased from the Queensland Museum Shop.

A Guide to Authors is displayed at the Queensland Museum web site

A Queensland Government Project Typeset at the Queensland Museum

HMS PANDORA PROJECT — A REPORT ON STAGE 1: FIVE SEASONS OFEXCAVATION

PETER GESNER

Gesner, P. 2000 06 03: HIV1S Pandora project — a report on stage 1: five seasons ofexcavation. Memoirs of the Queensland Museum, Cultural Heritage Series 2(1): 1-52.Brisbane. ISSN 1440-4788.

In 1790 the Pandora, a 24 gun frigate sailed from England to Tahiti in pursuit of the Bountyand its mutineers. After capturing some mutineers, the Pandora was wrecked in 1791 on itsreturn voyage attempting to navigate the Great Barrier Reef, east of Cape York Peninsula,Australia. The survivors sailed in an open boat from the Barrier Reef to Java and eventuallyreturned to England, where the mutineers were brought to trial. The discovery of the Pandorashipwreck in November 1977 and its' subsequent archaeological investigation by theQueensland Museum constituted an opportunity to expand on the Bounty saga and reconstructits material setting.Between 1983 and 1995 the Queensland Museum conducted exploratory excavations andsurvey over five field seasons, as a precursor to more intensive study. Excavation concentratedon the bow and stern sections where there was well preserved material evidence in the crewsliving spaces and personal storage areas. Maritime archaeology, shipwreck, 1TMS Pandora,Great Barrrer Reef, Royal Navy, HMS Bounty.

Peter Gesner, Museum of Tropical Queensland, 78-102 Flinders Street, Townsville 4810,Australia; 10 November 1999.

HISTORICAL BACKGROUND

HMS Pandora was a naval frigate dispatchedby the British Admiralty on a punitive voyage tothe South Pacific. Its mission was to find andrecapture HMS Bounty and bring to justice 25men who, in April 1789, had mutinied while theywere on the final stage of a voyage to transplantbreadfruit from Tahiti to plantations in the BritishWest Indies. The leader of the mutiny wasFletcher Christian, the Bounty's acting lieutenantwho felt slighted by his commander, CaptainWilliam Bligh, and, in an outburst of rage andfrustration, incited several members of his watchto take the ship and cast adrift 19 of theirshipmates, including Bligh. Following an un-successful attempt to establish a settlement onTubuai, one of the Austral Islands, the Bounty'smutinous crew fell out among each other fourmonths later. Sixteen mutineers elected to returnto Tahiti in September 1789, after which FletcherChristian sailed off with the Bounty to an un-known destination with eight others and theirPolynesian entourage. In the meantime Blighmanaged to return to England, where he reportedthe mutiny and the loss of his ship to the Lords ofthe Admiralty.

HMS Pandora sailed from Portsmouth in earlyNovember 1790 and arrived in Tahiti 011 23 March1791 after nearly five months at sea on the routearound Cape Horn (Fig. 1). During her six week

stay in Tahiti, fourteen of the Bounty's crew weretaken prisoner and locked up in a makeshiftwooden cell on the Pandora's quarterdeck, whichthe prisoners referred to as 'Pandora's Box'.

The Pandora left Tahiti on 8 May 1791 and,until mid-August 1791, cruised the South Pacificin search of the Bounty and the remainder of hermutinous crew. The search was unsuccessfulbecause Fletcher Christian and his followers hadfound a final refuge on uncharted Pitcairn Island,where they had burned and scuttled the Bounty inJanuary 1790.

On 29 August 1791 the homeward boundPandora was wrecked on the Great Barrier Reefwhile exploring an opening in the reef. Four ofthe prisoners and 31 of Pandora's crew diedwhen their ship sank after striking a submergedreef. Having spent two nights on a small sandcay(Escape Cay) in the vicinity of the wreck, thesurvivors —89 crew and 10 prisoners — set outfor the Dutch East Indies in the ship's boats. Afteran arduous 16-day, 2750km, open boat voyagethey arrived in Timor. They subsequently madetheir way to Batavia (Jakarta). From there theywere able to arrange for a passage home via theCape of Good Hope. At Cape Town the 10 sur-viving prisoners were transferred to a Britishwarship and transported to England to stand trial.Three of the mutineers were hanged and thePandora's officers faced courts martial. No

ENGLAND

HMS Pandoraleft England7th Nov 1790

Arrived Canary Islands r.

22nd Nov 1790

ATLANTI:OCEAN

Arrived Rio de Janeiro31st Dec 1790

•----------------

Pitcairn •^DucieIslandÎsland^- • sEaster

s

Island

2^ MEMOIRS OF THE QUEENSLAND MUSEUM

FIG. 1. Pandora's route from England to Tahiti.

efforts were made to salvage the Pandora. Thewreck remained undisturbed until its rediscoveryby scuba divers and documentary film makersSteve Domm, John Heyer and Ben Cropp, withthe assistance of the Royal Australian Air Forcein November 1977.

Following an archaeological survey commis-sioned by the Commonwealth Department ofHome Affairs & Environment and carried out bythe Western Australian Maritime Museum'sarchaeologist Graeme Henderson in April 1979,the wreck was positively identified as the Pandoraand declared a protected site under section 7 ofthe Historic Shipwrecks Act, 1976 (Henderson,1979).

Since 1982, management ofthe wreck has beenthe responsibility of the Queensland Museum.Nine archaeological excavations have beencarried out by the Queensland Museum to date,

with additional financial assistance from thePandora Foundation, the CommonwealthGovernment's Historic Shipwrecks Program andthe Queensland Museum's Board of Trustees.These excavations have established that an ex-tremely coherent and well-preserved collectionof artefacts and a substantial portion of the wreckedhull are buried in the seabed (Henderson, 1986;Gesner, 1988; 1993).

The wreck's historical and archaeological valuesunderpin its international cultural significance. Itis a major source of European and Polynesianmaterial culture associated with a British navalvessel engaged on a long voyage to the SouthPacific Ocean in the last quarter of the 18thcentury. The Pandora's last voyage falls within avery important period of European discovery inthe Pacific Ocean, also referred to as the GrandAge of European Pacific Exploration (Howse,1990; Smith, 1992). This period had started

HMS PANDORA: FIVE SEASONS OF EXCAVATION 3

•10^•^• 7•fi^4°

COAST OF NEW SOUTH WALESFROM CAPE TRIBULATION

ENDEAVOUR STRAITS

Shewing TILE LABYRINTHon a Larger Seale

'^(11 fra titsKAT4PI7i.(00 *LIM' 111167

7/4f

45^JO

•fat, n‘ntr.,.././

inn

4t. voz

IL4Z,

112 146'

deburn*

Cape Gee nvAdmil 11

•

Cape Weylnout

t.4,.I 4,5i* eAmofril2t...........

...... •

FIG. 2. Portion of late 18th Century chart of Australia's east coast showing the Pandora's track (upper centre)along the Great Barrier Reef between 26th and 28th August 1791, as laid down by Lt Hayward. (State LibraryNSW #50, Laurie & Whittle)

during the 1760s with Bougainville's voyage,flourished with Cook's three Pacific voyagesduring the 1770s, and La Perouse's epic voyagein the 1780s, and ended with Matthew Flinders'circumnavigation of Australia in 1802. It was thetime when 'enlightened' European philosophersand scientists were rediscovering the world'snatural, cultural and social environments, recog-nising their multifarious forms, and making aconcerted effort to systematically classify thediversity they encountered (MacLeod & Reh-bock, 1988).

The Pandora episode would probably havereceded into historical obscurity but for twoimportant facts. The track which the Pandorafollowed while searching for a passage throughthe Great Barrier Reef and Pandora Entrance tothe Torres Strait was copied onto a late 18thcentury chart of Australia's east coast (Fig. 2).The named features on the nautical charts

became a constant reminder of the fate andlocation of the Pandora.

More importantly however, the story of thePandora's last voyage is one of the three mainstories constituting the infamous saga of themutiny on the Bounty, arguably the most drama-tised and romanticised seafaring saga from theannals of maritime history — certainly an eventwhich became larger than life from the momentWilliam Bligh returned to England in March1790 with news of the mutiny. Following hisreports about the ordeal which he and his loyalcrew had experienced after being cast adrift in theBounty's launch, Bligh was considered a hero.Feted and acclaimed by society, he was applaudedfor his fortitude and leadership. Soon afterwardshe was given command of another expedition,comprising two vessels, HMS Providence andHMS Assistant, to complete the breadfruitcollecting mission he had originally been

4^

MEMOIRS OF THE QUEENSLAND MUSEUM

entrusted with when given command of theBounty (Oliver, 1988).

Several years later, however, the accounts ofthe ten Bounty crew brought home as prisoners bythe Pandora's Captain Edwards, cast doubts onBligh's version of events. Bligh's conduct as anaval commander was called into question andhis reputation was subsequently tainted byaccusations made against him by some mutin-eers; particularly by Peter Heywood and JamesMorrison, and by influential members and friendsof Fletcher Christian's family. The ensuingcontroversy fuelled renewed public interest in theBounty saga and resulted in a large collection ofnarratives, travelogues, letters, pamphlets andarticles. Many of these mention the Pandora'srole in the Admiralty's attempt to bring to justicethe Bounty's mutinous crew. Several are bymembers of the Pandora's crew who eagerlymade available their journals and accounts tosatisfy public demand for stories from the SouthPacific; in particular about the infamous mutiny.

Indeed, literature on the mutiny has grown tosuch an extent that a complete catalogue of theBounty saga, including references to the Pan-dora's voyage, would probably contain morethan one thousand entries. The list is still growingas contemporary historians continue to findreasons to analyse and re-interpret the mutiny andits aftermath (Dening, 1992: 397ff).

Thus, the main events of Pandora's last voyageare very well known. The names of the crew aswell as the chronology and sequence of events ofthe voyage can be gathered, not only from suchprimary sources as the ship's log (Adm. 180,Edwards' papers), but also from the publishedjournals, memoranda, eye witness accounts andletters written by Captain Edward Edwards andSurgeon George Hamilton (Thomson, 1915;Hamilton, 1998), by midshipman DavidRenouard (Maude, 1964) and by two of the Bountymutineers, Peter Heywood (Marshall, 1825;Tagart, 1832) and James Morrison (Rutter, 1935).

HMS PANDORA'S LAST VOYAGE

EN ROUTE (PORTSMOUTH TO TAHITI). Severaldays after taking up his command of the Pandoraon 10 August 1790, Captain Edwards wassummoned to the Admiralty in London for abriefing on special orders which would take thePandora into the South Pacific on what was to beher last voyage. The orders included instructions:

... to proceed to Otaheite and, not finding themutineers there, to visit the different groups of theSociety and Friendly Islands, and others in the

neighbouring parts of the Pacific and there seize andbring home in confinement all or some of thedelinquents ... (Adm/A/2831).

The unusual nature of the voyage requiredchanges to be made to the normal complement ofa frigate of the Pandora's class. No marines wereto be taken on the voyage, nor were the officersallowed the usual number of 'servants', i.e.apprentices (see Appendix 1). These berths wereto be filled by additional seamen. An extra lieu-tenant was also assigned and the usual number ofmaster's mates and midshipmen was doubled.These changes to the crew were made inanticipation of the need to provide an effectivecrew to bring the Bounty back to England afterher capture. The length, duration and theextraordinary nature of the voyage also requiredmore than the usual quantity of stores andsupplies to be taken on board. In addition, sparesand new fittings for the Bounty were required.Surgeon Hamilton's metaphor likening the crewto weevils, who first had to eat a hole in theirbread to make a space for themselves, is testimonythat the ship was filled to capacity on departure(Thomson, 1915: 92; Hamilton, 1998: 5).

Among the extra officers added to the Pan-dora's crew was Thomas Hayward, one of theBounty's midshipmen. As one of the 18 crew whodid not go with Fletcher Christian in the Bounty,Hayward had accompanied Bligh in the launchand had served as a witness at Bligh's courtmartial. By assigning Hayward to the Pandora asthe third lieutenant, the Admiralty was actingwith some forethought as they felt his par-ticipation would assist in the recognition of themutineers.

After leaving Portsmouth on 7 November 1790,the Pandora's first port of call was Tenerife in theCanary Islands, where she stayed for several daystaking on water, wine and fresh supplies inludingcitrus fruit, bananas and pomegranates (Thom-son,1915: 93-4).

On the leg between Tenerife and Rio de Janeiro,a large number of the crew, including assistantsurgeon James Innes, suffered from a contagiousfever. It was treated by surgeon George Hamiltonin a novel way by supplying the sick andconvalescent with tea and sugar. According toHamilton this was the first time tea had beenintroduced on a naval vessel at sea (Thomson,1915: 94). In spite of Hamilton's best effortshowever, James Johnson, a bosun's mate, diedduring the Atlantic crossing and was buried at seaoff Rio de Janeiro on 1 January 1791.

HMS PANDORA: FIVE SEASONS OF EXCAVATION^5

Edwards called in at Rio de Janeiro specificallyfor more fresh supplies as he was afraid the feverraging amongst the Pandora's crew would notpass before they reached the dangerous waters offCape Horn. After a short stay in Rio de Janeiro,the Pandora then ran along the coast of SouthAmerica toward Cape Horn. The crew's healthimproved because, according to Hamilton, of thecold weather off the Cape and the availabilty ofspecial supplies, including hot chocolate andspruce beer (Thomson, 1915: 100).

With most of the sick crew rapidly recovering,the Pandora passed Cape Horn without mishapin early February 1791 and after another un-eventful month at sea sighted Easter Island on 4March 1791. The voyage would certainly havebeen very different if Captain Edwards hadrealised that Pitcairn Island, the elusive Bountymutineers' final refuge, was within a relativelyshort sail of Easter Island. However, the fate ofFletcher Christian and his followers was toremain unknown until 1808 when the crew of theAmerican whaler Topaz put in at then unchartedPitcairn Island and met John Adams, the solesurviving mutineer, and the descendants of hisformer shipmates and their Polynesian entourage.

Several more islands were sighted to the westof Easter Island and their position duly recordedin the Pandora's log (Thomson, 1915: 88).Edwards named one of them Ducie Island afterhis patron Admiral Lord Ducie, under whom hehad served as a junior officer and to whose in-fluence he probably owed his current command.But because exploration and cartography werenot a priority, Edwards ignored the islands andcontinued on directly to Tahiti.

AT TAHITI: CAPTURING THE PRISONERS.By heading directly for Tahiti, Edwards wasactually giving effect to his orders to make aTahitian landfall as soon as possible. The reasonfor this urgency lay in the Admiralty's hopes thata sailor called Jonathan (or John) Brown wouldbe able to provide information on the location ofthe mutineers. The Admiralty had received areport that Brown had been put ashore fromanother ship (the Mercury) which had visitedTahiti in 1789. Edwards mentions this in hisjournal when recounting what happened immed-iately after the Pandora's arrival in Tahiti:

Jno. Brown, the person left at Otaheite by Mr Coxof the Mercury, and from whom their Lordshipssupposed I might get some useful information, hadbeen under the necessity for his own safety to associatewith the pirates ... I entered Brown on the ship's booksas part of the complement and found him very

intelligent and useful in the different capacities ofguide, soldier and sailor (Thomson, 1915: 30).

The Mercury had visited Tahiti in August 1789.Several weeks later Captain Cox had in fact comevery close to finding all of the mutineers — aswell as the Bounty — when the Mercury hadpassed the mutineers' first refuge on Tubuai in theAustral Islands. Fortunately for the mutineers,however, the Mercury passed at night, andalthough fires were sighted ashore, Cox did notbother to investigate. While at Tahiti, Cox hadheard of the return of Titreano — the Tahitians'name for Fletcher Christian — in the Bounty,without Bligh (Dening, 1992: 92-4). The storyhad puzzled Cox; perhaps he suspected that themutineers were at Tubuai but considered he wasin no position to act. He waited until he was inEngland to inform the Admiralty (Mortimer, 1791).

Within hours of the Pandora's arrival inMatavai Bay on 23 March 1791, five mutineersgave themselves up. The Bounty's armourerJoseph Coleman was the first to surrender. Hisexample was followed soon afterwards by twomidshipmen, George Stewart and Peter Heywood;then by the Bounty's barber Richard Skinner andthe nearly blind ship's fiddler Michael Byrne.Two days later three more seamen, ThomasEllison, Charles Norman and James Morrison,also surrendered. Morrison, Ellison and Normanhad spent their time on Tahiti with several othersbuilding a schooner in which they had had hopesof sailing for the Dutch East Indies. One of thePandora's midshipmen, David Renouard, recountshow the mutineers with much perseverance, hadbuilt a boat:

... handsomely shaped of about 18 tons which they hadnamed Resolution, to underscore their determinationto use it as a means of escaping from Tahiti (Maude,1964).

However, not all of the mutineers gave them-selves up. Several managed to elude thePandora's men by sailing off in the Resolution.But Brown informed Edwards that they had takeninsufficient water with them and would thereforeprobably soon return to the Papara district. ThePandora's launch and pinnace were sent toPapara and within a few more days another fourmutineers were captured. The remaining two,who had fled to the mountains, were finallycaptured as a result of information provided byBrown. By 9 April 1791 all of them, 14 in total,had been brought on board. This accounted forthe 16 Bounty crew left behind on Tahiti, asEdwards was given credible reports about thedeath of 2 others before the Pandora's arrival.

6^ MEMOIRS OF THE QUEENSLAND MUSEUM

Captain Edwards confiscated the Resolution.After the boat was supplied with canvas sailsfrom the Pandora's stores, she was renamed theMatavai and officially commissioned as thePandora's tender. Crewed by a master's mate(William Oliver), a midshipman (David Renouard),a quartermaster (James Dodds) and six ableseamen, she was to be used for inshore recon-noitering of the islands yet to be searched.

Some of the prisoners protested that theirTahitian friends had prevented them from givingthemselves up. According to Morrison:

Tommaree told us we must go to the mountains andkeep away from the ship. When we refused he said: Iwill make you go! And his men seized us and wasproceeding inland with us. (Rutter, 1935: 120).

This should not be considered as a lame excuse;several of the Tahitian tribal leaders whom themutineers had befriended, had a vested interest inthe Bounty men. Association with the mutineersgave them access to firearms for use againsttraditional enemies or rival leaders. Indeed,Charles Churchill and Matthew Thompson, 2 ofthe 16 Bounty men originally left behind on Tahitiby Fletcher Christian, had died as a result of theirinvolvement in local power struggles (Thomson,1915: 110).

The 14 prisoners were initially confined underthe quarterdeck in leg irons pending completionof a special prison cell, which Captain Edwardsordered to be built on the quarterdeck. This woodencell was soon referred to by the prisoners asPandora's Box. It was 11ft x 18ft (3.5m x 5.5m)on the deck and not quite high enough for ThomasEllison — at 5'3" (1.57m) the shortest mutineer

to stand straight in (Rutter, 1935: 123) (Fig. 3).The building of Pandora's Box took more than

one week and was only one of many tasks thePandora's crew occupied themselves with whileat Tahiti. They overhauled the Pandora's riggingand properly fitted out the schooner Matavai.Thus, the Pandora stayed at Tahiti for severalmore weeks before resuming the search for theBounty and the mutineers still at large.

During this time Captain Edwards interviewedthe prisoners as well as a number of Tahitianleaders to get information about the likely where-abouts of Fletcher Christian and the othermutineers. He also tried to get a clearer picture ofthe events which had occurred on the Bounty afterthe mutiny. In this he was aided by the journalsthe captured midshipmen, Stewart and Heywood,had kept. These had been found in the mutineers'sea chests, which Edwards had confiscated aftertheir capture. He also interviewed several

prisoners on specific points in an effort to find outabout what had happened since the 14 prisonershad arrived back on Tahiti and, most importantly,where the Bounty may have gone to since then(Thomson, 1915: 34).

