Long Island Investigation

Long Island was called Seals' Island by Batavia survivors and is a coral island approximately 1600m long and 180m at its widest point at the northern extremity.

Archival sources record that Seals' Island (Long Island) was the location for:

  1. Occupation and slaughter site of 45 Batavia survivors
  2. Gallows site of Batavia mutineers
  3. Mutineers’ prison.

Between 45 and 60 survivors are known to have sheltered on Seals’ Island (Long Island) with all but ‘seven boys and some women’ subsequently murdered (Drake-Brockman, 1963:159).

Long Island is historically, culturally and archaeologically significant in view of its confirmed association with the survivors of the wreck of Batavia. Although few artefacts have been found, the island has never been subjected to close archaeological scrutiny, except for surface exploration for potential occupation sites. No excavation has been undertaken on the island by the Western Australian Museum and the potential for such work remains.

Map showing where the surveys on Long Island had occurred over time

Long Island Surveys
Image copyright WA Museum


Early Finds

Long Island has twice been subject to a visual and remote sensing survey using a metal detector by the Department of Maritime Archaeology, WAM (in 2001 and 2005) (Souter, 2005). In these two survey efforts, which mark the first systematic investigation to be conducted on the island, the northern half of the island was covered. After the discovery of material believed to be contemporary with the Batavia mutiny in both surveys, it has been recommended to continue the survey efforts in order to cover the entire Long Island area.

In 2007, the Department of Maritime Archaeology scheduled to include a continuation of the Long Island survey as part of its NHL program. Three days were planned to survey the southern premises of the island. Like previous survey efforts, the aim was to locate evidence of habitation, including the execution site where the mutineers were tried and hanged according to 17th Century Dutch conventions of justice.

Previous finds have included a beardman jar shard (LIWAM 67) found in 1967; a lead morning star weapon (BAT 3923) found in 2001; square head nails and a ship’s fastening (BAT 80531, BAT 80532, BAT 80533 and BAT 80534) found in 2005. A large copper alloy pintle (BAT 3918), was donated to the Museum during the Amnesty and is believed related to the Batavia Fort stone portico, used to support a large hinged wooden gate. This object was reportedly found in shallow water off the anchorage and landing place on the eastern side of Long Island. Given the discovery of several iron fastenings in 2007 in the vicinity of the anchorage and landing place, this object takes on a renewed significance as part of the Long Island assemblage of Batavia artefacts.

A copper alloy object covered in corrision

Large copper alloy pintle (BAT 3918), believed related to the Batavia Stone Portico
Image copyright WA Museum

A research strategy and predictive model for the identification of a location for the gallows site was based on the likely construction (timber, iron fastenings, possible use of wreck timbers by ships’ carpenters), anticipated site formation processes (collapse, partial burial) and possible clothing and human remains (buttons, textiles, teeth, bones) if the bodies were left to rot on the gallows or buried (Souter, 2005: 9).

Although historical evidence from Pelsaert’s journal does not indicate whether the mutineers were left on the gallows, it is unlikely that they were taken down and/or buried after their execution. According to Dutch practice of the period 'earth was not given or permitted' for those who were executed (Mak, 2001:136). The gallows fields of Amsterdam, were well known for the decomposing bodies of criminals that were purposely left exposed to the elements.

An unusually shaped stone-like artefact

Morning Star 2001
Image copyright WA Museum

A shard recovered from a pot

Beardman Pot Sherd, 1967
Image copyright WA Museum

Survey Methodology

The survey methodology was to use a Fugro Omnistar 8300HP DGPS to fix all control, transect, and anomaly points with an accuracy of a few centimetres in horizontal plane. During the 2005 survey, information was provided by Barry Humfrey (Humfrey Land Developments) for existing survey marks on the island CPLI01 (Control Point Long Island 01) and CPLI02 (Control Point Long Island 02). These points were relocated for 2007 Long Island Survey and the total station was positioned on CPLI01 - the northern-most survey mark. Survey point CPLI02 was used to orient the total station survey.

A 100m fibreglass tape was used to lay the baseline down and two 30m tapes were placed perpendicular to the baseline three to five metres apart. These tape transects marked the metal detecting survey lanes and were set at right angle from the baseline with a compass. After each 100m baseline area was surveyed, the baseline tape would be moved in line from north to south along the island. In total, five 100m baselines (BL01-05) were laid down and surveyed. Endpoints of these baselines are referred to as BL01S and BL01N (for ‘Baseline 01 South Datum’ and ‘Baseline 01 North Datum’). Transect line endpoints were recorded as e.g. T01W-T01E (Transect 01 West – Transect 01 East). Artefacts or anomalies were recorded as A001, A002 etc. Baseline, transect line and anomalies were fixed using the DGPS during the first three days and using the total station on the last day (as the DGPS antenna did not pick up enough satellites to provide accurate data.)

