Victoria Strait, 22 April 1848: the lost Franklin Expedition

The case turns on a single sheet of printed paper, recovered eleven years after it was written. On 5 May 1859, in a stone cairn at Victory Point on the northwest coast of King William Island, Lieutenant William Hobson found a standard Admiralty message form with two messages written around the margins. The first, dated 28 May 1847 and signed by Lieutenant Graham Gore and Charles Frederick Des Voeux, ended with the words “All well.” The second, dated 25 April 1848 and signed by Captains Francis Crozier and James Fitzjames, recorded that HMS Erebus and HMS Terror had been deserted on 22 April, beset since 12 September 1846; that 105 souls under Crozier had landed at 69 degrees 37 minutes 42 seconds north, 98 degrees 41 minutes west; that Sir John Franklin had died on 11 June 1847; that the losses to that date were 9 officers and 15 men; and that the survivors would start the next day for Back’s Fish River.

No survivor reached a European or Hudson’s Bay Company post. In 1854 the surveyor John Rae brought back Inuit testimony that the marchers had starved and that the bodies bore signs of cannibalism in the final stages. Charles Dickens, in print, called the Inuit witnesses “covetous, treacherous, and cruel.” For the next 160 years the cause of death was contested across a sequence of forensic chapters: lead poisoning, scurvy, botulism, tuberculosis, the ships’ water system. In September 2014 Parks Canada announced the wreck of HMS Erebus in Wilmot and Crampton Bay, far south of the abandonment position. In September 2016 the Arctic Research Foundation located HMS Terror in Terror Bay. Both finds rested on Inuit oral history. In 2024 DNA from a mandible at Erebus Bay identified Captain Fitzjames, and the same mandible bore cut marks consistent with peri-mortem dismemberment.

The ships are on the chart. The cumulative cause of the 129 deaths is not. We keep three things separate: what is documented, what the evidence shows, and what is only a hypothesis.

The documented account

The Franklin Expedition was a Royal Navy attempt on the Northwest Passage, fitted out in 1845. Command went to Sir John Franklin, then 59, a polar veteran and former Lieutenant-Governor of Van Diemen’s Land. HMS Erebus sailed under Franklin with James Fitzjames as second. HMS Terror sailed under Francis Crozier, an Antarctic veteran of the James Clark Ross expedition of 1839 to 1843 in which both ships had already served. Both vessels carried iron bow plating, internal bracing against pack-ice pressure, and an auxiliary steam engine drawn from a railway locomotive of about 25 horsepower. The Erebus engine, named Croydon, came from the London and Croydon Railway; one strand of the secondary literature gives the London and Greenwich Railway instead, and the engine’s name rather than its operator is the firmer attribution. Each ship carried a new pressurised desalination and heating system fed by lead-soldered pipes. The Admiralty’s contract for tinned provisions, signed on 1 April 1845, went to Stephan Goldner of Houndsditch.

The expedition left Greenhithe on 19 May 1845 with 134 men. Five were discharged sick before the ships entered the North American Arctic, and the canonical complement of 129 is the figure used in every subsequent report; the port of discharge varies between sources from Greenhithe to Stromness to Disko Bay. In late July 1845 the whalers Prince of Wales (Captain Dannett) and Enterprise (Captain Robert Martin) met Erebus and Terror in Baffin Bay, awaiting conditions to enter Lancaster Sound. That was the last European sighting.

The expedition wintered in 1845 to 1846 at Beechey Island in Lancaster Sound. Three men died there: John Torrington (Terror) on 1 January 1846, John Hartnell (Erebus) on 4 January 1846, and William Braine (Erebus) on 3 April 1846. Their headboarded graves were discovered in 1851 by William Penny and Erasmus Ommanney during the Austin search expedition; some sources date the discovery to late August 1850. In 1984 and 1986 Owen Beattie of the University of Alberta exhumed the three men. The permafrost-preserved bodies returned an autopsy of pneumonia, with tuberculosis complicating Torrington’s case, and hair, nail, and bone samples with elevated lead concentrations.

In the summer of 1846 the expedition left Beechey, transited Peel Sound and Franklin Strait, and entered Victoria Strait between King William Island and Victoria Island. The Victory Point note records that both ships were beset on 12 September 1846, and that Franklin died on 11 June 1847; it does not state a cause. The original is held at the National Maritime Museum, Greenwich, and the transcription used here follows the museum’s facsimile via the Canadian Mysteries archive.