It is interesting to compare the comments aboutthe prison cell made by the gaolers, CaptainEdwards and surgeon Hamilton, with thecomments made by prisoners Heywood andMorrison. Not surprisingly they have oppositeperspectives and emphasise different aspects.Edwards reasoned that he:

... put the pirates in a round house ... for their moreeffectual security, airy and healthy situation, and toseparate them from and to prevent their having anycommunication with, or to crowd and incommode theship's company (Thomson, 1915: 34).

Hamilton relates that the cell was built so thatthe prisoners would:

... be secure and apart from our ship's company; andthat it might have every advantage of free circulationof air, which rendered it the most desirable place in theship. Orders were likewise given that they should inevery respect be victualled in the same as the ship'scompany, notwithstanding the established laws of theservice, which restricts prisoners to two-thirdsallowance (Thomson, 1915: 106).

These explanations may appear perfectlyreasonable according to the gaolers' perspective,however, a very different light can be cast bycontrasting them with Morrison's lament that:

Pandora's Box, was only 11 feet in length and 18 feetwide at the bulkhead, in which were two small scuttlesof 9 inches, and one on top of 18 or 20 inches square,secured by a bolt. When it was calm, the heat was sointense that the sweat frequently ran in streams to thescuppers, and soon produced maggotts, and thehammocks given to us were full of vermin, from whichwe could find no method of extricating ourselves ... Inthis situation we remained for 5 months ... Ourmiserable situation soon brought sickness on amongstus .... and, as the place was washed twice a week, wewere washed with it, there being no room to shift usfrom place to place, and we had no alternative butstanding up until the deck had dried ... (Rutter, 1935:122).

In a letter to his sister, Nessy, Heywood con-curred with Morrison's complaints about conditionsinside the cell. He stated that the prisoners:

... were all put in close confinement, with both legsand hands in irons, and were treated with great rigour,not being allowed ever to get out of this den. And,being obliged to eat, drink, sleep and obey the calls ofnature here, you may form some idea of the dis-agreeable situation I was in (Tagart, 1832:33) (Fig. 3).

The prisoners' treatment while locked up onthe Pandora is sometimes cited as one of theexamples of the harsh discipline 18th centurynaval crews were subject to. In this regard,Captain Edwards is often considered excessively

HMS PANDORA: FIVE SEASONS OF EXCAVATION^

7

FIG 3. Artist's impression of 'Pandora's Box' (by Robert Allen). The fourteen Bounty crew taken prisoner onTahiti were Joseph Coleman, Peter Heywood, George Stewart, Richard Skinner, Michael Byrne, JamesMorrison, Charles Norman, Thomas Ellison, Henry Hillbrandt (or Heildbrand), John Sumner, ThomasMcIntosh, William Muspratt, Thomas Burkitt and John Millward, four of whom (Skinner, Stewart, Sumner andHillbrandt) did not survive the wreck.

callous, not only because he put the prisonersunder close confinement during the Pandora'sfour month cruise in search of the Bounty, butmore so because he is reported as having shownvery little sympathy for the prisoners' plight afterthe Pandora had struck the reef and during thesojourn on Escape Cay, when the prisoners wereallegedly refused shelter against the harshtropical sun.

How much justification is there to considerCaptain Edwards as an excessive disciplinarian?Obviously, he should not be judged by 20thcentury standards. Most likely, he was simplyfollowing his orders, which were:

... to keep the mutineers as closely confined as maypreclude all possibility of their escaping, having,however, proper regard to the preservation of theirlives; that they may be brought home to undergopunishment due to their demerits (Adm. 2/120).

What obviously mattered to Edwards was thatthe appropriate judicial processes were to befollowed. It was not his prerogative to make,much less act on, judgements or pronouncementsabout the prisoners' guilt or innocence. Theywere to be brought home to undergo punishment

due to their 'demerits'. Thus, by keeping them'closely confined' to preclude escape, all theprisoners were treated in the same manner, inspite of the fact that Bligh had publicly vouchedfor the innocence of Norman, Byrne, Colemanand Macintosh. Peter Heywood later acknow-ledged this during his trial when he stated thatEdwards had not acted improperly by imprison-ing all of the mutineers but had simply followedthe Royal Navy's rules, or, as Heywood phrasedit, ... the dictates of the service' (Tagart, 1832:117).

Captain Edwards was probably also consciousof the need to lock up, guard and separate theprisoners because he was anticipating troublewith some of the prisoners' Tahitian friends. Mostof the mutineers had formed relationships withTahitian women and had also forged bonds offriendship (taio-ship) according to Tahitiancustom with Tahitian men, some of whom weretheir consort's father or brothers. Rival triballeaders had informed Edwards about a plot torescue George Stewart, which was apparentlybeing hatched by Stewart's `taio', who was alsoStewart's de facto father-in-law. Conceivably the

NEW CALEDONIA ••

roeIslands (6 June 1791)8olornon Islands (13 August 1791)

.........^................. Rotuma (5 August 1791)

"SamoartMlands (18 June 1791)

hPalmerston Island(21 May 1791)

Society Islands(23 March 1791)

...

........Cote, Islands(19 May 1791)

•

8^


FIG 4. Track of the Pandora's search for the Bounty in the South Pacific.

plot would put not only the prisoners at risk butthe Pandora as well (Thomson, 1915: 106-7).Viewed in this light then, Edwards' orders tostrictly separate and guard the prisoners can beinterpreted as the actions of a vigilant com-mander, primarily concerned for the safety of hisship and its crew. Hence the orders to prevent anycommunication between the prisoners and theirTahitian friends or consorts. Obviously some-times these orders, and the orders prohibitingcommunication between the prisoners and thePandora's crew, could not always be enforced:for example, as Edwards mentions, during thetimes that the prisoners were taken forward (tothe heads) to relieve themselves (Adm. MS 180,Edwards' papers, bundle 7).

SEARCHING THE SOUTH PACIFIC. Aftercompleting the ship's maintainance and hisenquiries, Edwards decided to leave Tahiti on 8thMay 1791, following the itinerary outlined in hisorders. These instructed him to proceed to theisland of Whytootackee' :

... calling ... at Huahine and Uliatea, where you neednot anchor as numbers of the natives may be expectedto come off to you, of whom you may probably get thenecessary information ... (Adm 2/120).

After `Whytootackee' (Aitutaki, in the CookIslands) the Pandora's crew searched the Union(Tokelau), Samoan and Friendly (Tongan)Islands (Fig. 4).

Using the Matavai or the Pandora's boats and,on one occasion, a sailing canoe hired in Tonga,Captain Edwards sent off armed shore-parties tolook for word or signs of the mutineers still atlarge. On these occasions the Pandora usuallyremained well offshore to avoid the fringing reefstypical of the South Pacific islands. As a result of

this practice the ship's cutter, with five men undercommand of midshipman John Sival, was lostwhen it failed to return from a search ofPalmerston Island (Fig. 4) on 24 May; it was notheard of again (Thomson, 1915: 86) (Appendix1).

Sival's cutter was the first of two of Pandora'sboats lost. The second one, the schooner Matavai,failed to rendezvous with the Pandora on 23 Juneoff Samoa. Command of the schooner had beengiven to William Oliver, a 20 year old master'smate. After becoming separated from the Pan-dora during a storm, he and his crew successfullynavigated the schooner from Samoa to Tofua,then along the New Guinea coast via the TorresStrait (Fig. 5), to Surabaya in the Dutch EastIndies. This voyage can be ranked with WilliamBligh's much vaunted open boat voyage to Timorin the Bounty's launch. A transcript of a journalkept by Oliver's second-in-command, 16-year-old midshipman David Renouard, has survivedand is of considerable historical interest;particularly because it describes several Pacificislands not previously recorded by Europeanexplorers (Maude, 1964). The schooner's crewarrived in Java on 20th September, several weeksbefore the survivors of the Pandora shipwreck;which occurred at about the same time as Oliverand his crew were struggling to get the schoonerthrough the Torres Strait.

However, their woes were not over yet; Ren-ouard recalls, when they made Surabaya:

Mr Oliver immediately waited on the governor toacquaint him with our misfortunes and to implore theprotection and assistance due to British subjects indistress. But the fate of the 'Bounty' had beencommunicated to (him), in consequence of which thegovernor suspected the truth of our story. The


9

FIG 5. Watercolour attributed to George Tobin showing the tender Matavai struggling through Tones Straitwaters. (Reproduced courtesy of the Australian National Maritime Museum)

appearance of our vessel, being built entirely ofOtaheitan wood, served to strengthen him in theopinion that we were in reality part of the 'pirates whohad seized on the 'Bounty' ...

As suspected Bounty mutineers, the crew wasimprisoned in Surabaya. But William Olivereventually managed to persuade the governor tolet them go on to Batavia. On the way there theyput in at Samarang where, during the last week ofOctober, they fortuitously met up with theirformer shipmates who had survived the wreck ofthe Pandora.

Captain Edwards eventually found out why theMatavai had failed to turn up at Anamooka.Oliver had been given orders to proceed toAnamooka (modern Nomuka), one of TonganIsland Group, in the event the tender failed torendezvous with the Pandora off Samoa. Butwhile bound for Anamooka, Oliver miscalculatedthe tender's speed and arrived at Tofua, aboutsixty miles downwind. Mistaking Tofua forAnamooka, the tender's crew waited there forseveral weeks, after which Oliver assumed some-thing untoward had occurred to the Pandora;consequently he decided to make for the DutchEast Indies (Maude, 1964).

In the meantime as the Pandora had waited forthe tender off Anamooka, several of the Pandora'screw experienced difficulties in their interactionwith the islanders. As related by George Ham-ilton, one incident involved Captain Edwards'clerk whose clothes were stripped from him andstolen.

Another incident was an assault on lieutenantRobert Corner, who apparently felt sufficientlyprovoked to shoot dead one of his assailants.None of these incidents appears to have pre-vented what Hamilton referred to as a 'brisktrade' being carried out between the islanders andthe Pandora's men (Thomson, 1915: 132-5). Ona previous occasion Hamilton also mentionedtrade with islanders, specifically the purchase ofobjects not related to provisions or the ship'sstores at:

...Whytootackee, an island discovered by CaptainBligh ... Here we purchased from the natives a spear ofthe most exquisite workmanship; it was nine feet longand cut in the form of a Gothic spire ... (Thomson,1915: 123).

The remainder of the Pandora's voyage throughthe south west Pacific was comparativelyuneventful. Shortly after leaving Anamooka on1st August an auction of the tender crew's

10^MEMOIRS OF THE QUEENSLAND MUSEUM

possessions was held. In his papers CaptainEdwards mentions the sale of William Oliver'sand David Renouard's personal possessionsamong the Pandora's officers. This was doneprematurely, in light of the fact that the crew ofthe tender were to be reunited with the Pandora'screw two months later.

Several discoveries of small islands were alsomade, including Rotumah (Fig. 4). On 13 Augustan island Edwards named Pitt's Island (VanikoroIsland) was sighted in the Santa Cruz Group.Hamilton describes it as mountainous andassumed it was inhabited because of smokeobserved in various parts of it (Thomson, 1915:140). However, a shore party was not sent out toinvestigate. Possibly some of these fires wereassociated with the survivors of the Frenchexplorer La Perouse's expedition, whose lastknown port of call had been Port Jackson, inFebruary 1788, after which his vessels LaBoussole and L'Astrolabe had not been heard ofagain. If the Pandora had stopped to investigate,further light may have been thrown on the mysteryof La Perouse's disappearance. But definitiveinformation on the loss of La Perouse's shipswould not be available until 1827, when thesandalwood trader Peter Dillon went to Vanikoroto investigate reports about the source of Europeanobjects he had found on neighbouring Tikopia(Dillon, 1829: 33f0. However, somewhat sur-prisingly, Edwards did not stop to investigate andordered a:

... course to the westward between the latitudes of 10 0

and 9033 S, keeping the mouth of Endeavour Straitsopen, by which means (he) hoped to avoid the dangersexperienced by Captain Cook in his passage throughthe reef in higher latitudes ... On 25th August at 9 in themorning, we saw breakers from the masthead ...(Thomson, 1915: 70).

These breakers were named Look-out Shoaland Edwards attempted to by-pass them bysetting a south-westerly course in the hope offinding a direct channel to Endeavour Straits. Onthis course the Pandora encountered the islandsand reefs around Mer at the eastern entrance tothe Torres Strait. Edwards named them theMurray Islands but did not land on them. In orderto by-pass them, a southerly course was followed.However, no suitable passage was found duringthat day or the next. He kept the ship well away atnight and came back to the reef's edge during theday. A large opening was sighted two days later.The yawl was launched and lieutenant Cornerwas given orders to reconnoitre inside the entrance.The Pandora hove to outside the opening. WhenCorner returned late in the afternoon, he signalled

from the yawl that a navigable passage throughthe reef had been found. It is interesting to notethat Edwards mentions that at this stage the yawlwas 'outside the reef' (Adm. MS 180, Edwards'papers, bundle 16). This indicates that by thistime the Pandora may have already drifted intoPandora Entrance while waiting for the yawl'sreturn. Whatever the case, as night was approach-ing, Edwards, ever prudent, ordered it back to theship, to get it on board before nightfall.

THE WRECKING. In ordering the yawl back tothe ship, Edwards was undoubtedly acting as aprecaution against loosing another of the ship'sboats. With 15 men already missing as a result ofmisadventures with the schooner Matavai andthe cutter, the Pandora could ill afford the loss ofanother boat and more men. Undoubtedly thisaccounts for the ship actually venturing into theentrance late in the afternoon to pick up Corner'syawl. With the sun low on the western horizon,visibility would have been greatly reduced andany reefs lying ahead difficult to see. Hove to, thevessel was especially vulnerable to the strongtidal current which must have driven her furtherwest into the entrance, where it was low tide atapproximately 5.30pm (L. Hiddens, pers. comm.).

It is not difficult to imagine that at this point thecrew may have been distracted by signallingbetween the ship and the yawl (Fig. 6). Moreimportantly, with sunset at approximately 6pmand the sun low on the western horizon afterapproximately 5pm, it would have been verydifficult for the lookouts to discern wavesbreaking on the small submerged coral outcropslocated in this part of Pandora Entrance to thewest of Pandora's location.

Edwards' account of the wrecking describeshow the vessel ran aground when:

at about twenty after seven the boat was seen closein under our stem and at the same time we gotsoundings in 50 fathoms. We immediately made sail,but before the tacks were on board the ship struck uponthe reef when we were getting 2 fathoms on thelarboard side, and 3 fathoms on the starboard side. Anhour and a half after she had struck there was eight feetof water in the hold and we perceived that the ship hadbeat over the reef where we had 10 fathoms water(Thomson, 1915: 72).

Thus, an unfortunate combination of factorscaused the ship to run aground on a coral outcropnot much larger than a cricket oval. This outcropis now unofficially refered to as Pandora's Reef andis surrounded on all sides by depths in excess of30 metres (15 fathoms). It is one of several,mainly submerged, small reef outcrops withinPandora Entrance.


11

FIG. 6. The Pandora 'hove to', waiting to take on board Lt Corner's yawl (artist's interpretation, watercolour byB. Searle, 1995).

Possibly, the vessel might have cleared thereef, or at least not have impacted on it as heavilyif she had run onto it several hours later whenhigh water would have been at approximatelyllpm. At this time there would have been betweenthree to four metres of water over the reef (Fig. 7).

Aided by the rising tide, the crew managed torefloat the vessel just before midnight and toprevent her from striking adjacent reefs byanchoring the stricken vessel in 15 fathomswith both anchors' (Rutter, 1935: 125). Morrisonadds how:

Coleman, Norman and Macintosh were ordered out tothe pumps and the boats got out. But as soon as CaptainEdwards heard that we had broke our irons he orderedus to be handcuffed and ironed again ... the Master atArms and Corporal were now armed with each a braceof pistols and placed as additional centinals over us,with orders to fire among us if we made any motion(Rutter, 1935: 126).

One can only speculate why Captain Edwardsordered handcuffing of the other prisoners whohad not been let out of their cell to help at thepumps and why he ordered several armed men on

top of the prison cell. Possibly he was afraid theymight steal a boat, interfere with the crew'sefforts to save the ship, or disrupt plans for theorderly evacuation of the ship.

By all accounts the crew behaved splendidly.They worked hard to prevent the sinking of theship; at the pumps, attempting to stop leaks,repairing equipment or hull damage below decks,fothering the hull, or heaving guns overboard tolighten the ship.

They continued to do so in spite of two fatalaccidents during the night. Hamilton describesvividly how the:

... guns were ordered thrown overboard; and whathands could be spared from the pumps were employedthrumbing a topsail to haul under her bottom toendeavour to fother her ... We baled between life anddeath ... She now took a heel, and some of the gunsthey were endeavouring to throw over board run downto leeward, which crushed one man to death; about thesame time, a spare topmast came down from thebooms and killed another man ... During this tryingoccasion the men behaved with the utmost intrepidityand obedience, not a man flinching from his post(Thomson, 1915: 143).

12^


Hi .fl tide 3.8 metres

2.5 Metres bei1,13:2,tue li)gh Tide Level

y 1.2 Metres

05 metres

Low Tide 0.4 Metres

Time ofWreck

5 pm^6 pm 7 pm^ 11 30 pm

FIG. 7. Tidal range at Pandora Entrance on 28 August1791. (Courtesy L. Hiddens)

The identity of the two men killed before thePandora sank is unfortunately not revealed inCaptain Edwards' list of the 31 crew lost with theship (Adm. MS 180, Edwards' papers).

Early next morning it was clear that nothingmore could be done to save the stricken vessel.Orders were given to abandon ship and to releasethe eleven remaining prisoners from Pandora'sBox. It appears the armourer's mate, Hodges,acted too slowly here. According to Morrison, theship began to sink before Hodges was able torelease all of the prisoners from their fetters:

At daylight the boats were hauled out and, most of theofficers being aft on top of the Box, we begged that wenot be forgot, when by Captain Edwards' ordersJoseph Hodges, the armourer 's mate, was sent down totake the irons off; but Skinner, being too eager to getout, got hauled up with his handcuffs on, and therebeing two following him close, the scuttle was shutand barred again. I begged the Master-at-Arms toleave the scuttle open when he answered 'Never fearmy boys, we'll all go to hell together !'. The wordswere scarce out of his mouth when the ship took a sallyand a general cry of there she goes' was heard. Burkittand Heildbrandt were still handcuffed and the shipunder water as far as the mainmast and it was nowflowing in fast on us when Divine providence directedWilliam Moulter to the place. He was scrambling upon the box and, hearing our cries, took out the bolt andthrew the scuttle overboard. On this, we all got outexcept Heildbrandt (Rutter, 1935: 127).

William Moulter's solicitude, which undoub-tedly saved several other prisoners from certaindrowning, was recognised in 1984 when one ofthe sand cays in Pandora Entrance was namedafter him. The cay at the eastern approach toPandora Entrance, referred to by Edwards asEntrance Cay, is now called Moulter Cay tocommemorate Moulter's humane deed.

ESCAPE CAY. Thirty-one ofthe Pandora's crewand 4 of the prisoners did not survive the wreck.Among the casualties were the two extra guardsposted on top of Pandora's Box, master-at-arms

FIG 8. George Reynolds' pen and ink drawing showinga shipmate in the water after the wrecking. (Privatecollection)

John Grimwood and ship's corporal WilliamRoderick, as well as prisoners, George Stewart,Henry Heildbrandt, John Sumner and RichardSkinner. The survivors, 89 crew and 10 prisoners,reached a small sandcay about three miles to thewest of the wreck. Incredibly, several of theprisoners managed to stay afloat in spite of stillwearing their handcuffs (Fig. 8).