As in previous surveys of Long Island, one person acted as surveyor operating the DGPS to measure in the points of the baselines and transects, and to record the positions of any anomalies or artefacts found. Another person operated the metal detector to locate artefacts assisted by two persons who would help to locate targets, and move and offset tapes of transect and baselines as the detector moved along the island.

Total station points were measured in by one of the metal detecting assistants and the surveyor. Communication between the operator of the total station and their assistant was facilitated by using hand-held UHF radios. According to survey instructions, the metal detector operator walked along either side of the transect lines moving the detector head coil from side to side to cover a 1.5m transect area approximately 10cm above the ground surface. This detection method and the previously described surveying method theoretically ensures a full coverage of the survey area, with an average 10 to 20cm overlap between transects to guarantee full coverage. At the end of each day, the survey data was downloaded from the DGPS or total station and imported into Arcview GIS software.

It should be kept in mind, however, that neither visual survey nor the use of a metal detector will provide conclusive evidence to answer the questions listed above as both methods are not capable of detecting buried organic remains, such as wood or skeletal material. Therefore, possible burial sites on Long Island may remain undiscovered using the current exploration method. Furthermore, organic materials are unlikely to have survived if exposed to the elements on the surface.

Two scientists performing a survey on Long Island

Photo from the Long Island Survey 2005
Image copyright WA Museum

Survey Constraints

a) Environmental

Sandy sections along the high ridge of the island are occupied by nesting shearwaters. It was necessary to operate the metal detector very carefully through these highly sensitive areas in order to avoid treading on and collapsing burrows, thereby slowing the survey process.

In addition to this, in some areas the island’s vegetation was too high or dense, particularly in those areas along the highest ridge of Long Island. Here, the metal detector head coil could not be held at a distance of 10cm or less from the ground. Coverage of these areas cannot be regarded as sufficient or accurate. Moreover, young chicks of nesting birds used these bushes as a hiding place at the time of survey. It was, therefore, decided not to include these areas in order to leave the flora and fauna of the island undisturbed.

Furthermore, the sandy beach and deposition of coral rubble directly inshore from it were not included in the Long Island survey as this area is continuously disturbed by dynamic wave action and tidal changes. It is highly unlikely that any historic cultural remains would have been preserved in this area.

b) Equipment

The metal detector used for the 2007 Long Island Survey is capable of detecting small, buried objects at a range of 10-30 cm. However, small isolated artefacts situated between the coral shingles or in soil at a depth greater than 30cm, or below a bush greater than 30cm height could not be detected.

One scientists performing a survey on Long Island

Photo from the Long Island Survey 2005
Image copyright WA Museum

c) Time

Time restraints did not allow the completion of a systematic survey of the southern-most area of Long Island (approx. 250m in length). This area remains un-investigated to date and it is recommended that the archaeological survey be completed in the near future.


A large concentration of wrought iron bolts and fastenings was found on Long Island in the course of the metal detector survey. Numerous remnants of iron nail heads and shafts, and three iron bolts were found in an area approximately 4x4m. These fastenings are consistent with 17th-19th Century ship construction; the nails having square-sectioned shafts and cylindrical iron bolts. The nails are poorly preserved and hardly have any diagnostic features. None of the nail heads, for example, still has its original shape or dimensions, and none of the nail tips were found. The last stage of corrosion of metal fasteners is usually represented by remaining nail heads and their upper shafts as this part is the thickest and has the largest material density - this is exactly what is found in the Long Island fastening assembly.

The best-preserved remnant of a wrought iron bolt probably represents a clinch bolt with a thick washer (BAT 80547; pres. dia. 0.054; pres.th. 0.021), similar to those found in the Vasa ship (McCarthy, 2005:64, 70). Vasa was built according to a Dutch bottom-based construction method and is contemporary to Batavia. The other two bolts found on Long Island are too poorly preserved; their heads and ends are missing (BAT 80540, BAT 80541). The preserved diameter, 0.033, of one of them (BAT 80541) is similar to the size of the bolts found on the Batavia ship’s hull.

The site is consistent with the predicted remains of a gallows site, being a sufficiently large concentration of a range of iron fastenings used for fastening structural timber pieces and possibly decking.

A corroded bolt with washer

Clinch bolt with washer, BAT80547
Image copyright WA Museum