The Bones in the Sediment

Seen from above, the search was a vast and unyielding process, a relentless effort that would continue because it could not stop.

From above, the Canadian Arctic Archipelago appears as a fracture zone between solid land and frozen sea, a region where sovereignty has always been negotiated rather than assumed. The administrative boundaries that govern this space—Nunavut, the Northwest Territories, the overlapping claims of federal authority and Indigenous self-government—represent a late-twentieth-century reconstruction of relationships that the Victorian navy would not have recognized as political at all.

In 2018, these boundaries acquired a new significance when the British government transferred ownership of HMS Erebus and HMS Terror to Canada and the Inuit, retaining only the right to repatriate human remains and any gold found aboard, along with a selection of relics.

The Inuit Heritage Trust received title. Parks Canada retained operational responsibility for archaeology. The partnership that emerged from this transfer rewrote the rules under which Franklin’s expedition could be investigated, replacing the race-against-time urgency of Victorian searchers with a deliberative process in which multiple parties held veto power over what might be disturbed.

The difference between this arrangement and its predecessors became concrete in August 2019, when a Twin Otter descended toward Terror Bay. The aircraft carried archaeologists from the federal research station, crossing the same bearing that Crozier’s party had followed south in April 1848. Below, at 68°54′13″N, 98°56′31″W, HMS Terror sat upright on the seabed at twenty-four meters, remarkably intact. Since the Arctic Research Foundation’s announcement of the discovery on 12 September 2016, the ship had become a fixed point in a shifting ethical landscape.

The wreck had been found to the south of King William Island in Terror Bay, in “pristine” condition. Erebus, found two years earlier in shallower water to the south, had already begun to yield artifacts: the ship’s bell, ceramic plates, a cannon, personal items that carried names and stories. But Terror remained more sealed, more complete, its hatches still closed, its interior still dark. The upright position suggested rapid sinking, perhaps with materials still in their final-use locations. The sealed compartments promised contexts that Erebus, broken and scattered, could no longer offer.

Marc-André Bernier, chief of underwater archaeology for Parks Canada, had overseen the operations since 2014. His teams worked under constraints that earlier Franklin searchers would not have recognized. The Victorian expeditions had collected human remains without ceremony, shipping skulls and bones back to England for identification and burial with full naval honors. McClintock’s party in 1859 had found skeletons in the boat on the western shore and taken what they could carry. Modern archaeology operated under different imperatives. The Nunavut Land Claims Agreement, the Historic Sites and Monuments Act, and the specific terms of the 2018 gift created a framework in which disturbance required justification, and human remains triggered automatic protocols of respect and potential repatriation.

The August 2019 operation had a limited window. Ice conditions in Terror Bay were favorable for only a few weeks each year. The team planned forty-eight dives over seven days, using cooperative weather to complete a three-dimensional photogrammetric map of the wreck. In 2018, a team had already examined the wreck using a remotely operated underwater vehicle (ROV) that collected photos and video clips. The goal now was documentation without intrusion: to understand the ship’s condition, to identify entry points, to plan for future work, but not yet to enter sealed spaces. The divers carried no tools for forced entry. Their cameras would record what light could reach, and the data would be processed in Ottawa through the winter.

On the third day, a survey pattern detected something unexpected.

The sonar image showed a linear feature inside the hull, approximately two meters in length, positioned on what would have been the lower deck near the officer’s quarters. The shape was ambiguous—too regular for natural debris, too indistinct for confident identification. It lay in sediment that had accumulated over 170 years, protected from currents by the ship’s intact structure. The archaeologists on the support vessel examined the image without speaking the obvious possibility. They had trained for this moment, theoretically, in conference rooms and policy documents. The partnership agreement specified procedures: immediate suspension of activities in the immediate area, notification of the Inuit Heritage Trust, consultation with community representatives, and a formal assessment of whether further investigation was warranted or ethical.

The feature was never definitively identified. Parks Canada would not confirm what the sonar suggested, and the raw data remained classified within the research files. But the possibility alone was sufficient to trigger the protocol. The divers shifted their operations to other sections of the wreck. The 3D mapping continued, but that compartment was marked for exclusion pending decision at a higher level. The season ended with the map incomplete in that sector, a deliberate gap preserved in the digital record.

This pause, this deliberate gap, marked a fundamental transformation in how the Franklin search could proceed. The Victorian evidence lag had measured the months between an Arctic event and its reception in London, the time required for ships to carry news through the Davis Strait and down the Atlantic coast. The modern delay operated differently. Information moved instantly from Terror Bay to Ottawa, from sonar screens to conference calls. What took time was not transmission but authorization: the process of consultation, the weighing of competing claims, the construction of consensus among parties with different relationships to the dead.