Peter Heywood described the survivors' plightin a letter he later wrote to his sister, adding in themargins of his letter:

... two little sketches of the manner in which HMSPandora went down on 29th August, and the appearancewe who survived made on the small sandy key withinthe reef, about ninety yards long and one hundredathwart, in all ninety-nine souls. Here we remainedthree days, subsisting on a single wineglass of wine orwater, and two ounces of bread daily, with no shelter.Captain Edwards had tents erected for himself and hispeople, and we prisoners petitioned him for an old sailwhich was lying useless, but he refused it, and all theshelter we had was to bury ourselves up to the neck inthe burning sand, which scorched the skin, we beingquite naked, entirely off our bodies, as if dipped in largetubs of boiling water (Tagart, 1832: 71) (Fig. 9).

The identity of the sandcay Heywood describedhas not yet been definitively established. However, itis likely to be Preservation Cay (Fig. 10). Thesurvivors called it 'Escape Cay' to distinguish itfrom Entrance Cay' (Moulter Cay), the sandcaymarking the entrance to Pandora Passage.Edwards placed both cays on latitude 11°23'S(Thomson, 1915: 89). However there is a problemwith this determination as neither Moulter Cay norPreservation Cay are actually on the latituderecorded by Edwards. On modern charts theiractual latitude is at approximately 11°25'S (GreatBarrier Reef Marine Park Authority, 1985). Thediscrepancy of two minutes can be attributed to theaccuracy of Edwards' instrument, which wasprobably an octant or quadrant.


13

FIG. 9. Thomas Heywood's sketches of the wrecking and the survivors on the cay. (Courtesy of the NewberryLibrary, Chicago)

The identification of Preservation Cay as EscapeCay becomes more likely with the addition ofdetails from Morrison's accounts. Morrisonmentions that the cay was the middle of threesimilar cays within Pandora Entrance:

... the whole of it being no more than a smallsandbank washed up on the reef which, with a changeof wind, might disappear, it being scarecely 150 yardsin circuit and not more than 6 feet from the level ofhigh water. There are two more of the same kind ofwhich this is in the middle; between it and the one tothe southward is a deep channel through which a shipmight pass in safety (Rutter, 1935: 128).

Looking at the modern configuration of cayswithin Pandora Entrance (Fig. 10) (Great BarrierReef Marine Park Authority, 1985) and assumingthis reflects the situation in August 1791, thereare three cays and two sandbanks. At spring tidesthe sandbanks sometimes dry, but generally theyare submerged and do not warrant being calledsand cays proper. For this discussion, they havebeen disregarded. Morrison's description appearsto suggest that the three cays in question arealigned along a north-south axis. However, thereare only two cays oriented approximately alongthat axis (Fig. 10); the third cay (Moulter Cay)being well to the east. One explanation for Mor-rison's description could be that one of the twosandbanks to the north of 'Melbourne Cup' Cay(Cay 11-088 on BRA Q 102) (unofficially namedMelbourne Cup Cay by members of the Queens-land Museum's 1983 Pandora expedition tomark the location of a piggy-back horse race onMelbourne Cup race day in 1983) was a

substantial cay in 1791, but has eroded since then.If this was the case at the time of Pandora's loss,then Melbourne Cup (Cay 11-088) Cay may beEscape Cay. However, this explanation isunlikely in view ofMorrison's mention of a 'safe'and 'deep' channel between the middle cay and... the one to the southward'. The channel

referred to is unlikely to be any of the narrow,east-west running passages between MelbourneCup (Cay 11-088) Cay and Preservation (Cay11-091) Cay or between the string of small reefsdirectly to their south. None of these east-westaligned passages is wider than 100m. While theymay be reasonably safe for modern motor-drivenvessels, a navigator attempting to pass throughthem under sail would be reckless in the extremebecause of the strong tidal flows and eddies whichoccur inside the passages. It therefore seems moreplausible that Morrison's use of 'southward' indesignating the 'safe' channel's position shouldactually be interpreted as a channel leading to thesouthward rather than lying to the southward.The wide opening between Moulter Cay andPreservation Coy is certainly wide enough anddoes lead to the south.

Possibly then, this opening is the channelexplored by lieutenant Robert Corner, which thePandora would have taken the next day if she hadnot run aground while manoeuvring to take onboard the yawl. The three cays Morrison men-tioned are most likely to be Moulter (Entrance)Cay, Preservation (Escape) Cay (Cay 11-091)

FIG. 10. Pandora Entrance, modern situation (map drawn from GBRMPABRA Q102).

and the unnamed cay, now referred to unofficiallyas Melbourne Cup (Cay 11-088) Cay (Fig. 10).

During the two days which the survivors spenton Escape Cay, the majority of the men were idle,although a number of them were undoubtedlyinvolved in making more seaworthy the fourboats saved from the Pandora. While one boatwent off to fish, another was sent back to thewreck the next day to see if anything worthwhilecould be salvaged (Fig. 11). They returnedwith part of one of the Top Gallant Masts ... and acat ... found sitting on the crosstrees ...' (Rutter,1935: 129). The cat is not mentioned again in anyother source, and its fate is unknown.

THE SURVIVORS' JOURNEY TO ENGLAND.Escape Cay to Mount Adolphus Island. Bysending lieutenant Corner in the yawl toreconnoitre Pandora Entrance, Edwards was notbeing overly cautious. Undoubtedly most watch-ful commanders would have done the same,especially in light of the remarks about theintricacies and dangers of navigating in GreatBarrier Reef waters made by navigators of JamesCook and William Bligh's calibre.

Captain Edwards was to use the informationgathered by Corner to give effect to what mayhave been another aspect ofhis orders, which was

to find, in a more northerlylatitude, an easier passagethrough the Great Barrier Reefthan the track from Provi-dential Channel to EndeavourStrait followed by Cook in1770. Edwards described thepassage the Pandora'ssurvivors followed, sayingthat they left Escape Cay on:... the 31st August ... at half past

ten in the forenoon (and)embarked and steered N.W. by W.and W.N.W. within the reef. Thischannel through the reef is betterthan any hitherto known. In therun from thence to the entrance ofEndeavour Strait there is a smallwhite island on the larboard end ofthe channel which lies in latitude11 0 23' S' (Thomson, 1915: 75)(Fig. 12).

Edwards omits to name theisland at the end of thischannel, but with reference tomodern charts it is reasonableto assume that the trackfollowed by the Pandora'sboats took them through whatis now named Denham Pass.

Edwards" small white island' (Thomson, 1915:75) can therefore be identified as eitherCholmondsley Island or Wallace Islet (AUS 835).

However, the passage pioneered by thePandora's boats was never generally adopted norrecommended in 19th century sailing directions.Edwards' claim about the superiority of thepassage was not followed up; presumablybecause of a general perception among merchantcaptains that advice from ships' commanderswho had lost their ships ought not to be heeded.More importantly perhaps, how could Edwardshave hoped to be compared to such accomplishednavigators as Cook and Bligh (whose reputationas a skilled navigator and very capable hy-drographer was never adversely affected by thedebate concerning his qualities as a naval com-mander). Edwards' and Hamilton's descriptionsof the survivors' route refer to landmarks andgeographical features noted by both James Cookand William Bligh. Other landmarks and featuresare also noted and named.

Distributed among the four boats, the 99 sur-vivors set off for the mainland; before nightfallthe boats formed into single file, the launch infront with each boat successively in tow (Fig. 13).At daylight on 1 September 1791, the mainland

'Melbourne Cup Cay(11-088) . r ..

i„:, ....Preservation• • •,,(11-091)^,....'(Escape Cay

te.:=

14^



15

FIG 11. George Reynolds' water colour of the masts protruding from the waves. (Private collection)

was sighted and the two yawls sent ashore atFreshwater Bay to search for water, while thelaunch and pinnace made for Mount AdolphusIsland.

Morrison and Hamilton's accounts are mostinformative about the survivors' progress afterEscape Cay. Their accounts give interestingdetails about Cape York, the Torres Straits andtheir inhabitants.

Morrison recalls that the survivors:... embarked in the following manner: McIntosh,

Ellison and myself in the pinnace with Capt Edwards,Lieut. Hayward and 19 officers and men ...; in the RedYawl went Burket & Millward with Lieut. Larkan and19 officers and men; in the Launch, Peter Heywood,Josh Coleman & Michael Byrn, with Lieut. Corner &27 officers and men; in the Blue Yawl, Norman andMuspratt, with the master and 19 officers and men,making ninety-nine souls in all ... next morning, the 1stof September, we made the land ... and the two yawlswere sent in to the land, while we stood on towards anisland where we hoped to get water. In the afternoonwe were joined by the yawls who had got water andhaving filled their vessels followed us ... We stood infor a bay to search for water and as we approached abeach found that there were some inhabitants on it, ...The natives appeared on the beach to the amount of 18or 19 men, women and children, who appeared to beall of one family; they came off freely to the boatswhen we found that the colour of their skin washeightened to a jett black by means of either soot or

charcoal ... some had holes in their ears which werestretched to such a size as to receive a man's arm. Wemade signs that we wanted water which they soonunderstood, and half an ancker being given to one ofthem & some trifles by way of encouragement, he soonreturned with it almost full which (we poured) into abrecco and we gave it to him again. He then called ayoung woman who stood near him and sent her for thewater ... Meanwhile two of the men began to preparetheir weapons, and a javelin being thrown ... severalmuskets were fired ... (Rutter, 1935: 129-30).

Morrison's ability to recall such details as thenumber of men in the boats, which prisoners wentwith which officers and in which boat isremarkable. Looking at his published account,the amount of detail he is able to recall aboutmany aspects of the voyage suggests he had ajournal, notebook or some paper to make notes.This is corroborated by Captain Edwards' mentionthat he suspected that some of the prisoners wereengaged in a secret correspondence with some ofthe Pandora's crew (Adm. MS 180, Edwards'papers bundle 7). Alternatively, he may later havehad access to one of the Pandora's officer'sjournals; perhaps when he was writing his'memorandum' while a prisoner awaiting hiscourt martial in Portsmouth (Du Rietz, 1986).

Possibly Hamilton's notes or journal were ac-cessible to Morrison, as Hamilton records similar

Coral Sea

„tr:

Rant

^

•`:, •^Wednesday Is

.^....

Little

...^

...... ................^........^ lot Adolphus „â â sea. 9 ' ) ea

^ Adolphus Landed and ate a plum-like iriri;

knives end buttonsBooby I^

. oxo.nged

s,....... ...........

Fres^Bayte=ter^ te•

2 yawls sentPrince of Wares^ -le^hwater

ashore for waterIsland

CAPE YORKPENINSULA

50^100 krn

2 neWawrecksite,^29 August. 1791

4 boats left for Time30 August, 1791

a

16^


SULAWESI

C:4frar

sel

C:2^• 'zBall^Flores^p

Timor........... .................

....

Torres Strait

-X Pandora wrecksite

Coral Seasee inset irri9.3

/AUSTRAUA

To England ---

(crew divided between four vessels)

Sumba Isiah

INDIAN OCEAN

FIG. 12. A, B, Survivors' route to Timor.

details about the survivors' progress from EscapeCay:

... we embarked and laid the oars upon the thwarts,which formed a platform, by which means we stowedtwo tiers of men ... At meridian we saw a cay[Cholmondsley Island] bounded with large craggyrocks ... At eight in the morning, the red and blue yawlswere sent ahead ... to investigate the coast of NewSouth Wales ... On entering a very fine bay[Freshwater Bay] we found most excellent waterrushing from a spring at the very edge of the beach.Here we filled our bellies, a tea kettle, and two quartbottles. The pinnace and launch had gone too far aheadto observe any signal of our success; and immediatelywe made sail after them ... In two hours we joined thepinnace and launch, who were lying to for us ... Afterrunning along, we came to an inhabited island [MountAdolphus Island] from which we promised ourselves a

supply of water. On our approach, the natives flockeddown to the beach in crowds ... we made signals ofdistress to them for something to drink, which theyunderstood; and on receiving some trifling presents ofknives and some buttons cut off our coats, theybrought us a cag of good water. They would nothowever bring it the second time, but put it down onthe beach and made signs to us to come on shore for it.This we declined as we observed the women andchildren ... supplying the men with bows and arrows.In a few minutes, they let fly a shower of arrowsamongst the thick of us. Luckily we had not a manwounded ... We steered from these hostile savages toother islands [Little Adolphus Island] (Thomson,1915: 148).

There is a slight difference between Edwards'account and Morrison's and Hamilton's account


17

FIG 13. The Pandora's boats in tow while making for Timor. (Private collection)

of the survivors' first contact with the Australianmainland, the Torres Strait's islands and theirinhabitants. Edwards glosses over the hostilenature of the encounter.

Hamilton and Morrison are more forthcoming,although they do not distinguish whether thepeople were Aborigines or Torres StraitIslanders. However, Hamilton's mention of thebows and arrows used against the Pandora'sboats suggests that the encounter was with TorresStrait Islanders, as Aborigines did not use thiskind of weapon. From Hamilton's account, itappears that after the hostile encounter on MountAdolphus Island an armed party probably landedon Little Adolphus Island to look for water beforesetting off for Horn Island (called Laforey'sIsland by Hamilton).

Mt Adolphus Island to Timor. Edwards namedLittle Adolphus Island Plum Island, after anabundant fruit — probably the bountiful TonesStrait `nonda' plum — found growing there bythe shore party. They left the island beforenightfall and steered a westerly course towardsthe Prince of Wales Islands, west of Cape York.Several of the islands were named by Edwards,however, not all of these names were adopted tobe passed on and recorded on modern charts.

Edwards recounts how in the evening they:... steered for the islands which we supposed were

those called Prince of Wales' Islands by Captain Cook,and before midnight came to, near one ofthese islands,in a large sound formed by several of the surroundingislands, to several of which we gave names, and calledthe sound Sandwich Sound. It is fit for the reception ofships, having from five to seven fathoms of water.There is plenty of wood on most of the islands, and bydigging we found very good water ... (Thomson, 1915:149).

Although Edwards probably chose the namesof the various other islands making up the Princeof Wales Group, he does not mention any of themin his text. However, Hamilton does mentionthem individually:

We steered for those islands which we supposed werecalled the Prince of Wales' Islands; and at about two inthe morning came to anchor ... along side of an islandwe called Laforey's Island [Horn Island] ... Themorning was ushered in with the howling of wolves ...Lt. Corner and a party were sent at daylight to searchagain for water... As we landed, we discovered afootpath which led down into a hollow ... and ondigging we had the pleasure to see a spring rush out ...On traversing the shore, we discovered a moray, orrather a heap of bones ... among them two humanskulls, the bones of some large animals, and someturtle bones. These were heaped together in the form ofa grave, and a very long paddle, supported at each endby a bifurcated branch of a tree, laid horizontally alongit ... There is a large sound formed here, to which wegave the name of Sandwich Sound [Flinders Passage(AUS 293)], and commodious anchorage for shippingin the bay, to which we gave the name Wolf's Bay,Hammond 's Island lies north west ... Parker's Island[Wednesday Island (AUS 293)] from north by west tonorth and by east ... Sandwich Sound is formed byHammond's, Parker's and a cluster of small islands onthe starboard hand at its eastern entrance [TuesdayIslets (AUS 293)] and near the centre of the sound is asmall dark-coloured, rocky island [Channel Rock(AUS 293)] (Thomson, 1915: 151).

From Hamilton's account, it is obvious that ashore party landed on Horn Island (Hamiltoncalls it Laforey Island) and traversed part of itsnorthern shoreline. It is assumed that the shoreparty landed just west of King Point on its northeastern shore, where the coast forms a shallowbay towards the south (i.e. Hamilton's Wolf'sBay). However, they do not appear to havetraversed very much of the shore as Hamiltonomits to mention Thursday Island, which, if theysaw it, they must have assumed to be part of


Hammond Island because of its height. Fromtheir vantage point at Wolf's Bay the channelbetween Thursday and Hammond Islands wouldnot have been apparent; Thursday Island wouldtherefore have appeared an integral part ofHammond Island.

The survivors made off from Horn Island assoon as they had filled their containers withwater. They left Sandwich Sound by the entrance,now called Flinders Passage, between Hammondand Wednesday Islands (AUS 293), enteredPrince of Wales Channel and last sighted theAustralian coast late in the afternoon when, westof Goods' Island, they finally cleared Prince ofWales Channel (Thomson, 1915: 77) (Fig. 12B).

Although there is no record of third lieutenantThomas Hayward's thoughts at the sight ofPrince of Wales Island receding from view, itwould be understandable if he had been over-heard muttering to himself that as far as he wasconcerned this was the last time he wanted to seteyes on this part of the world. After all, havingbeen with Bligh in the Bounty's launch in 1789, itwas the second time within as many years that hehad been in Torres Strait waters in such dis-tressing circumstances. For as Morrison recounts:

... the heat of the weather made our thirst unsupport-able and as the canvas bags soon leaked out, noaddition of allowance of water could take place, and tosuch extremity did thirst increase, that several of themen drank their own urine ... (Rutter, 1935: 131).

The survivors' progress through the ArafuraSea was comparatively uneventful in spite of theextreme deprivations suffered as a result ofhunger, thirst and the heat of the sun. Some of theprisoners' lot was even worse as they not onlysuffered hunger and thirst, but were also made tolie down on the boats' floor with their armspinioned. Morrison relates how he and Ellisonapparently had done or said something toprovoke Captain Edwards' wrath, and weresubsequently ordered to be pinioned with a cordand lashed down in the boats' bottom (Rutter,1935: 131).

Timor was finally sighted after 12 days in theArafura Sea. The boats eventually made theDutch East India Company's (VOC) settlement atCoupang several days after the survivors hadspent a day and night ashore near a small villageabout seventy miles to the southwest of Coupang,where they bartered for water and food with localTimorese (Thomson, 1915: 156).

Upon arrival at Coupang, their reception byVOC officials was cordial. The Dutch authoritiesdid everything possible to ensure a speedy and

full recovery from the ordeal of the 15 day openboat voyage. They were soon fit enough to makethe journey to Batavia (Jakarta) where theywould be able to embark on VOC vessels boundfor Europe.

During the five weeks which the survivorsspent in Coupang, Edwards took charge ofanother group of prisoners. This group of eightmen, one woman and two children had arrived inCoupang two weeks before the Pandora'ssurvivors. Led by William Bryant, they wereescaped convicts from New South Wales and hadmade the long sea voyage in a small fishing boatfrom Port Jackson along the east coast of Aus-tralia to Torres Strait and across the Arafura Sea(Martin, 1991).Timor to England. While on the journey fromCoupang to Batavia, the ten surviving Bountyprisoners were confined to the lower deck of theVOC ship Rembang, which, but for the exertionsof the Pandora's survivors who helped to crewher, was almost wrecked in a storm off the northcoast of Java. The Rembang was then forced toput in at Samarang for repairs. Initially this delaywas a setback for the Pandora's survivors. How-ever, it later proved beneficial as it resulted in afortuitous reunion with William Oliver's crewfrom the Matavai, who had been lost off Samoaand given up for dead nearly four months earlier.

For William Oliver the pleasure of the reunionwith the rest ofhis shipmates was short-lived. Hisname is to be counted among the thirteen crewlisted as deceased after the survivors reached theDutch East Indies (Appendix 1). Some probablydied as a result of the extreme conditions they hadbeen exposed to during the open boat voyage(Thomson,1915: 145), but most died of diseasescontracted in Batavia, where the survivorsarrived on 7 November 1791. Oliver's second incommand, David Renouard, was lucky to survivethe rigours of Batavia's pestilential climate. Ableseaman Barker, one of the Matavai's crew, haddied before Oliver's group got to Samarang.(Adm. MS 180, Edwards' papers, bundle 7).

Once in Batavia, Captain Edwards made everyeffort to get berths on ships bound for Europe forhimself and his crew. He negotiated with theVOC and agreed to sign on most of the Pandora'ssurvivors as crew on three of the Company'sships returning to Holland from Batavia. (Thom-son, 1915: 81). Edwards negotiated a passage forhimself, George Passmore, Gregory Bentham,two midshipmen and his prisoners, again in ironfetters, on board the Vreedenburg. (Thomson,


1915: 163) The remainder of the Pandora's crewwere divided into three groups, each under one ofthe three lieutenants, and were mustered as crewonto the VOC ships Horssen, Hoornweg andZwaan.

Brown, the beachcomber who had been enteredon Pandora's muster list at Tahiti, was paid off inBatavia. Only one man was too sick to travel andwas left behind in hospital in Batavia. (Thomson,1915: 87). No records about the fate of these twomen have been found to date.

When the Vreedenburg arrived at Cape Town,Edwards transferred his party to HMS Gorgon,homeward bound from New South Wales. Thereare accounts of this voyage, several of whichmention Edwards' presence on board — one byMary Ann Parker, the wife of the Gorgon'scaptain, who apparently enjoyed Edwards' com-pany during the voyage home (Parker, 1795).Unfortunately Mrs Parker omits to mention theprisoners, responsibility for whom Edwardspartially ceded to Captain Parker.

Edwards was the only officer to transfer hisparty, including the escaped New South Walesconvicts, to the Gorgon. The remainder of Pan-dora's crew remained on board the other threehome bound VOC ships as mustered crew andeventually got home to Britain via Holland.Together with sixteen of Pandora's men, RobertCorner was mustered on the Horssen. Theyarrived in Den Briel on 21st July 1792, whereCorner drew an advance of £30 on the Admiralty.He used this money to pay for his men'ssubsistence and their passage to England fromHolland (Adm. 106/1317). The groups underLarkan and Hayward's charge arrived in Englandsometime afterwards; Hayward's group arrivedin Den Briel in the Hoornweg, while Larkan'sgroup arrived in Rammekens in the Zwaan on 12August 1792.COURTS MARTIAL AND SUBSEQUENTCAREERS. HMS Gorgon arrived in England inearly June 1792. The fate of the ten Bountyprisoners remained undecided for another threemonths; they were detained as prisoners on HMSDuke in Portsmouth Roads until their courtsmartial. Hearings started on 11 September 1792.

Michael Byrne, Joseph Coleman, Charles Nor-man and Thomas McIntosh were immediatelyacquitted because William Bligh had vouched fortheir innocence. Tom Ellison, John Millward andThomas Burkitt were found guilty, sentenced todeath and were eventually hanged on board HMSBrunswick in October 1792 (Dening, 1992:

39-40; 46-48). In spite of also being found guiltyof m utiny and sentenced to death, Peter Heywoodand James Morrison were recommended to theKing's mercy and were eventually pardoned;William Muspratt was acquitted on a legaltechnicality.

Peter Heywood remained in the navy and roseto the rank of post captain. He retired in 1819after a distinguished career as a sea-going officer.His service record included duty as signal mid-shipman, master's mate and acting lieutenant onvarious ships in the Channel Fleet until 1798.Upon receiving his lieutenant's commission hewas given command of the brig HMS Amboyna,attached to the India Station at Madras. Whileserving at the India Station he surveyed northernAustralian waters as the commander of HMSVulcan between 1800 and 1801. During this timehe had occasion to visit many islands in the DutchEast Indies, including Timor, where he hadlanded ten years earlier as one of CaptainEdwards' prisoners. In 1807 Captain Heywoodcollaborated with James Horsburgh, the BritishEast India Company's hydrographer, assistinghim with information and the production of thecharts which accompanied Horsburgh's SailingDirections 'for the navigation of the Indian Seas'(Tagart, 1832: 164ff).

James Morrison also stayed in the Navy andhad an exemplary career as a warrant officer,fighting in a number ofmajor naval battles duringthe Napoleonic Wars. He drowned in 1807 whileserving as the gunner on HMS Blenheim. Whilein prison in 1792 awaiting his court-martial, theReverend William Howell, vicar of St John's inPortsea, assisted him with the preparation of ajournal and a separate memorandum detailing hisexperiences of the mutiny in the Bounty, his lifeas a mutineer in the South Pacific and hisimprisonment on board the Pandora (Rutter,1935; Du Rietz,1986; De Lacy, 1997).

Edward Edwards was exonerated for the loss ofhis ship. However, he never commanded at seaagain. He was twice appointed as a 'regulating'(i.e. recruiting) captain; first for Argyleshire andthen for Hull. His various requests to theAdmiralty for another sea command were neveracted upon. He was, however, eventually pro-moted to Admiral's rank (Syrett & Niardo, 1994).A reputed link between Edwards and the'Pandora Inn' in Restronguet, near Falmouth,UK, of which Edwards is said to have been thelandlord during his retirement, has not beenconfirmed. This link appears apocryphal as the


correspondence with the Admiralty, in whichEdwards sought further appointments as a sea-going officer, is post-marked from his Londonresidence or from Edwards' home in Lincolnshire.He died in 1815 and was buried in St Remegiuscemetery in his native village, Water Newton inHuntingdonshire. Edwards' obituary in TheLincoln, Stamford & Rutland Mercury of 21stApril 1815 mentions that he had felt 'to the latestperiod of his life', the effects of the hardships heexperienced during the open boat voyage toTimor.

John Larkan later served as first lieutenant onHMS Defence under Lord Gambier; he sawaction at the Battle of the 1st June (1794) andsoon afterwards was promoted to commander.This was his last sea going appointment. Hereturned to Ireland where he subsequently wasgiven command of an Irish division of SeaFencibles (Marshall, 1825: 250).

Robert Corner appears to have continued in anactive naval career; there is evidence that LloydsPatriotic Fund awarded him one of their pres-entation swords for distinguished conduct whileserving as a lieutenant on board HMS Thisbeduring action against the French frigate Veloce in1803 (May & Annis, 1970: 70). He ended hiscareer as Superindendant of Marine Police inMalta and died in 1820 (Marshal,1825: 38).

Thomas Hayward's naval career was cut shortwhen he perished with his crew, while in com-mand of the sloop HMS Swift, which was lostwith all hands during a typhoon in the SouthChina Sea in 1797 (James, 1886: 462).

George Hamilton published an account of thePandora's voyage in 1793 (Thomson, 1915;Kenihan, 1990; Hamilton, 1998). Upon his returnto Britain, his next naval appointment was assurgeon on board HMS Lowestoff, which waspart of the British Mediterranean fleet under LordHood. Hamilton was discharged in April 1794having lost an arm, probably after being woundedin action against French fortifications on Corsica(Pigott, 1995: 24ff). It is assumed he returned tohis native Northumberland sometime during 1794,where he lived in retirement on a naval pensionuntil his death in 1796 (Kenihan, 1990: xvii).

George Passmore's career did not remainunblemished; he appeared before a court martialin 1793. The nature of his offence is not recordedbut was apparently unrelated to his conduct as thePandora's master. Whatever the offence, it wasserious enough to cause the Navy Board todecline an appointment as master of HMS

Daedalus, in spite of representations from thecaptain and recommendations by Admiral SirCharles Knowles (Adm. 106/1317).

The rest of the Pandora's crew have recededinto historical obscurity, although undoubtedlyanecdotal information about some of them, e.g.on master's mate George Reynolds, is extant inprivate collections as well as in public recordsrepositories. Archival information on many ofthe others should also be recoverable, as theirnames will appear in such documents as musterlists and pay books, as well as in private collect-ions, or obituaries in contemporary newspapers.

However, what remains of them in a verytangible way are their personal possessions andprofessional equipment — their material culture— which was lost with the Pandora. Much of thisappears to have been preserved within and aroundthe hull and can be retrieved using systematicarchaeological methods and analysed and in-terpreted using models developed in materialculture studies. This, then, is the rationale forconducting an archaeological investigation of thePandora wreck, which provides the functionalcontext of a well preserved collection of objectsin use by a microcosm of late 18th century Britishsociety (Rodger, 1986: 14).

Finally, as far as the Pandora is concerned,neither the Admiralty nor the colonial authoritiesin New South Wales apparently made anyattempts to salvage material from the wreck.There is no evidence that the wreck was disturbeduntil its rediscovery in November 1977 bydocumentary film makers Steve Domm, JohnHeyer and Ben Cropp, with Royal Australian AirForce assistance.

ARCHAEOLOGICAL INVESTIGATION

SITE DESCRIPTION AND DISTRIBUTIONOF THE WRECKAGE

LOCATIONAL INFORMATION. The wreck islocated within Pandora Entrance (Figs 10, 12B),approximately 5km to the NW of Moulter Cay.This sandcay is on the outer Great Barrier Reefabout 140km ESE of the tip of Cape York innorth-eastern Australia.

Although the wreck lies inside Pandora Entrance,it is exposed to the force of swells from the CoralSea generated by easterly winds, prevalentbetween July and December. Within a radius ofapproximately 200m to the SE, E, N and W, thewreck is surrounded by four small submergedreefs which provide some protection againstocean swells. These reefs also deflect the flow of

AI,ICHOR?

SANCHORP BAKERG HENDERSON

10 METRES

CHAIN PLATE 9

/ SHEATHING

COPPERPUMP TUBING

ANCHOR?,

C)^CANNON

CANNON^CANNON

CANNON

Us

00

11‘RUDDERFITTINGS

0 'kANN

C,^Q^- CHAINPLATES

CAPSTAKI

,) ,V) \lcH°R pCHAIN PLATE

COWLING?


21

FIG. 14. Pandora wrecksite 1979: predisturbance site plan drawn from photomosaic. (P. Baker & G. Henderson)

tidal currents. The pattern of these currents hasnot yet been determined definitively, althoughpreliminary measurements have been recorded(Ward et al., 1998). However, divers have exper-ienced their considerable strength, sudden onsetand unpredictability. Consequently conditionsfor marine archaeological recovery operationsare challenging. Underwater working conditionsare further complicated by depths of between30m and 35m which impose limitations on (airbreathing) diving operations.

An area with a radius of 500m, centred on thesite at the intersection of latitude 11°2240S andlongitude 143°59'35"E, has been declared aprotected zone under Section 7 of the Common-wealth's Historic Shipwrecks Act (1976). Entryto this zone is subject to a permit issued by theDirector of the Queensland Museum, who is theMinister's delegate under the Historic ShipwrecksAct (1976), responsible for historic wrecks inCommonwealth waters off Queensland.

SITE DESCRIPTION. The bulk of the wreck liesburied within a layer of predominantly carbonatesand which contains a significant proportion ofalgal, coral, mollusc and formaniferal detritus.This layer is thought to be up to 3m thick, extend-ing in all directions around the site and appearingto be fairly mobile. To date the depositionalhistory of this sand layer is imperfectly under-stood, although preliminary analysis has beenundertaken (Ward et al, 1998).

From the ordered distribution of wreckage onthe seafloor it is evident that the hull was intactwhen it sank, coming to rest on the seabed on itsstarboard side. It appears to have settled into,and/or have been gradually buried by, deposits ofcoralline sand. However, while these depositswere building up within and around the hull, thevessel's top sides, including its upper deck levels,appear to have disintegrated as a result of marine

borer activity. Currents, and to a lesser extent,wave motion probably also played a role in itsdisintegration. The ordered distribution on theseafloor of the visible features of the wreck attestto a process of sediment accumulation in andaround the lower hull and the gradual disin-tegration ofthe hull's exposed upper deck levels.

The most prominent features of the site areseveral large iron objects standing proud of theseabed, including a bower anchor, the vessel'sBrodie stove, a number of cannon and variousother large, coral encrusted objects. There isnothing random about this distribution; itindicates that these heavy iron objects droppeddown directly from their original positions on thequarterdeck, focsle deck and upper deck as thevessel disintegrated (Fig. 14).

The bower anchor, which is lying adjacent tothe sternpost, is thought to be the anchor dropped'underfoot' to steady the ship after the Pandorahad been refloated from the reef by her crew(Thomson, 1915: 72; United Services Magazine,1843: 418).

The buried hull remains and the bulk of theartefact assemblage lie within an area 20m x 50mon a gently (SE) sloping, featureless, sandyseafloor. This 1,000m2 area is the main site.Depending on the height of the tide, depths varybetween 30-35m. The outline and orientation ofthe hull remains are well delineated within thesite. This applies especially to the hull's stern,which is indicated by the large, readilyidentifiable, section of copper sheathing whichcovered the hull's sternpost. The visible part ofthe sternpost is lying on its starboard side re-vealing the embossed draught marks and'XIV' of its port face (Fig. 15).

Two attempts have been made with remotesensing equipment to determine the extent of thehull remains buried in the seabed (Fig. 16).

FIG 15. The Pandora's stempost sheathing (portside view) with em-bossed draught marks.

22^


However, these efforts only gave indicationsabout the outline and edges of the hull remains.Together with evidence obtained by spot trench-ing, especially in the stern and midships areas, itappeared that 25-30% of the hull was preserved(Gesner, 1993).

Although it is reasonable to assume that thisoutline is accurate, there are still a number ofunanswered questions about the structural

integrity of the hull remains.However, a very coherent andwell-preserved artefactassemblage — the Pandora'scontents — definitely lies buriedwithin, around and under thepreserved lower hull.

There are also indications thatcurrents may have swept some ofthe Pandora's smaller, lightercontents, especially from theupper deck, away from theconfines ofthe preserved hull. It isassumed that a dispersal trail,consisting primarily ofpartially orwholly buried small artefacts, willlead away from the wreck in thedirection of the strongest currentset. Although some observerssuggest this is a south-westerlyset, this has yet to be determined

by more reliable oceanographic methods.

IDENTIFICATION. A survey to confirm theidentity of the wreck and assess its archae-ological significance was commissioned by theCommonwealth government in 1978. GraemeHenderson, the Western Australian MaritimeMuseum (WAMM) maritime archaeologist whoconducted that survey in April 1979, sub-sequently made recommendations about future

00FIG. 16. I sometric drawing of the Pandora wrecksite with expected outline and orientation of the hull remains.

(B. Jeffery)


23

FIG 17. A, bronze rudder pilule with embo ssed broad arrow and punched number '24'; B, punched '24'(detail).

use and management of the wreck (Henderson,1979).

Rudder fittings from the wreck, retrieved byDomm, Heyer & Cropp in November 1977, weresent to the WAMM's laboratory for conservationand analysis. One of these fittings, a pintle, wasmarked with a series of dots punched onto thesurface in the form of the number '24'. It also hadthe name 'FORBES' in raised letters, in additionto a raised broad arrow identifying the pintle asthe property of the British government (Fig. 17A,B). These were the most positive clues that thewreck was a 24 gun Royal Navy frigate of thePandora's size. Archival information sub-sequently provided confirmation that the wreckwas the Pandora, as a foundry operated byWilliam Forbes was documented as a supplier offittings to the Deptford yard where the Pandorawas built in 1778.

After the initial survey, WAMM photographerPatrick Baker compiled a photomosaic of the site(Fig. 18). By analysing the distribution of objectsrecorded on the photomosaic, Henderson con-cluded that the wreck had been intact after it hadsettled onto the seabed. Although its topsidesappeared to have disintegrated, it was assumedthat its lower levels would very likely besubstantially intact and would certainly contain awell-preserved collection of artefacts represent-ing the cargo and the crew's possessions.

Three of the main identifiable features in thephotomosaic were the ship's Brodie stove, thePandora's best bower anchor and a section of thesternpost. Draught marks on the sternpost inRoman numerals indicated not only that a sub-stantial part of the lower hull probably lay buriedin the seabed but also that the hull had settled ontoits starboard side.

The Pandora wreck was regarded to be ofconsiderable archaeological significance.

Because its position had become known, extrasurveillance and additional protection wereinvoked in July 1981 through Section 7 of theHistoric Shipwrecks Act (Henderson, 1986:129-35). Excavation of the wreck's contents wasalso recommended (Henderson, 1979: 33-4).

Responsibility for management of the wreckwas subsequently delegated to the QueenslandMuseum, which established a Maritime Archae-ology Section in April 1982 and appointed ascurator its senior exhibition designer, RonaldColeman.

FIVE SEASONS OF EXCAVATION

DELINEATING THE EXTENT ANDCONDITION OF THE WRECK:

1983 SEASON

October 1983 saw the beginning of the firstseason of archaeological excavation. The Com-monwealth Government, through the Departmentof Home Affairs & Environment, funded the ex-cavation. Planning had started at the QueenslandMuseum in 1982 under the direction of RonColeman. In terms of personnel the expeditionwas a national effort, with participation of vol-unteers as well as professionals from museumsand State heritage management agencies inWestern Australia, South Australia, Victoria andTasmania. The archaeological excavation wasdirected by the Western Australian MaritimeMuseum's (WAMM) maritime archaeologistGraeme Henderson.

Sailing from Cairns in a converted prawn trawler,the RV Flamingo Bay, a 22 person archaeologicalteam arrived in Pandora Entrance in late October1983. Using Baker's 1979 photomosaic (Fig. 18)as a guide, the expedition's main objectives were:1, to establish a permanent grid reference systemover the main area of wreckage; 2, to determine,

24^


FIG 18. Predisturbancephotomosaic. (P. Baker, 1979)

by limited trenching, the extent and orientation ofthe hull remains; and, 3, to assess the conditionand range of artefacts to be expected from the site(Henderson, 1986: 136ff).

METHODS. A reference grid network wasestablished over a 20m x 50m area of the wrecksite.

A network of reference poles was establishedaround the wreck using a pre-measured spacingdevice called a 'horizontal plumb frame', whichenabled poles to be erected at 5m, 2m or lmintervals. They were guided into the seabed withanother pre-fabricated device, a 'vertical plumbframe', which ensured that the poles were placedvertically.

Using a tensioned stainless steel wire as aguideline, three parallel lines of 50m length, each10m apart, were set out. The first line wasdesignated as the primary baseline and was usedto establish, by trilateration, the other two parallellines. Nine grid reference poles were establishedat 5m intervals along the primary baseline andnumbered according to their distance along theX-axis of the site (Fig. 19). Six more referencepoles were established by triangulation from theprimary baseline along each of the other twoparallel lines.

The grid reference poles were subsequentlylabelled with their X and Y co-ordinates. Datumwas marked on the 0.0 pole and then transferredto all other grid reference poles. This datumserved as the benchmark for recording anartefact's Z co-ordinate. Site profiles were ob-tained by measuring, at each of the referencepoles, the height of the datum mark above theseabed.

Pre-fabricated rigid, aluminium grid frames,measuring either 2m x 5m or 2m x2m were erectedover three selected areas of the wreck site.

Trenches were excavatd by scuba diving archae-ologists using airlifts (Green, 1990: 130-5) withinthe grid frames. Artefact location was recordedwith reference to the corners of the grid frames.