The question of what could be learned had to be weighed against the question of what should be done. Lead isotope analysis of bone could clarify whether the crew had been poisoned by their tinned provisions or by the ships’ distilled water systems, a debate that had persisted since the 1980s forensic work on Beechey Island. Subsequent research had suggested the distilled water systems as a potential source, with one scientist arguing it was “impossible to see how one could ingest from the canned food the amount of lead… required to raise” blood lead to poisoning levels. DNA analysis could identify individuals and perhaps trace family lines. Examination of skeletal pathology could reveal the progression of scurvy, the effects of cold, the evidence of terminal starvation including the extraction of marrow from bone ends. All of this would advance scientific understanding of the expedition’s final catastrophe.

Yet the same information could be pursued through less invasive means. The artifacts already recovered from Erebus—a hairbrush with initials, a boot sole, a piece of equipment modified for use on ice—told stories without requiring the disturbance of graves. The Inuit testimony, re-examined through modern anthropological methods by researchers such as David C. Woodman in the 1990s, offered narrative frameworks for understanding the final months. Woodman’s books, Unravelling the Franklin Mystery (1992) and Strangers Among Us (1995), had reconstructed probable sequences from the scattered accounts, challenging earlier theories by arguing that most of the 105 survivors cited by Crozier had survived past 1848, possibly re-manning a ship. The physical remains offered certainty, perhaps, but at a cost that the partnership agreement measured in terms of cultural respect and spiritual consequence.

The Inuit Heritage Trust’s position, as reported in subsequent documentation, emphasized continuity. The witnesses of the 1850s had been ancestors of the current membership. The stories passed down were not merely historical sources but obligations of relationship with the dead. The ships themselves had entered Inuit oral tradition as palangiq, things that had drifted into their waters and become part of their landscape. The 2018 gift recognized this continuity, transferring title not as a museum acquisition but as a return to proper custody. Within this framework, the disturbance of human remains required justification that went beyond scientific utility.

The British government’s retention of repatriation rights added another layer. If remains were found and identified as British subjects, a diplomatic process would begin. The Ministry of Defense’s Joint Casualty and Compassionate Center would be notified. Descendants would be sought. A decision would be made between burial in place, repatriation to England, or some intermediate arrangement. This process, developed for military aircraft crash sites and sunken warships with known casualties, had never been applied to a nineteenth-century exploration disaster of this scale. The paperwork alone would take years.

Bernier’s team continued their surface work. In 2017 and 2018, they had recovered artifacts from Erebus that spoke to individual lives without requiring the disturbance of human remains. A lieutenant’s epaulette, dropped or discarded in the final abandonment. A medicine bottle with traces of its original contents. A piece of wool fabric, still dyed the dark blue of naval uniform, preserved in the cold and darkness. Each item was photographed in situ, mapped into the 3D model, and raised only after documentation that would allow virtual reconstruction of its original context. The procedures had been developed for the Titanic and other deep-water sites, adapted for Arctic conditions, and now applied with the additional constraint of Inuit oversight.

The minute artifacts carried disproportionate weight. A hairbrush with the initials “F.H.” suggested Francis Hornby, mate on Erebus, who had been alive when the ships were abandoned in April 1848. A boot with a distinctive repair pattern matched descriptions in the expedition’s outfitting records, suggesting a particular cobbler’s work. These were proxies for the men themselves, standing in for bodies that could not be approached. The archaeological value lay in their specificity: not a generalized sailor’s possession but this sailor’s possession, found in this location, suggesting this action in final days.

The debate about the sonar feature continued through 2019 without resolution. Scientific conferences heard papers on the potential of ancient DNA from Arctic shipwrecks. Bioethicists discussed the special status of historical human remains, neither fully archaeological nor fully modern. Inuit representatives participated in these discussions not as consulted parties but as decision-makers with standing equal to the government archaeologists. The partnership model, developed in response to colonial patterns of extraction and appropriation, had created a new kind of delay—not the evidence lag of Victorian times, when news took months to reach London, but a deliberate pause for ethical consideration.

The Victorian searchers would not have recognized this hesitation. For them, the recovery of human remains had been a duty to the dead and their families, a way of converting anonymous disappearance into named tragedy with proper rites. McClintock’s 1859 expedition had found skeletons in the boat place and taken skulls for possible identification. Later searchers had collected bones from Beechey Island and elsewhere, shipping them to England for burial with honors that the original deaths had denied. The scientific value of such remains had been recognized—phrenological examination, crude by modern standards, had been attempted—but the primary framework had been one of rescue from anonymity and neglect.