Using Scuba equipment, a team's dive timewas usually limited to 18 minutes 'bottom time'on the first dive of the day. Most diving teamswere able to carry out a second dive in the lateafternoon; however, due to nitrogen loading fromthe first dive, bottom time was limited to amaximum of 15 minutes. Depending on the typeof work a diver was performing, an individual'sfirst dive was sometimes less than 18 minutes as adiver could run low on air before available timehad elapsed. Decompression diving extended adiver's time on the bottom, sometimes by 10minutes; however, this was not a satisfactorysolution to increasing dive time because a sub-sequent decompression stop at 6 or 3m depth foranother 20 minutes was required, thus preventingother divers from descending while decompres-sion was being completed by the previous team.

RESULTS. Trench /. The first probe, a trenchmeasuring 2mx5m, was excavated around thesternpost (Fig. 19). The rationale for this trenchwas to dig from the foot draught mark,which was visible at the seabed surface, downalong the port face of the sternpost (Fig. 15) andprogressively follow the draught marks to exposethe junction between the sternpost and the keel.Due to the almost horizontal lie of the sternposton the seafloor, its edge initially trended at aslightly oblique angle into the seabed. However,at approximately the 'IX' foot draught mark agreater angle was observed, indicating that thejunction between sternpost and keel lay sub-stantially deeper in the seabed; approximately 1.5to 1.8m (Fig. 20).

HMS PANDORA: FIVE SEASONS OF EXCAVATION

50.20

40.20

30.20

> N

50.10

-20.10

20.20

10.20Trench 4^',Trench 3

SternpOs

10.10

0,10^ Stersrt anchor

- - ----

-

•• 4,11

••

•

•s_s„. So-.1 0 Brodie stoviii3

114---

Tench 2

Bow anchor^50.0

40.0

25

0.20

0.0

s-s—•

20.0

5

30.0

0.0rn

FIG. 19. Pandora wrecksite: grid reference network and trenches excavated during the 1983 expedition.

Within several days of probing in trench 1, itbecame apparent that the junction of the sternpostand keel would not be reached easily. Due to thesediment's loose and coarse-grained composition,it could not be contained without shoring thetrench's sides. Thus, with constant collapse of thesides, a wider than intended area was exposed. Adeep trench over a small area of the wreck couldtherefore not be excavated.

Within one week of excavation it was obviousthat deeply buried hull remains, especially thekeel, keelson, deadwood and heel of the stern-post, would not be exposed easily. As sheathingfrom the hull's portside became exposed, it alsobecame apparent that at approximately the 'IX'foot draught mark the sheathing from the portside of the hull had been pulled down in crumpledlayers toward the hull's centre line. Thisindicated that the port timbers had collapsedtoward the centre line of the hull (Fig. 20).

The dense concentration of artefacts exposedas the sediment was removed also preventedpenetration into deeper layers of the wreck. Themajority of diving time was spent on recording inthree dimensions the location of the exposedartefacts before they were retrieved from theseabed.

The trench extended no deeper than 0.5m intothe seabed. It was estimated that the junction ofsternpost and keel lay under a layer of sediment atleast another meter thick; consequently theorientation of the hull remains was not deter-mined during the first season.Trenches 2 and 3. A second trench (2m x 5m) wasstarted in the bow area (Fig. 19). It appeared tocontain substantial timbers at a depth of approx-imately 0.5m. However, the timbers were notactually sighted as they were covered by coppersheathing; but tactile probing suggested theirpresence under the sheathing. A distinctivefeature of this area of sheathing was that it openedoutward toward the port side, whereas in the sternsection the portside sheathing had collapsedtoward the centreline of the hull. This suggested apossible twisting of the hull as it broke up, i.e. thehull's sides falling away from the centreline in thebow and falling toward the centreline at the stern.

In addition to sheathing remains, the bow trenchcontained two large earthenware storage jars(MA138 & 139) as well as several fittings fromthe rigging, remnants of an iron box and a full setof spare bronze rudder fittings. Small areas oflarge section timbers were briefly sighted underthe earthenware jars; however, dimensions couldnot be adequately recorded as the timbers were

2

4


FIG 20. Hypothesised collapse sequence of the stempost and port side stern timbers.

too deeply buried and time constraints preventeda concentrated effort.

Subsequently, a 2mx2m trench (Trench 3) wasexcavated in the midships area around a cannonlying on the sea floor.

This midships trench contained a spectaculardiversity of very small and fragile objects, denselyclustered under one of the twenty '6-pounder'cannons (MA 148) with which the Pandora wasarmed. More than 120 objects were retrievedfrom the sand under and immediately adjacent tothe cannon. These objects were all retrieved fromthe top (i.e. <50cm) layer of sediment (Fig. 21).

Artefacts retrieved during the 1983 season weregiven field numbers and later accessioned into theQueensland Museum's Maritime ArchaeologyCollection through assignation of `MA numbers'(i.e. Maritime Archaeology Collection) MA010-157 and MA300-354.

Other Projects. Before excavation started, anetwork of grid reference poles was erected overan area of 20m x50m. The network was estab-lished by hammering into the seabed threeparallel 50m lines of poles, each spaced at 10m

intervals and each line comprising at least sixpoles. The poles, each 4m long and with adiameter of 40mm, were driven in vertically to adepth of about 1.25m (Fig. 19).

A photogrammetric project was undertaken byPaul Clark. Using a specially constructed phototower, to which a camera bar was fitted con-taining two Nikonos III underwater cameras (Fig.22), Clark methodically covered the entire 20mx50m wrecksite and obtained 250 stereo-pairs ofphotographs.

DISCUSSION. Although not entirely successfulin terms of providing definitive informationabout the extent of the hull remains, the 1983season did give a wealth of information about therange and condition of the artefact assemblage.The survival in the midships trench (Trench 3) ofsuch fragile organic material as a pencil (MA81)indicated that at deeper levels of the wreck therewas a favourable environment for the preserv-ation of ship's timbers and other, more fragileorganic objects. This was confirmed by the soundcondition of timbers briefly sighted under anearthenware olive oil jar in the bow trench(Trench 2) at a sediment depth of nearly 0.55m.


033 40 53^• 39

• 152 • 78

• 76• 55

• 54

• 85• 38

• 72

• 36

• 46

• 105

•88

• 86

• 118

• 116 •112102 •104^•91• 0103 on

• 94

• 125

5980• 6 .95

• 40^ •48• 51

• 43

• 42

• 75

• 74• 114

• 111

• 81

• 117• 115

.79 032

041

• 49

• 58

• 82• 45^

• 77• 83

• 84

• 98

• 64 •70• 65

• 59^

•146• 151

• 66• 31 •67

560^•69

092

148

• 73

• 89

• 97

• 107• 108

FIG. 21. Artefact plot in trench 3 (1983 season). MA31, bronze coak; 32, stoneware jar; 33, stoneware jar; 36,stoneware jar; 38 wrought iron concretion; 39, six case bottle sherds ; 40, creamware plate sherd; 41, sixstoneware jar sherds ; 42,101 musket balls; 43, six musket balls; 45, stoneware jar; 46, eight pane glass sherds;48, two wine bottle sherds; 49, apothecary bottle; 51, glass bottle sherd; 53, coconut husk pieces; 54, three casebottle sherds; 55, two glass bottle sherds; 58, timber fragment; 59, four cast iron round shot ; 60, two pane glasssherds; 61, wide-mouthed bottle sherd; 64, seven case bottle sherds; 65, stoneware jar; 66, unidentified organicmaterial; 67, eight wood fragments; 69, stoneware jar with cork; 70, five wood fragments; 72, hairbrush/shoehorn; 73, ivory brush; 74, syringe; 75, turned ivory knob; 76, case bottle base; 77, two brass buttons; 78,brass buckle; 79, three bone buttons; 80, three wood fragments; 81, wooden pencil; 82, pane sherd; 83,apothecary bottle; 84, butt plate for musket; 85, trigger guard; 86, Polynesian wooden tool; 87, case bottle sherd;88, glass bottle sherd; 89, two case sherds; 90, neck and shoulder case bottle; 91, wide-mouthedbottle sherd; 92,twelve musket balls; 94, three barrel stave fragments; 95, two bricks; 96, case bottle ; 97, wood fragment; 98,wood fragment; 102, tourniquet clamp; 103, clear glass sherd; 104, ivory brush handle; 105, glass bottle stopper;107, two bottle sherds; 108, case bottle base; 111, eleven case bottle sherds; 112, glass bottle base; 114, licecomb; 115, glass bottle stopper; 116, stoneware jar; 117, two glass bottle sherds; 118, copper sheeting fragment;125, brick; 146, cannister shot; 148, cannon; 151, wrought iron bolt; 152, bar shot.

It was obvious that the context of the artefactsand their condition and diversity called for asystematic approach to artefact retrieval and tothe recording of locational information. The im-mediate challenges were to decide on the mostefficient method of making an accurate spatialrecord of artefact distribution and how to interpretthe apparent jumble of material.

Upon analysis of the spatial artefact inform-ation. many of the finds from the midships trench(Trench 3) appeared to be associated with thePandora's surgeon, George Hamilton. Thisinterpretation was supported by 18th century navalarchitect's plans of a vessel of the Pandora'sclass. These plans show where cabins werelocated on the lower deck and to which officers,including the surgeon, they had been assigned

28 MEMOIRS OF THE QUEENSLAND MUSEUM

0^1^2rn

FIG. 23. Plan of lower deck showing the linear distanceon deck between the surgeon's cabin and the stern-post. 1, Rudder; 2, sternpost; 3, breadbin; 4, Purser'scabin; 5, Master's cabin; 6, Surgeon Hamilton'scabin; 7, 2nd Lieutenant Corner's cabin; 8, 1st Lieu-tenant Lakan's cabin; 9, 3rd Lieutenant Hayward'scabin; 10, removed cabins.

SYSTEMATIC EXCAVATION: 1984 AND1986 SEASONS

FIG 22. Stereographic phototower in use.

according to the 'establishments', i.e. standingAdmiralty instructions (Lavery, 1987: 156-68).

With this information, a reliable clue wasavailable about the orientation of the buried hullremains and about the deposition of artefactswithin the hull remains. Thus, working on theassumption that the midships trench (Trench 3)area corresponded with the foremost portsidecabin on the lower deck, assigned to an officer ofGeorge Hamilton's rank (a warrant officer), thedistance between the midships' trench and thesternpost was thought to be indicative. Allowingfor some displacement due to the wrecked hull'sangle of heel to starboard, which likely wouldhave caused artefact clusters from the port side ofthe lower deck to be deposited toward the hull'scentre line, the distance was found to approx-imate the actual distance between the sternpostand the surgeon's cabin on the naval architect'splan (Fig. 23).

1984 SEASON. Excavation commenced inNovember. This second major season was againfunded by the Commonwealth Department ofHome Affairs & Environment and organised bythe Queensland Museum. This expedition wasalso a national effort with a number of volunteersand professional maritime archaeologists andconservators seconded to the expedition frommuseums and heritage management agencies inWestern Australia, South Australia, Tasmania andQueensland. Overall project coordination wasagain by the Queensland Museum's maritimearchaeology curator Ron Coleman, with archaeo-logical direction again provided by the WAMM'sGraeme Henderson. A team consisting of 22diving archaeologists, photographers, conserv-ators and technicians spent 35 days at the wreckusing as a mother vessel the MV Reef Seeker, a65-foot catamaran, purpose-built for divingexpeditions on the Great Barrier Reef.

METHODS. During the 1984 season a system-atic approach to excavation was adopted. Thewreck site was divided into 250 numbered gridsquares, each measuring 2m x 2m (Fig. 24). Thesewere used as fixed reference areas and alsoserved as a guide for excavation teams. The gridsconsisted of rigid aluminium angle-line lengths,each cut, mitred and welded to form 2mx2msquares. Grid squares could be bolted to eachother on any side; thus allowing for pairsextending along the site's X axis or Y axis.Individual grid squares had four 40mm holes cutinto their corners allowing them to be fitted over

HMS PANDORA: 5 SEASONS OF EXCAVATION^

29

50m

0.0

FIG. 24. Grids superimposed over the estimated outline of the hull remains Primary excavation areas: 1984 —grids 1-12, 25-32, 45, 47, 49, 51 & 52 excavated; 1986— grids 61-76; 1993— grids 104, 106 & 125 excavated;1995— grid 70.

the aluminium grid reference poles or theintervening grid supports. A gusset in each of thecorners containing stainless steel set screwsprovided for levelling of each grid squarethrough tensioning of the set screws against thevertical reference poles.

Scuba equipment continued to be used for alldiving operations. This enabled most divers todive twice a day for a total of approximately 32minutes per day; typically a first dive wouldcomprise 18 minutes bottom time; the day's sec-ond dive would be for approximately 14 minutes.Two weeks into the fieldwork, decompressiondiving was also conducted, extending total dailytimes spent on the seabed by an additional 8 to 10minutes.

Divers used a water dredge to remove sediment(Green, 1990: 136). The water dredge was divercontrolled as well as surface controlled. Diverswere able to regulate suction on the seabed byoperating a valve, and at the surface the amountof water pumped down to the water dredge wascontrolled by a throttle. Sediment was directed toa spoil heap which, after systematic sifting, wasredeposited as backfill into the trench during thelast days of excavation work.

RESULTS. Excavation started in grids 1 & 2 andproceeded systematically to grid pair 11 & 12,moving along the site's Y axis; i.e. approximatelyathwartships (Fig. 24).

Although grid square 1 lay at least 12m to theSE of the estimated outline of the hull remainsand at least 18m away from the artefact clusterthought to correspond with the surgeon's cabinarea, it was deliberately selected as the startingpoint because allowance had to be made for theknown starboard heel of the hull. It was assumedthat this heel may have caused deposition ofartefacts well outside the estimated starboardedge of the hull remains Also taken into conside-ration was the effect of tidal currents which werelikely to have swept artefacts from the dis-integrating hull, depositing them either side ofthe hull remains.

During the first ten days of excavation noartefacts were located in the top sediment layer(i.e. <0.5m) in grids 1-12. The direction of ex-cavation was therefore changed to that of parallelto the site's X axis (i.e. approximately along theship's centre line) and so excavation of grids13-20 was not started. This excavation was doneby erecting grid pairs 27 & 28 and grids 30-34, to

FIG 25. Bulkhead fireplace in-situ (grid 28).

30^


which grids 49-52 and 71 were later added. This'leap frogging' of grids was achieved by boltinggrids 29 & 31 to the sides of grids 10 and 12 andby bolting grid 27 to grid 8 (Fig. 24).

During the third and fourth weeks as excav-ation progressed toward the centre of the site, i.e.into grids 30-34 and 47-52, increasing numbersof artefacts were revealed. Artefacts were locatedin the top sediment layer (i.e. <0.5m depth) withinall of these grid squares. Among the objectslocated directly adjacent to the estimated outlineof the hull remains was a bulkhead fireplacewhich was uncovered lying directly on top of thebower anchor next to the sternpost (Fig. 25).

A substantial diversity of artefacts was retriev-ed. Using Nick Rule's 'direct survey method', theywere co-ordinated to the nearest centimetre in threedimensions (X, Y and Z axes).

Artefacts retrieved during the 1984 season weregiven temporary field numbers and later acces-sioned into the Queensland Museum's MaritimeArchaeology Collection through assignation ofaccession numbers MA500- MA779.

DISCUSSION. The 1984 season proved thatsystematic artefact retrieval within the gridsquares was a useful and rewarding approach.Plots of artefact location revealed progressivelylarger concentrations closer to the estimatedoutline of hull remains; especially in grids 30-34and 47-52, 55 and 71 (Figs 26 & 27).

Most importantly, the first of the Pandora'ssurviving timbers, later identified as ceiling,

frames and outer plankingon the starboard quarter,were uncovered in grid 71during the last day ofexcavation. This was thefirst direct physical con-firmation of the expectedoutline of the hull remains.

Excavation in grids 52and 55 exposed the edgeof copper sheathing on thestarboard quarter as wellas the sheathing whichhad been attached to thedeadwood and under thePandora's starboardtransom. Although notransom timbers, dead-wood or outer plankingwere encountered, thesheathing appeared tohave retained its original

position and shape. This was apparent from thealignment of the sheathing tacks which indicatedthat the planking had simply deteriorated afterthe weight of the transom timbers had causedcollapse of the stern (Fig. 20).

An analysis of fastenings retrieved during theseason was undertaken by Mike McCarthy of theWAMM. This analysis also contained a dis-cussion about the amount of hull likely to bepreserved; estimated at approximately 25% ofthe original (McCarthy, 1983).

A shallow seismic survey, using 3.5KHz equip-ment, was conducted by consultants Dr DavidJohnson and Kevin Hooper from James CookUniversity's Geology Department. The surveydelineated a 'tapering, tubular object beneath thesite' and concluded that while it is possible thatthis object represented buried coral growth orfallen reef blocks, it is most likely to be theremains of the ship now covered by surficialsands up to 5m thick' (Johnson, 1984: 2).

1986 SEASON. The third Queensland Museumexpedition was an international effort, funded bythe Commonwealth and Queenslandgovernments, again under the leadership of RonColeman. Archaeological direction was by theWAMM's Graeme Henderson, with co-directionby the author, then the Queensland Museum'sassistant curator of maritime archaeology. Theexpedition team was assisted by a group of inter-national volunteers from the UnitedKingdom-based youth adventure training

418 4618

HMS PANDORA: 5 SEASONS OF EXCAVATION^31

• /:^stern^t

cannon

••

• \^•. •stendard

•

N'near o^. '

—^tTit'Zi 0sheathing—vvheeJ

existing —sheathing

M=C11111:7171•=10 1^2 3â 5m

/ Pbowcons tion

brodiestove

--

existingsheathing

bow

4.0 46

FIG. 26. Artefact plot recorded during the 1984 excavation. A, parts of the ship; B, ship's furniture and fittings; C,weapons and accessories, tools and instruments, domestic equipment, utensils and accessories, clothing andaccessories, artificial curiosities.

/stem ost• •

'cannon•

ptlandriiisternanchor^N^••••

I •

cop*;

pump,

•sheathing.

•pulleyV

—wheel•

0^1 2^3^4 5rn

sternanchor

/^•

• -SlernOtoal

•

•

•

< stand4f

sheathing.•^ pu1140—wheal

• •••

1111=111101101111=111=10 1 2 3 4 5m


4.18

4.0

4.0

organisation Operation Raleigh (OR). This wasthe largest expedition to the wreck site,involving 37 divers, 17 ships' crew and 20 ORstaff and assistants. In terms of divingoperations, the expedition was also the mostcomplex and challenging to date, due partly tothe number of divers and partly to the oversizedmother vessel, OR's Sir Walter Raleigh.

RESULTS. Using the same Scuba diver-controlled water-dredging techniques as theprevious seasons, systematic excavation withinthe grid system was continued. In a six weekperiod during October and November 1986, a4mx16m trench, encompassing grids 61-74,was excavated (Fig. 24). Excavation started atthe baseline in grids 61 & 62 and the trench wasprogressively expanded by addition of gridpairs along the Y axis until grid pair 73 & 74was reached. Depth of excavation increased asthe estimated outline of the hull remains wasapproached. Thus, in grid pair 61 & 62sediment was removed to a depth ofapproximately 0.3m, whereas in grid pairs 69 &70 and 71 & 72 the depth reached was nearly1.4m. It was felt a trench transecting the hullremains would give a good indication of theextent of the preserved hull.

Compared to the 1984 season, a similardiversity of artefacts was retrieved. As ex-cavation approached the edge of the preservedhull, a similar density of clustered artefacts wasalso encountered. The grids spanning areaswithin the confines of the hull remainsappeared to contain a greater proportion ofpersonal possessions. This was apparent ingrids 69 & 70, where artefacts were clusteredon top of the lower deck planking and againstthe side of the vessel (i.e. the spirketting) in anarea which corresponded approximately withthe aft-most cabin on the starboard side of thelower deck. If the establishments were followedduring the Pandora's South Pacific voyage,this cabin was probably occupied by thePandora's second lieutenant, Robert Corner(Fig. 27).

The timbers uncovered in grid 71 during theprevious season were re-exposed. In doing so,more timbers were uncovered in adjacent gridsquares. Grids 68-70 revealed the wreck's first

A

srern ost

stemanchor

Pump ispollay

existingsheathing

MI=110======10 1 2 3 4 5m

418

418

FIG. 27. Artefact plot recorded in the 1986 season. A,parts of the ship; B, ship's furniture and fittings; C,weapons and accessories, tools and instruments,domestic equipment, utensils and accessories,clothing and accessories, artificial curiosities.

4.0


Area of

—--■11011K

crumpled sheathing

\

D_0^73

Mizzen mast

1

A!>C7L,.$i....^„,4 1

---,„^„Ill

dg^knee 72

(I!^W^/^\44 Carronade

•.. alk figr. . . _----■ ... ...-----._.

110rî

--------;----.„

68

FIG. 28. 1986 trench after excavation. B = deadwoodbolts; C = concretion; F = frames; P = outer planking;S = starboard sheathing; W = lower deck waterway;SP = spirketing.

immediately identifiable timbers, which includ-ed outer planking, a lower deck lodging knee,cant frames, ceiling planking and remnant of themizzen mast (Fig. 28).

Most importantly, grid 68 revealed the top edgeof the uppermost run of starboard sheathing, in-dicating that in the starboard quarter the hull waspreserved from slightly above the waterlinedown to the keel. The sides of the hull along theport quarter were not exposed as any remainingtimbers and plank from this side were obscured

under sheets of crumpled sheathing in grids 73 &74 (Fig. 29).

Although not the easiest of seasons in terms oflogistics and coordination of diving operations,the 1986 season was more productive than the1983 and 1984 seasons. Seven hundred andeighty-six artefacts were retrieved over a fiveweek period, compared to 256 in 1983's six-weekexpedition and 283 in 1984's five-weekexpedition. Almost 50% of the 1986 collectioncould be provenanced to grids 69 & 70. As duringprevious seasons, the 1986 collection was acces-sioned into the Queensland Museum's MaritimeArchaeology Collection through assignation ofaccession numbers MA 1000-1786.Other projects. Photogrammetry again featuredprominently as a project during this season(McCafferty, 1988). Several features of thepreserved hull were shown prominently on aphoto-mosaic (Fig. 30) which was compiled fromphotographs of the excavated trench (grids63-74). Among the features clearly delineated onthe photo-mosaic were the starboard edge of thesheathing, starboard cant frames, inner planking,a lower deck lodging knee, a lower section of themizzen mast as well as several large copper boltsused to fasten the deadwood to the keel. Thesebolts were considered a clear indication of theheel to starboard of the hull, 34° off the vertical(Fig. 31).

DISCUSSION. Artefact distribution in the 1986trench revealed that larger concentrationsoccurred along the starboard edge of the hullremains. This clustering of artefacts was obviousagainst the inboard edge of the hull remains ingrids 69 & 70, suggesting that, as the vessel sankand settled onto the seabed, objects kept in thewardroom and the second lieutenant's (RobertCorner's) cabin had fallen onto the (lower) deckand collected on the deck planking against thestarboard side of the hull (Fig. 27).

Comparison of the artefact plots made from therecords of the 1986 excavation and from the 1984season's plots, indicated differences between thetypes of artefacts lying inside, and outside, thehull remains. Those inside the hull remains werepredominantly small and light, or very large andheavy objects (e.g. cannon). Artefacts lying out-side the hull were mainly the ship's fastenings,such as heavy bolts, or internal furnishings in-cluding the bulkhead fireplace and a sandstonewater filter. The latter items had both probablybeen installed in the great cabin on the upperdeck. However, the main differences were their

FIG 29. Crumpled portside sheathing in grid pair 73-74.

Ammimmons

W "."-° 3

1. 4

FIG. 30. Photo-mosaic of the 1986 trench compiled from stereo-photographs. (M. McCafferty and AAMSurveying & Mapping)

34^


original function. The small, light artefacts found as personal possessions or as professional equip-within the hull could predominantly be classified ment belonging to the Pandora's officers; most

35HMS PANDORA: FIVE SEASONS OF EXCAVATION

FIG 31. Row of deadwood bolts (looking aft) showing the wreck's 34° angle of heel to starboard.

likely to second lieutenant Robert Corner.However, some intrusion of material from thewardroom and the officers' pantry, adjacent to thesecond lieutenant's cabin, or from the warrantofficers' cabins (possibly purser GregoryBentham's or master George Passmore's cabins)opposite the second lieutenant's cabin, could notbe discounted.

Because of their weight, the heavy objects hadsimply dropped down as the upper and quarter-decks disintegrated, and then settled on top of, ornext to, clusters of objects which had been kept inthe spaces or cabins on the lower deck or theplatform deck below. This accounts for the denseartefact cluster found immediately around thecannon retrieved during the 1983 season. Thecannon had settled onto an artefact cluster mainlycontaining surgeon's equipment.

The ships' fastenings and cabin furnishings (inparticular the fireplace and water filter) as-sociated with the upper deck appear to have beendeposited outside the hull remains because of thestarboard heel of the hull. Both the fireplace andstone water filter were most probably fitted in thegreat cabin on the upper deck. With the dis-integration of bulkheads it is likely they rolled

across the upper deck and came to rest against thetop starboard side of the hull. Subsequently, whenthe hull's top sides disintegrated, they bothdropped off the exposed upper deck onto theseabed adjacent to the edge of the hull remains(Fig. 32). The fireplace's position can be con-sidered particularly revealing as it was founddirectly on top of the stock of the anchor whichwas dropped 'underfoot' before the Pandorafoundered. The level of the stock is significant asit probably indicates the level of the seabed at thetime of the wreck. The keel apparently came torest between 8m to 10m from this anchor.

The apparent dearth of smaller, lighter artefactsfrom higher deck levels outside the confines ofthe hull remains can not be easily accounted for.Several scenarios may explain this pattern. Mostlikely, they were dispersed from the wrecked hullas they dropped out of the hull and becameexposed to currents; alternatively they havealready deteriorated completely. The fact thatsome small and light material was found clus-tered around the bulkhead fireplace may be anindication that large objects from the upper decklevel acted as barriers, trapping some of the smalland light material which dropped out of the hull,

36^


-ce5=

FIG. 32. Artist's impression of the wreck disintegration sequence (cf. Fig. 33). (R. Brunke, QM)

thereby preventing them from being dispersed bycurrents. Alternatively, they were swept underthe stern of the hull, where they were subsequent-ly trapped by the final collapse of the starboardquarter. Excavation under the hull remains,especially under the transom, will ascertainwhether this has been the case.A HYPOTHETICAL DISINTEGRATIONSCENARIO. Exposure of a section through thehull remains in the stern area, and subsequentanalysis of the artefact distribution, suggests adisintegration scenario (Fig. 33) that posits fourdistinct stages in the disintegration of the hull (cf.Gesner, 1991: 22). They are as follows:

Stage 1 — Between one and two years aftersinking, masts and rigging have broken off andfloated away; weakened quarterdeck planking

and top sides' timbers break off and are dispersedby currents or trapped against the hull (1).Internally, artefacts accumulate on decks inclusters (2) or against bulkheads and otherpartitions. Gradual internal silting up occurs asorganic materials decay and fine waterborneparticles are trapped and settle within enclosedhull spaces. Artefact clusters begin to be buriedby these trapped and settling fine-grainedsediments (3).

Stage 2— Two to five years after sinking, majorparts of the quarterdeck, the upper deck and topsides have deteriorated or collapsed, primarilydue to marine borer activity. Some of the top sidetimbers fall away onto the seabed or are sweptunder the hull in the stern. (4) Currents begin todeposit coarser sediments into semi-enclosed


37

Stage 2

Stage 3 Stage 4

FIG. 33. Hypothesised disintegration sequence.

hull spaces (5). In more open spaces, currentsbegin to effect light artefacts as they becomeexposed or as they drop out of the hull. These aregradually dispersed. Gradual seabed build-upoccurs under and along the sides of the hull (6).Fine particles continue to be deposited withinenclosed spaces inside the hull, gradually mixingwith or being overlain by coarse sediment asenclosed spaces become more exposed.

Stage 3 — Five to fifteen years after sinking,processes started in stage 2 intensify, with morecollapse of timbers (e.g. the transom) anddispersal of artefact clusters from the upper decklevel. For instance, the bulkhead fireplace fromthe great cabin falls out as the hull's upper sidesdisintegrate and is deposited on top of the boweranchor's stock. A substantial collection ofartefacts, particularly from the upper deck, fallsout of the hull and is trapped under the stern (7)prior to its final collapse. Due to lack of seabedbuild-up supporting the transom and cant frames,these collapse. The weight of the transom causesthe upper sternpost to break off at approximatelythe 'IX' foot draught mark, and fall down tostarboard onto the seabed. Fine particle build-upwithin confined spaces continues, especially atlower deck and platform deck levels, and a mix-ture of fine and coarse sediment accumulateswithin semi-enclosed spaces (8). Build-up con-tinues around the hull. Heavy iron objects, forinstance cannon (9) and the galley stove, dropdown onto lower deck levels and consolidateartefact clusters which have accumulated against

the ship's sides and bulkheads at lower decklevels.

Stage 4 — Fifteen to twenty years after sinking,the wreck approaches stabilisation. All remain-ing spaces within the hull remains have beenfilled with compacted sediments or have becomeoverlain or mixed with coarser grained sediments(10). Seabed build-up has stopped. Marine boreractivity is minimal due to sediment cover whichhas insulated the lower hull timbers and organicmaterials stored within the hull remains fromoxygen-rich water. About one quarter of the hullis left intact. Artefact clusters from lower andplatform deck levels are contained within theiroriginal spaces, especially along the starboardside. Several fittings, for instance the ship'ssternpost and anchor, remain exposed andprotrude from the seabed where they start to becolonised by coralline algae to form concretions(11). Mobile sediment layers are alternativelydeposited or eroded over the wreck depending onwater movement.Implications of the disintegration scenario. Ifthis is the way the hull actually deteriorated,artefact deposition should approximate the hy-pothesised disintegration sequence. Therefore, ifartefacts were deposited in clusters within theentire length and breadth of the hull remains(approximately 35m x 12m), it should be assumedthat 130 of the 250 grid squares (Fig. 34) con-stitute the primary area within the 20m x 50m site.

This area should contain the preserved remainsof the hull, encapsulating the bulk of the ship'sfittings, furnishings, cargo and the crew's

38^


'N5 95 96 115 115 136 136 155 156

4 73 74 93 94 113 114 134 153 154 173 174 193

52 71 72 91 92 111 112 131 132 151 171 172 191 192

70 89 90 109 110 129 130 149 150 169 170 190 209 210

48 67 66 88 107 108 127 128 147 148 167 168 187 188 07

46 65 66 85 88 105 106 1 . 145 146 165 166 185 186 205 • c.

44 63 84 83 84 103 104 123 124 143 144 183 164 183 184 203 204

42 61 82 81 82 101 102 121 122 141 142 161 162 181 182 201 202

10^

20^

30^

40^

50

H primary excavation areas^secondary excavation areas

FIG. 34. Primary and secondary site areas within the 250 grid squares.

20

10

belongings. Because artefacts from the lower andplatform deck levels were buried within the intactlower hull structure, especially along the star-board side, they were probably subject to the leastamount of disturbance during disintegration.Consequently, the distribution of artefacts withinthe primary site area should represent an excel-lent opportunity to record the functional contextof such items as the tools, stores and personalpossessions of the crew.

Further, artefacts known to have been used orstored on the quarterdeck and the upper deckshould be located under the hull remains, inparticular under the collapsed transom in grids50, 52 and 69-72 (Fig. 34). They may also belocated outside the primary area, where they weredeposited by currents after they dropped out ofthe hull and were no longer protected by bulk-heads and partitions. Others may be trapped underthe collapsed stern section.

The hypothesised disintegration sequence im-plies that within the primary site area the artefactsoriginally used and stowed on the lower deck andthe platform deck will be found juxtaposed to, oron top of, strata containing artefacts from theupper deck or the quarterdeck (Fig. 35). How-ever, this will not apply where artefacts fromthese two levels have been dispersed by currents;in such cases dispersed objects will most likely beencountered adjacent to the primary area on theedges of the 20m x 50m site area or even entirelyoutside the site. Nor will this apply in the case oflarge, heavy items, e.g. the cannon and the Brodiegalley stove, both originally placed on the upper

deck which, because of their bulk and weight,were less susceptible to currents and simplydropped down on top of artefacts clustered onlower deck levels.

It can be expected that the 130 grid squaresmaking up the primary site area will contain themost dense concentrations of artefacts; especiallythe grids spanning the preserved edge of thestarboard side of the hull in the stern and bowareas (grids 87-92, 162, 164 and 181-184).

This expectation is supported by the analysis ofthe artefacts retrieved from grids 68-72 duringthe 1986 season, which demonstrated that a sub-stantial number of artefacts can be classified asobjects reflecting the social fabric of the ship —i.e. the crew's personal belongings. In the 1986trench, this class of artefact was found to pre-dominate in grids spanning the inboard starboardedge of the hull remains (grid pair 69 & 70).This is evidence of the clustering of artefactscharacteristic of stage 1 of the hypothetical dis-integration sequence.

The densest artefact concentrations reflectingthe ship's social fabric, especially the personaland professional possessions used on the lowerdeck and/or stowed on the platform deck, aretherefore likely to be found in grid nos 85-94 andgrid nos 105, 107, 109, 111 and 113, which spanthe aft area (the officers' quarters) and grids163-168, 183-188 and 203-206, which span thefocsle area (the 'ordinary sailors' quarters). Objectsstowed in the hold areas and within partitionedareas on the platform decks are to be encounteredin even deeper strata in these grid squares. For

Currentsea floor level

LegendArtefacts representing social fabric(i.e. crew's possessions)

Artefacts representing cargoÂrtefacts representing upperdeck levels(i.e. ship's stores)^

(artefacts dispersed by currents)


FIG. 35. Assumed stratigraphy within the primary and secondary site areas. Sections through the site (looking aft)at grids: 68 to 78 (A); 104 to 116(B); 182 to 192 (C).

example, during the 1986 season, excavationreached a depth of nearly 1.65m in one corner ofgrid 72 with no indication of having reached theceiling planking or platform deck. In some gridsexcavation can therefore be expected to extendan additional 0.5m, possibly to a maximum depthof as much as 2.2m into the seabed. Gridsspanning the midships' section are not expectedto contain any artefact clusters of the crews'belongings, as this part of the hull was primarilyhold space for the ship's equipment and stores.

As far as artefacts from the higher deck levelsof the ship are concerned, it can be expected that

some will be encountered under the hull remains,especially in the stern section (grids 69-74) andsome will be encountered a distance away fromthe preserved hull remains.

MANAGING PANDORA'S BOX: THE 1993EXPEDITION

The Queensland Museum's next major exped-ition was not until January 1993. The objectiveswere changed from excavation to management-oriented ones. This situation had several con-sequences. The most important managementissue was the recognition that discontinuation of

40^


FIG. 36. Diagram showing the excavation-backfillingprocess. A, predisturbance situation: artefact assem-blages (1, 2, 3) etc and hull remains sealed fromoxygenated waters by compact sediment layer 0.5-1.0m thick. B, excavation of artefact assemblage 1:sediment layer gradually removed and deposited onspoilheap (2); artefact assemblage (1) recorded in 3dimensions, retreived over a period of ca. 4 to 5weeks. C, backfilled excavation area (3) using sedi-ment from spoilheap; from observation, backfill is ofless compact nature and of different material com-position due to loss of fmes.

excavation would place at risk the hull remainsand the artefact assemblage still buried withinand around it. Accelerated deterioration of buriedorganic wreck material due to destabilisation ofsediment cover was recognised as the major risk,especially in areas which had already beenexcavated and subsequently backfilled.

Faced with the prospect of continued lack offunds, the decision was made to close down thesite. However, before this could be done, it wasnecessary to assess the extent excavation hadadversely effected the stability and effectivenessof the sediment layer covering the wreck; ofparticular concern was the longevity of the hullremains and the artefact assemblages en-capsulated within it. One urgent issue was todetermine whether micro-biological deterioration,especially of the organic remains and particularlythe previously buried parts of the hull remains,had been accelerated as a result of the exposure tooxygenated waters (Fig. 36).

The objectives of the 1993 expedition weretherefore essentially management oriented, in

particular towards gathering data to base meas-ures to counteract possible adverse impact on theburied hull remains. It was felt that this data couldbe used to inform effective strategies for longterm conservation and management of the site.

The 1993 season's objectives were to:1) gather sediment samples to assess the bio-logical stability of the wreck and its associatedartefacts;2) experiment with physical means to protect thewreck site, especially the backfilled areas, fromenvironmental impacts, particularly currents andwave motion;3) record more remote-sensing images of theburied hull remains; and,4) Undertake limited trenching, in the mid shipsand bow sections, to confirm that the edge of thepreserved hull lay where it was expected.

FIELDWORK. Fieldwork from the expedition'smother-vessel TSMV Pacific Conquest wascompleted in January 1993. During thisexpedition the majority of diving operations werecarried out by divers using surface suppliedbreathing apparatus (SSBA), which enabled asubstantial increase of each diver's bottom time.Using SSBA each diver would be able to diveeffectively, once a day, and spend between 40-45minutes on the seabed without risking extremeexposure to potentially harmful nitrogen loadingand without relying on Scuba technology fordecompression diving. This technique was a con-siderable improvement on diver safety,efficiency and effectiveness.

The expedition team was lead and archae-ologically directed by the Museum's secondcurator of maritime archaeology, the author, andcomprised an 18 member expedition team ofvolunteers and professionals from Queensland aswell as interstate agencies in Western Australia,South Australia, Tasmania and New South Wales.

SEDIMENT SAMPLING. Sediment sampleswere extracted manually from pre-disturbanceand post-disturbance locations (Fig. 37) to gatherdata for a microbiological research project titled'The effect of sediment disturbance on bacterialactivity; its relationship to preservation of buriedorganic 'wreck materials'. This project wasdesigned in collaboration with a consultant, DrDavid Moriarty, and the University of Queens-land's Department of Microbiology, to determinewhether biological differences within sedimentlayers were evident between pre- and post-disturbance site locations. The project wascarried out as research toward a PhD thesis in


41

„V OW' It**

4161:41 >77 4:1712***/0 .4U45 WM4041°00

IONT*1011*

Stern

11°2 44tilt biltl‘0111W+e44%11:‘

„I 01 lag 00 !AO ■I*Oil

5 0 m

FIG. 37. Sediment sampling locations. I, pre-disturbance sample location (off site); II, pre-disturbance locationnear hull remains; III, post-disturbance location 1983 excavation; IV, post-disturbance location 1986 ex-cavation; V, post-disturbance location 1993 excavation.

microbiology at the University of Queensland byJodie Guthrie (Guthrie et al., 1994, 1996).

REMOTE-SENSING SURVEY AND TRENCH-ING Geologists from James Cook Universityvisited the site for three days with a 3.5KHzsub-bottom profiler to conduct a remote sensingsurvey. The profiler was towed across 10 tran-sects at 5m intervals on the surface directly abovethe site area. By conducting the survey acrossthese 10 transects, a close-plot image wasobtained of the sub-surface site area (Carter &Hooper, 1993).