Modern archaeology had inverted this priority. The dead had rights that limited what the living could do, even in the name of historical understanding. The partnership agreement expressed this inversion in legal terms, but its roots lay in broader shifts: the repatriation movement of the late twentieth century, the recognition of Indigenous title, the development of bioethical frameworks that treated human tissue as carrying continuing claims of personhood. Franklin’s men, dead before any of these frameworks existed, were nonetheless governed by their application.

The sediment that preserved them became a kind of custody. The cold, dark water of Terror Bay, with its low oxygen and minimal biological activity, had created conditions for exceptional preservation. Wood, fabric, paper, bone—all survived where they would have decayed in warmer seas. But this preservation also created temptation. The interior of Terror was a time capsule, potentially containing not just the material culture of 1845 but the biological remains of the crew themselves.

The 2019 season ended without entry. The 3D map was completed for the accessible portions, showing a vessel still remarkably intact: masts standing, hatches closed, the hull seated firmly in sediment that had built up around it. The sonar feature remained unexamined. The file moved between Ottawa and the regional Inuit organizations, between scientific committees and heritage boards, accumulating the documentation that would eventually support a decision or justify continued deferral.

In England, the Franklin family descendants maintained their interest. Some had provided DNA samples for comparison with any remains that might be recovered. Others had expressed preference for burial in place, arguing that the Arctic had become the sailors’ final home by their own efforts and suffering. The Ministry of Defense held the formal authority for repatriation decisions, but in practice deferred to the partnership arrangement in the north. The imperial power that had sent the expedition now waited on permissions from the territory it had once claimed as wilderness.

The scientific questions remained urgent in their way. The lead poisoning hypothesis, advanced by Owen Beattie’s 1980s forensic work on the Beechey Island graves, had been challenged by subsequent research suggesting the ships’ distilled water systems as an alternative source. The tinning of provisions, once blamed for catastrophic contamination, might have been less significant than the lead joints in the desalination apparatus. Only analysis of remains from the final phase of the expedition—men who had survived the first winters and died on the retreat south—could resolve this question. The Beechey Island bodies had been early casualties, buried in January 1846 when the expedition was still functioning. The real catastrophe came later, in 1847 and 1848, and its victims lay somewhere in the ice or in the ships.

The Inuit testimony, as collected and analyzed by Woodman and others, suggested multiple scenarios. Some accounts spoke of a ship sinking off the western coast of King William Island with men aboard. Others described a camp on the southern shore where survivors had lingered before attempting the final march. The location of Terror in Terror Bay, far from the known line of retreat, required explanation: had some of the crew returned to the ship? Had it been re-manned as a refuge? Later reviews of archaeological data from the 1990s had concluded that separate sites on the coast likely marked where boats on sleds ended up after the crew tried to reach Back River. The physical evidence, if it could be reached without violation, might answer questions that the testimony had left open.

But each potential answer carried ethical weight. To enter the ship was to risk disturbance. To identify remains was to initiate claims of repatriation and burial. To analyze bone was to extract information that the dead had not consented to provide, even if such consent could be meaningfully constructed for men who had died before the invention of modern bioethics. The partnership agreement had not resolved these tensions but had institutionalized them, creating a structure in which decisions required multiple approvals and in which the default position was preservation in place.

The 2020 field season was cancelled by pandemic restrictions. The 2021 season operated under reduced capacity. The sonar feature in Terror remained unexamined, its interpretation still provisional, its existence known only to those with access to the research files. The 3D map circulated in academic and heritage contexts, demonstrating what could be achieved without intrusion. The artifacts from Erebus continued their conservation treatment in Ottawa, their analysis proceeding through methods that required no additional disturbance of the wreck site.

The search had found the ships. It had not found, in any confirmed sense, the men. The graves on Beechey Island held three bodies from the expedition’s first winter. The scattered remains found by McClintock and subsequent searchers accounted for perhaps a dozen more. The rest—over a hundred men—lay somewhere in the ice, on the land, or in the waters where their ships had finally settled. The possibility that some remained aboard, preserved in the cold and dark, transformed the wrecks from archaeological sites into potential cemeteries.

This transformation had consequences for how the search could proceed. The technical challenges of Arctic diving, always severe, were now compounded by legal and ethical constraints that had no Victorian precedent. The scientific value of potential remains, always significant, was now weighed against obligations of respect that came from multiple sources: international heritage conventions, Canadian law, Inuit tradition, and the specific terms of the 2018 gift. The partnership model, developed to correct colonial patterns of extraction, had created a new kind of relationship with the past—one in which the dead retained claims that limited what the living could do in their name.

The establishment of protocols for dealing with potential human remains created a concrete consequence: the inquiry was now permanently framed by a modern ethical dilemma, handing off the need for a final synthesis in which the desire to know must be justified against the obligation to let rest.