Trenching was carried out to verify the pro-jected outline of the starboard hull remains. Asmall trench was excavated in grids 104, 106 and125 (Fig. 24) to locate the starboard edge of thesheathing at mid ships. After several days ofexcavation, the edge of the sheathing wasexposed in the expected area (Fig. 38).

Subsequently, it was possible to follow the topedge of the sheathing for nearly eight metrestowards the bow section, i.e. into grid 163. Thusexcavation confirmed that the orientation of thehull remains on the starboard side was asprojected.

In the course of fieldwork, a second methodwas used to determine the depth of the hullremains. An hydraulic probe (Green,1990: 137)was placed vertically on the seafloor andactivated along the 22, 24 and 30m transects.Divers Henderson and Sullivan operated theprobe at each selected location until it struck anobstacle. The probe holes were spaced atintervals of approximately lm on transectsbetween 6 and 16m long; the probe was able topenetrate to a depth of between 0.5-0.7m beforestriking obstructions.

With this information and the images from thesub-bottom profiler, it was then possible todetermine more accurately the extent of the hullremains (Fig. 39). This was done by tilting thehull at the 34 degree angle of heel measured in1986 from the angle of the deadwood bolts in grid74.

Excavation conducted in the mid ships' trench(grids 102, 104 & 106) also revealed that this areadid not contain similar artefact clusters to thoselocated in the grids spanning the starboard hullremains in the stern section. No artefacts werelocated on top of the remaining midship's timbers;several fastenings and a copper cauldron

42^


FIG 38. Starboard edge of the sheathing exposed atmid ships (grid 125).

(MA3010) were located in grid 104 well outsidethe confines of the preserved hull (Fig. 37).

This information further confirmed the hypoth-esised disintegration sequence, indicating that noartefact clusters similar in density or diversity tothose encountered in 1986 were located on top ofor within the confines of the hull remains at midships. This was not surprising since the vessel'smid ships' section (i.e. the well between quarter-deck and focsle deck) was not used as a livingarea. Clusters of personal or professionalpossessions would therefore not be expected here.

PHYSICAL PROTECTION. Although furtherconsolidation of the 1986 trench had been con-sidered a special priority at the beginning of theexpedition, time had been lost during this shortseason due to three days of adverse weatherconditions. Consequently, little work could beundertaken to consolidate and stabilise trenchesback-filled and covered during previous ex-cavations.

The temporary nature and untextured surfaceof the plastic protective sheet used in 1984 and1986 was not satisfactory. It hindered the

FIG. 39. Sections (looking aft) through the stern (A),midships (B) and bow areas (C), indicating the extentof expected hull remains. Solid lines = remains,dotted lines = deteriorated hull.


43

compacting of the layer of backfill by preventingbonding between it and the underlying, un-disturbed sediment. This had been conclusivelyobserved during a monitoring dive in April 1988when the author accidentally dropped a smallhammer while swimming across the site severalmetres off the seabed. The hammer almostdisappeared into the sediment after it struck theseabed. Having seen where the hammer came torest, it was retrieved and in doing so thequicksand-like state of the seabed was observed.Upon further inspection, it was found that thequicksand-like sand layer only occurred wherebackfill had been deposited on top of plasticlining. It appeared that the backfill was looselysuspended on top of the plastic lining and therebyunable to bond with the underlying sediment. Inview of this, the use of plastic sheet wasdiscontinued and the mid ships' trench (grids104, 106 and 125) was stabilised by back fillingand consolidated with approximately 8m2 of anylon mesh fabric (shade cloth) instead. Theshade cloth was spread out and weighted downwith sand-filled polyurethane bags.

During an inspection dive in August 1994conducted by Warren Delaney, the MaritimeArchaeology Section's technical officer, the bagsweighting down the shade cloth were observed tobe almost completely buried, indicating that theyacted as traps for current borne sediment. Arecently deposited layer of approximately 30cmhad covered the shade cloth and sandbags,indicating the effectiveness of this method.

RE-INTERMENT OF HUMAN SKELETALREMAINS. Another expedition objective wasthe consecration and placement on-site of amemorial obelisk containing the skeletal remainsof one of the Pandora's crew. These remains hadbeen retrieved from grid 70 during the 1986excavation. After a 'Service for burial at sea'(RAN, 1979) conducted by HMAS Cairns' chap-lain Mark Wallbank, the obelisk was loweredfrom the deck of the expedition's mother-vesseland placed on the seabed adjacent to the wrecksite (Fig. 40). Set into the obelisk was a perspexplaque inscribed:

The 24 gun frigate PANDORA was sent into the SouthPacific in pursuit of the BOUNTY. After a search of 4months the frigate was returning to England with 14mutineers imprisoned in a temporary cell constructedon her quarterdeck. The PANDORA struck a sub-merged reef to the east of this site. This monumentcontains the skeletal remains of one of PANDORA'screw who perished with 30 of his shipmates and 4mutineers when the PANDORA foundered and cameto rest here on 29 August 1791. This wreck is a

FIG. 40. Consecrated obelisk containing humanremains being lowered on-site from the mothervessel.

protected historic site (Historic Shipwrecks Act, 1976)and has been partially excavated by the QueenslandMuseum.

Prior to re-interment, the skeletal remains hadbeen examined. Anatomical studies indicated thatthey were of an adult male of between 22-25years of age, who had been approximately 1.65mtall and had possibly suffered from rickets as anadolescent (Wood & Hodgson, 1996).

The identity of this individual has not yet beendetermined. It is suggested here that he may havebeen one ofthe two sailors whom surgeon GeorgeHamilton mentions were killed while the Pan-dora's crew were trying to save the ship after shehad been refloated from Pandora Reef (Thom-son,1915 : 143). It is possible that, in the interestsof keeping up the crew's morale, the two deadsailors were taken below, to be buried at sea afterthe ship had been saved. To date, the names ofthese two individuals has not determined.

Alternatively, it has been suggested that theskeletal remains may have been of one of the fourmutineers who drowned when the ship sank


(Wood & Hodgson, 1996). However, this isunlikely in light of information provided byJames Morrison, one of the surviving mutineers,who mentions that immediately after thePandora foundered, the roof of the prison cell(Pandora's Box) had floated up and had beenused as a raft by several of the survivors. (Rutter,1935: 128) Thus, any of the prisoners who weredrowned inside the cell while the ship wassinking would either have floated out of thewreck when their corpse became bloated upondecomposition, or have been devoured by fish.According to Morrison the only prisoner whoremained trapped inside the cell was the Bounty'scooper, Henry Hildbrandt (Rutter, 1935: 127).The remains are therefore more likely to be one ofthe thirty one crew whom Captain Edwards listsas lost with the vessel (Appendix 1).

REFINING SYSTEMS: THE 1995 SEASON

The Museum's next season was of the sameduration as that of 1993 and comprised, with fivechanges, the same personnel. The team spent 16days at the wreck.

The expedition was made possible through theassistance of the Queensland Museum's Board ofTrustees, which made available a special advanceto the Pandora Foundation to carry out furtherexcavation work.

The Pandora Foundation had been establishedin August 1994 as a result of a ministerial init-iative. The Foundation's initial purpose was tocarry out fundraising for the Museum. In par-ticular, its intention was to capitalise on a subsidyoffer made by the Queensland Government, whichin March 1994 agreed to allocate $1 million, as a$1 subsidy of every $2 raised by the Foundation.

Revitalisation of the Pandora project wasinitiated by the Hon. D. Wells (then Queensland'sAttorney General and Minister for the Arts) afterthe Museum's senior management conducted areview of three options for the continuation ofarchaeological work at the Pandora wreck site.Two of these options included the retrieval, con-servation and display of the hull, along similarlines to the Mary Rose Project; the third optionaddressed recovery and conservation of thewreck's moveable contents only.

The review took into account the wreck'scultural significance as well as the earningpotential of its public display. It concluded thatany informed decisions regarding the raising ofthe hull could only be made with reliable in-formation about the exact extent and structural

coherence of the hull remains. In light of the factthat such information would not be availableuntil the contents of the hull had been retrieved, adecision was made to concentrate first on therecovery of the hull's contents. Consideration ofthe feasibilty of recovering the hull remainsalong similar lines as the Mary Rose Project inPortsmouth (UK) was therefore postponed tofurther review, to be conducted at an unspecifiedtime, if appropriate.

The review's summary concluded that it was:... premature to discuss the potential of the 'Pandora'in terms of a Mary Rose-like project because there isno definitive indication of how much hull is preserved... The two conclusions at this stage are: 1) raising thehull is considered unjustifiable, firstly in terms of itspotential to advance knowledge and, secondly, ineconomic terms with respect to developing a touristattraction, and 2) retrieval/conservation/display of theartefact assemblage is justifiable in terms of thecultural significance of the material and threats fromvandalism or further inevitable weathering.

With the formulation of a less ambitious and,arguably, more realistic project outcome, theMuseum's Curator of Maritime Archaeology (theauthor) was tasked to make an assessment ofresources required for archaeological retrievaland conservation of the wreck's moveable contentsonly.

This assessment found that with favourableoperating conditions and requisite resources, fivemajor expeditions could complete retrieval of arepresentative sample of the wreck's contents.Ideally these expeditions — each of approx-imately six weeks duration — should becompleted by a multi-skilled expedition teamcomprising a minimum of 25 site-experienceddiving archaeologists and/or assistants. The costof the five expeditions (in 1994 figures) wasestimated to be approximately $320,000 per ex-pedition with additional outlays of approximately$180,000 per year to provide for additional temp-orary specialist staff. At least three additional fulltime staff and several consultants, as well as atimely and expeditious program of artefactconservation and curation were deemed to berequired. The estimates assumed continuedavailability of voluntary community and cost-free peer group support from interstate museumsand heritage management agencies (cf. CreativeNation, 1994: 77), on-going support from alreadyavailable Queensland Museum human resources(at least six fulltime staff) as well as resources inthe Queensland Museum's Maritime Archaeology,Conservation, Photography, Marketing andPublic Programs Sections.


FIG. 41. Illustration of assumed deposition of the pump chamber. (S. Elmer)

The 1995 expedition was conducted as a dressrehearsal for a projected series of five majorexpeditions, as well as to maintain the revitalisedproject's momentum and continuity and toprovide publicity and promotional material forthe Pandora Foundation's planned 1996 fund-raising campaign.

Several objectives were formulated to suit theshort duration of the 1995 expedition, includingthe testing of a newly established 4-point moor-ing system. Installed by a marine engineeringcontractor in December 1994, the mooringsystem was designed to hold the expedition'smother-vessel in position directly over the wreckwhile surface-supplied diving operations werecarried out.

Excavation work was conducted in two areasof the site between 20th January and 5th February1995 by a team of 16 divers, again operating fromthe TSMV Pacific Conquest and using SSBAequipment and the same type of excavation plantas on previous excavations.

FIELDWORK. Mid Ships Area. A previouslyexcavated trench spanning grids 106 & 125 waspartially re-opened because an object partiallyuncovered during the 1993 season was thought tobe a carronade. Initially it appeared the carronadecould easily be retrieved. However, furtherexcavation, subsequently identified the object asa length of a pump chamber; most probablyassociated with the Pandora's chain pump(McKay & Coleman, 1992: 78).

The pump chamber's location prompted spec-ulation about its deposition. It was found lyingathwart the edge of starboard sheathing, partiallywithin the confines of the preserved hull. Mostimportantly, as this length was round, it wasdetermined to be the bottom end of the chamber(cf. McKay & Coleman,1992: 78, fig. Eli/1).The top, square, end of the chamber could not beseen as its length was assumed to extend aconsiderable distance into the seabed, outside theconfines of the preserved hull. This indicated that

the entire pump may have toppled over after thehull's topsides had disintegrated. Consequently,the pump's cisterns probably fell off the upperdeck in a similar manner as the cabin fireplaceand were deposited adjacent to the starboard hull(Fig. 41).

In addition to the pump chamber, an array ofconcreted iron or metal fittings associated withthe pump assemblage was uncovered. Theseconcreted fittings were attached to the chamberand presented a veritable cat's cradle of objectsobscuring the top end of the pump chamber.Further excavation work to disentangle this arrayarid thereby delineate the chamber and locate thecistern was deemed beyond the scope of the 1995expedition's objectives. Consequently, grid 106was back-filled and stabilised with mesh clothand sand filled bags.Stern Area. The selected excavation area in thestern was in grids 68 & 70 where two largeconcretions (each measuring approximately1.3mx 1.15m x0.5m) were located; each concret-ion was expected to lie on top of a complex offragile artefacts.

The largest concretion, thought to be a carro-nade, lay within grid 70 and was excavated first.The primary objective here was to retrieve thecomplex array of fragile objects expected to lieunderneath it. This was imperative before the guncould be raised.

During excavation, it became apparent that adiverse concentration of objects were clusteredunder the concretion. Retrieval of the objectsrequired more time than anticipated as severaldelicate objects, for example an intact coconutand a ceramic apothecary's jar containing mercury,were found firmly attached to the concretion; itwas imperative to dislodge these prior to theraising of the concretion. Gradually the base ofthe concretion was exposed and smaller objectswere dislodged from it and retrieved. As the baseof the concretion was not attached to underlyingconcretions, it was prepared for raising.

FIG 42. Concretion being secured to the lifting litter.


A prefabricated iron and timberlitter was assembled on the seabednext to the concretion. Havingattached air bags and lifting slingsunder it, divers then filled the airbags and lifted the concretion offthe seabed, gently guiding it ontoa litter. Once on the litter, the airbags were deflated and theconcretion was wrapped inwebbing and secured (Fig. 42).The concretion and litter wasestimated to weigh between0.6-0.9 tonne. It was then raisedby crane onto the aft deck of thePacific Conquest.

Excavation was subsequentlycontinued in the area where theconcretion had been situated. Itcontained dense and diverse clus-ters of artefacts, each requiringadditional time and concentrated effort prior toretrieval.

Because the recovery of the concretion(MA4190, corronade) had required more effortthan anticipated at the beginning of theexpedition, time constraints prevented work onthe second large concretion. However, duringexcavation around the first, a swivel gun(MA4028) was noticed in the adjacent grid. Uponinspection it appeared easy to retrieve. One diveteam prepared the swivel gun for raising whichwas achieved on the last day of excavation usinganother specially pre-fabricated litter, liftingslings, air bag and crane.

SEDIMENT SAMPLING AND OTHER WORK.Sediment samples were collected from variouslocations within the wreck for microbiologicalanalysis. In addition to locations where sampleshad been collected during previous expeditions, aseries of samples was collected from a 12mtransect laid out across the bow section betweenthe 40.0 and 50.10 grid reference poles. Thistransect covered unexcavated areas as well asexcavated areas (Guthrie et al., 1996).

Several minor tasks were also completed; inaddition to inspection and monitoring of thestability of grid reference poles, one of the tasksinvolved consolidation of a number of cardinalgrid reference poles by placing 30kg concretefootings at their base. This was achieved on the0.0, 50.0, 0.10, 50.10, 16.6 and 16.12 grid refer-ence points. Another task involved the clearing ofswim lines previously placed on-site to facilitate

location of the mooring points. This was carriedout as a safety precaution to prevent fouling oftheSSB equipment umbilicals.

DISCUSSION. With the raising of the concretedcarronade (MA4190) from grid 70 and the swivelgun (MA4028) from grid 72, one of the expedition'smain objectives was achieved. It demonstratedagain that a large concreted object could beextracted from the wreck without causing unduedisturbance to underlying or adjacent objects.

Achieving this objective was largely due to theability of the Pacific Conquest to remain inposition directly over the wreck during divingoperations. This proved a crucial factor foreffective diving and artefact recovery operationsinvolving SSB equipment, as well as for directlifts of large, heavy objects from the seabed ontothe deck of the mother vessel.

Another important factor contributing to theexpedition's success was the composition andexperience of the expedition team; with morethan half of the team experienced in work-divingat the Pandora wreck site.

The deposition of the artefact cluster foundunder the concretion was subsequently analysedand found to be continuous with the clusterretrieved from the adjacent area during the 1986expedition. Predominantly personal possessions,this cluster is thought to be part of the sameassemblage which can be provenanced to thewardroom, the officers' pantry and the secondlieutenant's cabin.


ONGOING WORK

During the 1996 and 1997 seasons, systematicexcavation continued on the stern. In the 1998and 1999 seasons systematic excavation con-tinued on the stern and commenced on the bow.Artefacts from these excavations are in the pro-cess of conservation, analysis and description,and will be the subject of forthcoming papers inthe Cultural Heritage Series ofthe Memoirs oftheQueensland Museum. The final excavation willtake place in 2001. The reason for the emphasison the bow and stern sections of the ship is thatthese are the areas most likely to contain artefactsrelating to the social fabric of the crew. Byexcavating the focsle areas where the Pandora'sordinary sailors lived, ate and slept, it was likelythat artefact clusters corresponding with thesailor's sea chests would be found. These werethe only spaces within the Pandora where thesailors could keep personal belongings. Most ofthe Pandora's officers were privileged to have acabin as their personal space. However, theordinary sailors had to make do with sea chests tostore their personal belongings. Moreover,excavation in the bow penetrated preserved plat-form deck spaces under the lower deck, where anumber of the ships specialists (for example thecarpenter, gunner, sail-maker and bosun) hadtheir work and storage areas (Gesner, 1997).

By focussing primarily on the stern and bow ofthe wreck, both which contain material evidencewhich reflects the personal lives and professionalactivities of the crew, an analysis of the personaltastes and habits of the crew can be undertaken.

Analysis of their day-to-day lives will add towhat is known about the crew from historicalsources such as the Pandora's muster list or theeyewitness voyage accounts. The artefacts willbe of paramount importance in illustrating objectuse and acquisition during a period when Britishsociety was undergoing tremendous social andeconomic change due to the commercialrevolution which was flourishing after ca. 1760,and which was a precondition for the advent ofthe Industrial Revolution (McKendrick et at.,1982). Analysis of personal possessions willprovide insights into the values and attitudesprevalent among a fairly typical late 18th centurynaval crew, and thereby provide data to assesswhether sailors, a sub-cultural group described as'fringe dwelling' (Rodger, 1986:15), weresusceptible to the same economic, social andcultural changes influencing broader sections of

British society during the last decade of the 18thcentury.

There are no plans to systematically excavatethe ships mid section in the near future. Theexpectation is that this part of the wreck willcontain items reflecting aspects of nauticaltechnology carried in the ships hold such as waterbarrels, cordage, spare anchors, shot and ballast.Although information on this type of material isof interest to nautical archaeologists, it was notgiven the same priority as the material reflectingthe Pandora's social fabric. There exist other,more comprehensive, archival and printedsources of information about 18th centurynautical technology, so there is less urgency toretrieve the material expected to be there.

ACKNOWLEDGEMENTSThe archaeological work on the Pandora ship-

wreck has been carried out by the QueenslandMuseum with generous financial assistance andadministrative support from the PandoraFoundation, the Queensland Museum Board ofTrustees, the Queensland Government and theCommonwealth Government. Archaeologistsand other specialists and professionals from allover Australia, and overseas, have participatedin, and supported the five Queensland Museum'sexpeditions between 1983 and 1995. The authorgratefully acknowledges the assistance providedby the following organisations, colleagues andvolunteers.Volunteers. Beryl Turner, David Bosisto, AlisonDarroch, Capkin van Alphen, Sayward Ayre,Alan Boon, Jerry Calvert, Hubert Hofer, SueKelly, Joe Sedlacek, Wendy Black, AlistairBrown, David Luddington, Terry ew, MaureenBoyce, John Murray, Robert McKinnon, MartiWalker, Kathy Kramer, Len Zell, Peter Sullivan,Kaye Walker, Heath Bell, Elizabeth Evans-Illidge, Jodie Guthrie, Paula Tomkins, MichaelSt.James, Howard Smith, Nigel Erskine, CosCoroneos, Ingrid Ward, Peter Veth, AnnabelWood, Mark Lawrence, Ian Lawrence, VivienneMoran, John Read, Dennis Lee Sye,David Wood,Coleman Doyle, Stirling Smith, Brad Duncan,Jaco Boshoff, John Gribble, Kevin Hubbard, BrianDermody, Graham Schulz, Greg Chappelow,Rhiannon Walker, Jane White. Brian Richards,Simon Adey-Davies, Gavin Ericsson, SeanRubidge.Maritime Archaeological Association ofQueensland. Paul Brown, Tom and MarthaGood, Bob Gerrard, Colin Ward, Don Norman,Maree Edmiston, Fiona Scott, Kevin McDougall,


Mark McCafferty, Alan Mole, Anne-Marie Grice,Richard Priddy, David & Enid Waddington, JohnGritton, David Bell.Commonwealth Government. Dept. HomeAffairs & Environment (Mrs Jennifer Amess);Dept. Arts, Sports, Tourism & Territories (LesNielsen, Ray Mulligan, Alan Roberts, DarrylMcIntyre, Mrs Barbara Williamson); Dept.Communications & the Arts (Australian CulturalDevelopment Office, Kay Daniels, Dawn Casey,Wayne Cassidy, Martin Hiscutt, Patricia Snigg);Environment Australia (Peter Mitchell, JuliaSearle, Barry Reville); Australian NationalMaritime Museum (Kevin Fewster, Mary-LouiseWilliams, Antonia Syme, Kieran Hosty, SueBassett); Great Barrier Reef Marine Park Author-ity; Royal Australian Navy (HMAS Cairns,HMAS Moreton, RAN Hyographic Service).Western Australian Government. WA MaritimeMuseum (Graeme Henderson, Patrick Baker, ScottSledge, Geoffery Kimpton, Mike McCarthy, JonCarpenter, Dena Garratt, Jeremy Green, CorioliSouter).Tasmanian Government. Tasmania NationalParks & Wildlife Service (Paul Clark); Dept. ofPrimary Industries, Water & Environment (MikeNash).Victorian Government. Victoria ArchaeologicalSurvey (Shirley Strachan, Peter Hervey);Heritage Victoria (Ross Anderson).South Australian Government. Dept. of Environ-ment & Heritage (Martin Bryne, Bill Jeffery,Terry Arnott).New South Wales Government. NSW HeritageOffice (Rosalind Strong, David Nutley).Queensland Government. Dept. of Health(Townsville General Hospital, Tom Fallowfield,David Griffith, Stuart Lavender); QueenslandNational Parks & Wildlife Service; Dept. ofEnvironment & Heritage; Queensland Rail.Queensland Museum. Warren Delaney, JanetCampbell, Gary Cranitch, Christine Ianna, FreyaBruce, Jennifer Blakeley, Amanda Pagliarino,Jessica Turner, Ron Coleman, Sally May, BrookBatley, Marisa Giorgi, David Hallam, NevilleAgnew, Peter Illidge, Vivienne Moran, AlisonMann, Andrew Viduka.Others. The Dive Bell (Colin Hodson); ElizabethE Coral Cruises (Bob & Mandy Evetts, JohnEvetts); David Flatman Productions; James CookUniversity, Dept.of Geology (Prof R. Carter);James Cook University, Dept. of Anthropology& Archaeology; National Tidal Facility.

LITERATURE CITEDADMIRALTY Adm 2/120. (PRO: Kew).ADMIRALTY Adm 168/147. Specification for a 24 gun

ship. (PRO: Kew).ADMIRALTY MS 180. Edwards' papers, Naval Hydro-

graphic Office. (Taunton: Somerset).ADMIRALTY A/2831. Captain Edwards orders.

(PRO: Kew).ADMIRALTY 106/1317. In letters from naval officers.

(PRO: Kew).CARTER, R.M. & HOOPER, K. 1993. Shallow seismic

survey of the Pandora wreck site. Pp. 37-51. InGesner, P 1993 (ed.) Managing Pandora's Box.Unpubl. report, Queensland Museum.

CREATIVE NATION, 1994. Commonwealth CulturalPolicy Report. (Dept of Communications and theArts: Canberra).

de LACY, G 1997. Plagiarism on the Bounty, a note onthe composition of Morrison's Journal. Mariner'sMirror 83(1): 84-89.

DENING G. 1992. Mr Bligh's bad language: passion,power and theatre on the Bounty. (CambridgeUniversity Press:Cambridge).

DILLON, P. 1829. Narrative of La Perouse's Exped-ition, Bibliotheca Australiana 53. (N. Israel:Amsterdam, 1972).

DU RIETZ, R.E. 1986. Peter Heywood's Tahitianvocabulary and the narrative by James Morrison;some notes on their origin and history. (Banksia 3,Dahlia Books: Uppsala).

GESNER, P. 1988. The Pandora project: reviewinggenesis and rationale. The Bulletin of the AustralianInstitute for Maritime Archaeology. 12(1): 27-36.

1991. Pandora: an archaeological perspective.(Queensland Museum: Brisbane).

1993. Managing Pandora's Box, report on the 1993Pandora expedition. Unpubl. report to theQueensland Museum Board.

1997. HMS Pandora. Pp. 305-07. In Delgado, J. (ed.)British Museum encyclopaedia of underwater &maritime archaeology. (British Museum Press:London).

1998. Managing Pandora's Box: an exercise in eco-archaeometry. Pp. 230-236. In Bound, M. (ed.)Excavating ships of war (International MaritimeArchaeology Series, 2). (Anthony Nelson:Oswestry, Shropshire).

GREAT BARRIER REEF MARINE PARK AUTHOR-ITY, 1985. Far Northern Section Zoning Plan,BRA Q102, August 1985. (Commonwealth ofAustralia, ISBN 0-642-52433-5).

GREEN, J. 1990. Maritime archaeology: a technicalhandbook. (Academic Press: London).

GUTHRIE, J.N., BLACKALL, L., MORIARTY, D. &GESNER, P. 1994. Wrecks and marine micro-biology: a case study from the Pandora. Bulletin ofthe Australian Institute for Maritime Archaeology18(2): 19-24.

GUTHRIE, J.N., BLACKALL, L., MORIARTY, D. &NICHOLS, P.D. 1996. Decomposers of shipwreckPandora. Microbiology Australia 17: 17-19.


HAMILTON, G 1998. A voyage round the world in HisMajesty's frigate Pandora (facsimile edition of theoriginal 1793 edition). (Hordem House: Sydney).

HENDERSON, G 1979. Report to the Department ofHome Affairs and Environment on the 1979 ex-pedition to establish the identity and archaeologicalpotential of the Pandora wreck. (Unpubl. report:Dept of Maritime Archaeology, Western AustralianMaritime Museum).

1986. Maritime archaeology in Australia.(University of Western Australia Press: Perth).

HOWSE, D. (ed.) 1990. Background to discovery: Pacificexploration from Dampier to Cook. (University ofCalifornia Press: Berkeley).

JAMES, W. 1886. Naval history of Great Britain.(London).

KEN1HAN, G (ed.) 1990. George Hamilton: a voyageround the world in H M Frigate Pandora. (WilliamTorrens: Adelaide).

LAVERY, B. 1987. The arming and fitting of English shipsof war 1600-1815. (Conway Maritime Press:London).

MACLEOD, R. & REHBOCK, P.F. 1988. Nature in itsgreatest extent; western science in the Pacific.(University of Hawaii Press: Honolulu).

MARTIN, J. 1991. Escape from Botany Bay: `memor-andorns' by James Martin (Mulini Press: Canberra).

MAUDE, H.E. 1964. The voyage of the Pandora's tender.Mariner's Mirror 50(3): 217-235.

MAY, W.E. & ANNIS, P.GW. 1970. Swords for seaservice. (Her Majesty's Stationary Office: London).

McCAFFERTY, M. 1988. Direct linear transformation, acase study based on HMS Pandora. The Bulletin ofthe Australian Institure for Maritime Archaeology12(1): 51-55.

McCARTHY, M. 1983. Ships fastenings: a preliminarystudy. Bulletin of the Australian Institute ofMaritime Archaeology 7(1): 28-34.

McKAY, J. & COLEMAN, R. 1992. The 24-gun frigatePandora. Conway anatomy of the ship series.(Conway Maritime Press: London).

MORTIMER, G 1791. Observation and remarks madeduring a voyage (...) in the brig Mercury com-manded by John Henry Cox. (T. Cadell: London).

OLIVER, D. 1988. Return to Tahiti, Bligh's secondbreadfruit voyage. (Melbourne University Press:Melbourne).

PARKER, M.A. 1795. A voyage round the world in theGorgon, Man of War, (Facsimile edition pub-lished by Hordern House Rare Books, AustralianMaritime Series, 1991: Sydney).

PIGGOTT, L. 1995. The surgeon's equipment from HMSPandora. Bulletin of the Australian Institute forMaritime Archaeology 19(1): 23-28.

RODGER, N.A.M., 1986 The wooden world: an anatomyof the Georgian Navy. (Collins: London).

RIMER, 0. (ed.) 1935. The journal of James Monison,bosun's mate of the Bounry. (Golden CockerellPress: London).

SMITH, B. 1992. Imagining the Pacific: in the wake of theCook voyages (Melbourne University Press:Melbourne).

SYRETT, D. & NIARDO, R.L. (eds) 1994. The com-missioned sea officers of the Royal Navy,1660-1815. Naval Records Society. (Scolar Press:London).

TAGART, E. 1832. Memoir of the late Captain PeterHeywood R.N. (Effingham Wilson: London).

THOMSON, B. (ed.) 1915. Voyage of HMS Pandora;being the narratives of Captain Edward Edwards andsurgeon George Hamilton. (Francis Edwards:London).

WARD, I.A.K, LARSCOMBE, P. & VETH, P. 1998.Towards new process-oriented models fordescribing wreck disintegration: an example usingthe Pandora wreck. Bulletin of the AustralianInstitute for Maritime Archaeology 22(1): 109-115.

WOODS, W. & HODGSON, G 1996. The skeletonfrom the wreck of HMS Pandora. Pp. 303-307. InCovacevich, J., Pearn, J., Case, D., Chapple, I. &Philips, G. (eds) History, Heritage & Health;proceedings of the Fourth Biennial Conference ofthe Australian Society of the History of Medicine.(The Australian Society of the History ofMedicine: Brisbane).

UNITED SERVICES MAGAZINE, 1843. ThePandora again! United Services Magazine andnaval and military journal 172: 411- 420.


APPENDIX 1

PANDORA'S CREW AT THE TIME OFDEPARTURE. Key: + = did not survive thewrecking; # = died on the journey to Englandafter the wrecking, either in the Dutch East Indiesor on the way back from Batavia; * = was withWilliam Oliver on the tender Matavai and not onthe Pandora at the time of the wreck; ** = diedbefore the wreck; ++ = lost in the cutter before thewrecking.

Captain: EDWARDS, EdwardCaptain's clerk: EDMONDS, Edward King1st Lieutenant: LARKAN, JohnLieutenant's yeoman: GERMAIN, John2nd Lieutenant: CORNER, Robert3rd Lieutenant: HAYWARD, ThomasMaster: PASSMORE, GeorgeMaster's mates: REYNOLDS, George

RICKARDS, ThomasBOWLING; EdwardOLIVER, William # *

Surgeon: HAMILTON, GeorgeSurgeon's mate: INNES, JamesPurser: BENTHAM, GregoryPurser's Steward: BOWLER, Robert +Midshipmen: ATKINS, James

FENWICK, John #McKENDRICK (or Kendrick), GeorgeMATSON, RichardPYECROFT (or Pycroft), HenryRENOUARD, David *SIVAL, John ++

Quartermasters: ANDREWS, John +DODDS, James *HAYTHORN, William *LINDSEY, Thomas

Quartermaster's mates: MANNERS, JamesMARSHALL, JohnSPENCE, CharlesWATSON, Thomas

Caulker: BELL, ChristopherBoatswain (Bosun): CUNNINGHAM, JohnBosun's mates: MOULTER, William

JOHNSON, James **CUNNINGHAM, Joseph ++

Coxwain: WISDOM, WilliamCarpenter: MONTGOMERY, Alexander #Carpenter's mate: FARRENS, JohnCarpenter's crew: BROWN, Robert +

YOWELL, JohnGunner: PACKER, JosephGunner' s yeoman: FOSTER, RobertGunner's mate: STEWARD, WilliamQuarter gunners: CAMERON, John

NICHOLAS, HenryPUDNEY, WilliamREEVE (or Reeves), Robert

Sailmaker: JACKSON, Thomas

Sailmaker's mate: ARBUTHNOT, Alexander +Cooper: REYNOLDS, HenryMaster at Arms: GRIMWOOD, John +Corporal: RODRICK, William +Cook: MAHONEY, Denis #Annourer CLEMENTS, William #Armourer's mates: STAPLETON, Philip

HODGES, JosephAble Seamen:ADAMS, HENRY **

BARKER, William *#BEACKIE, ThomasBRIXLEY, Thomas +BROWN, AllanBROWN, James *BROWN, John (note 2 below)CALLIGHAN, TimothyCAMPBELL, JohnCARROLL (or Carrol), Thomas +CARTER, William +COBB, FrancisCOLLINS, HenryCULLEMORE, James +CURLING; John GodfreyDEDWORTH, William +DENEEN, John *DOUGHTY, ThomasDUNNETT, JohnDURLINQ Daniel +EGLINGTON, George +FARRELL, William #FARRENS, Robert *FLETCHER, William +FUSS, JamesGEORGE, JosephGOOD, James ++HAMMOND, Samuel +HANZELL, JacobHENRY, Patrick +HOUSTON, HughINGLESBY, (or Ingoldsby) WilliamKEMP, WilliamKING, PeterLAYSELL, JamesLOFT, Henry *LION (or Lions), William +MANSON, JohnMURPHY, John #MILTON, Robert #MITCHELL, MosesMURRAY, JamesORCHARD, RobertPATTERSON, John #PAXTON, Alexander +PERRYMAN, William +PILCH, William #POWELL, HenryPUMMELL, CharlesRANCE, SilasREEVES (or Reeve), William +


ROBINSON, AlexanderRUFFEY, JohnSCOTT, James ++SKELTON (or Shelton), William +SMITH, JohnSMITH, ThomasSPRACKLEY, ThomasSWAN, William +SWEENEY (or Swiney), RogerTAYLOR, RobertTHOMPSON, LewisVARLINQ ThomasWASDALE, William ++WELLS, DanielWELLS, WilliamWILMOTT, JohnWORLDHAVE (or Wooldhave), Richard *Ordinary Seamen: CRAY, WilliamFEA, Robert +GORDON, James +MACKIE, Richard +MAYSONER, Martin +MILLER, James +THOMPSON, William +WEBBER, Robert +

Landsmen: BANDY, Joseph +CONNELL, James #DAVIS, John #JONES, Evan +

Notes: 1, Two crew on board at the time of departurefrom Portsmouth were not on board on 29th August1791, having died during the voyage: James Johnson(bosun's mate) died off Rio de Janiero and HenryAdams (Ordinary seaman) died off Vanikoro. Bothwere buried at sea.2, John (or Jonathan) Brown, listed above, was onboard at the time of the wreck but was not part of theoriginal crew. Brown was the crew member from theMercury, who had been put ashore at Tahiti where hewas living as a beachcomber Brown assisted CaptainEdwards with information leading to the rounding upof the five mutineers who initially eluded capture bythe Pandora's men. Brown was mustered as an ableseaman at Tahiti. He apparently elected to stay behindin the East Indies when the Pandora wreck surviviorsgot to Batavia. His fate is not recorded.

Crew of the schooner Matavai (tender) when lostoff Samoa in June 1791; re-united with Pandora' screw in Samarang, October 1791:Master's Mate: OLIVER, WilliamMidshipman: RENOUARD, DavidQuartermaster: DODDS, JamesAble Seaman: BARKER, William (died in Samarang

after the open boat voyage in the Matavai)FARRENS, RobertBROWN, JamesDENEEN, John

LOFT, HenryWORLDHAVE (or Wooldhave), Richard

Crew of the cutter when lost off PalmerstonIsland on 24th May 1791; never seen or heard ofagain.

SIVAL, John (midshipman)CUNNINGHAM, JosephGOOD, JamesSCOTT, JamesWA SDALE, William

APPENDIX 2SOUTH PACIFIC ISLANDS VISITED ORSIGHTED BY THE PANDORA.Ducie Island —Named by Captain Edwards. Sighted on

16 March 1791; no crew ashore.Tahiti —Arrived on 23 March, departure on 8 May 1791;

extensive interactions with Tahitians.Moorea — Sighted on 8 May 1791. No crew ashore.Huahine —Arrived on 9 May, departed on 10 May 1791;

boats sent ashore Ulietea).Otaka — A boat's crew briefly ashore during 10 and 11

May 1791.Bora Bora — Tatahu, the 'King' of Bora Bora, visited on

board the Pandora.Maurua — Sighted on 12 May 1791. No crew ashore.Aitutaki — A boats' crews briefly ashore on 19 May

1791; extensive description by Hamilton of thecollecting of an ornately carved spear. (Thomson,1915: 123)

Palmerston Island — A boat's crew ashore (Corner);Bounty's driver yard found. The cutter searchedthe outer islands (Sival); Hayward and Cornerwent ashore on other islands. Sival returned withstrange 'curiosities' (a canoe and Figurines)(Thomson, 1915: 124). Sival's jolly boat lost offPalmerston I. 22nd/23rd May; never heard ofagain.

Duke of York Island — Visited on 6 June; boats ashore;deserted huts discovered.

Duke of Clarence Island (Tokelau) — Visited on 12 June;boats ashore; islanders seen, but no contact withcrew.

Chatham's Island (Savaii/Samoa) — Visited on 18 June;contact with islanders.

Tutuila (Samoa) — Visited 21 June; contact withislanders; many 'curiosities' acquired, includingbirds and parrots. Lost contact with the Matavai,Pandora's tender commanded by William Oliver.

Falafagee Island — Sighted on 26 June.Anamooka (Tonga) — Visited on 29th June; Hayward in

canoe to Ha' apai; extensive contact with islanderson Anamooka; watering party attacked (Lt Cornerkilled one assailant).


Tofua - While the Pandora was at Anamooka, LtHayward made a separate voyage to Tofua, one ofthe Friendly Islands, in a hired sailing canoe.

Vavua (Tonga) - Sighted.Manua (Tonga) - Islanders on board.Tutuila (Samoa) - 15th July; found evidence of La

Perouse's visit (De Langle massacre).Vavua - Visited 19 July; named Curtis Sound; islanders

came on board with parrots; named Howe'sIslands by Captain Edwards.

Bickertons' Island - Sighted 21 July.Pylstaart Island - Sighted 23 July.Tongatabu - Landed for provisions 27 July.

Anamooka (Tonga) -29 July; second visit, waited until3 Aug for missing tender, extensive contact withislanders.

Niva'foo - Visited 5 August, Edwards named it Proby'sIsland; short stay.

Wallis Island - Visited 6 August; islanders on board.Rotumah - Visited 8 August; numerous canoes came off

to Pandora.Mitre Island - Sighted 12 August.Cherry Island - Sighted 12 August.Vanikoro I - Sighted 13 August.Meriam (Murray Island) - Sighted 25 August; named

by Captain Edwards.