After the 22 April 1848 abandonment, the 105 survivors marched south down the west coast of King William Island toward Back’s Fish River. They left a trail of skeletons, abandoned sledge gear, a ship’s boat mounted on a heavy oak sledge at Erebus Bay, Royal Navy buttons, silver cutlery engraved with officers’ crests, books, and a copy of The Vicar of Wakefield, recovered piecemeal over the next century and a half.

In April 1854, working overland near Pelly Bay and Repulse Bay, the Hudson’s Bay Company surveyor John Rae met Inuit carrying Franklin Expedition artifacts. His principal informant was an Inuk recorded by Rae as In-nook-poo-zhe-jook, in modern orthography Inukpujijuk. The Inuit had not seen the events themselves; they relayed accounts from others: about forty white men had been seen dragging a boat and sledges south down the west coast of a large island, had died of starvation near the mouth of the Great Fish (Back) River, and the condition of the bodies and the contents of the kettles indicated cannibalism in the final stages. Rae purchased relics directly from his informants, including silver plate engraved with officers’ initials and a small Hanoverian Guelphic Order medal attributed to Franklin (the catalogue attribution at the National Maritime Museum is the controlling reference). His report was published in The Times on 23 October 1854. Charles Dickens replied in Household Words in December 1854, rejecting the Inuit testimony as the product of “covetous, treacherous, and cruel” people. Lady Franklin publicly disputed Rae. The Admiralty paid Rae the 10,000 pound reward.

Modern scholarship has fully vindicated Rae and his Inuit informants.

After the Admiralty closed its own search, Lady Franklin commissioned the steam yacht Fox under Francis Leopold McClintock at her own expense. McClintock wintered in eastern Bellot Strait and dispatched sledge parties in the spring of 1859. Hobson found the Victory Point note on 5 May 1859. McClintock’s parties recovered the open ship’s boat at Erebus Bay with two skeletons inside, conducted further Inuit interviews, and published The Voyage of the Fox in the Arctic Seas that year. Charles Francis Hall continued the Inuit interviews through the 1860s. Lieutenant Frederick Schwatka’s American Geographical Society expedition, with Heinrich Klutschak, William H. Gilder, Frank Melms, and the Inuit guide Ebierbing, spent more than a year on King William Island and the Adelaide Peninsula (1878 to 1880) and mapped Starvation Cove.

In 1987 Owen Beattie and John Geiger published Frozen in Time, setting out the integrated hypothesis that cumulative lead poisoning from the Goldner tins had driven the cascade of deaths after April 1848. The hypothesis dominated public understanding for two decades, then was challenged on multiple fronts: William Battersby (Journal of the Hakluyt Society, 2008) argued the principal lead source was the ships’ own pressurised water system rather than the cans; Millar, Bowman, and Battersby (Polar Record, 2014 to 2015) re-analysed the bone-lead data and found the levels elevated but not acutely fatal and broadly consistent with mid-Victorian baseline exposure; and Christensen and colleagues (PLOS One, 2017 to 2018), using high-resolution confocal X-ray fluorescence and isotopic work, concluded lead is unlikely to have been the decisive cause. Russell Taichman and colleagues (Arctic, 2017) proposed chronic tuberculosis with secondary adrenal insufficiency, or Addison’s disease, to account for the wasting and collapse documented in the Inuit testimony.

On 2 September 2014 an Inuit tent ring in the southern Parks Canada search area produced an iron davit fitting from a Royal Navy ship. Within days a Parks Canada side-scan survey located HMS Erebus in Wilmot and Crampton Bay, off the Adelaide Peninsula, in 11 metres of water, far south of the documented April 1848 abandonment. Prime Minister Stephen Harper announced the find on 9 September 2014. The locating information was the Inuit oral tradition collected over thirty years by Louie Kamookak of Gjoa Haven, integrated with previously dismissed testimony from Rae, McClintock, Hall, and Schwatka. The Erebus bell was raised that November. Kamookak was named to the Order of Canada and the Order of Nunavut and died on 22 March 2018.

On 3 September 2016 the Arctic Research Foundation’s research vessel Martin Bergmann, acting on a tip from the Inuk crew member Sammy Kogvik of Gjoa Haven, who had seen a mast-like spar protruding from sea ice years earlier, located HMS Terror in Terror Bay off the southwest coast of King William Island in about 24 metres of water. The discovery was announced on 12 September 2016 and confirmed by Parks Canada on 18 September. The position is approximately 92 km south of the abandonment and about 50 km from the Erebus wreck.

In 2018 Parks Canada and the Inuit Heritage Trust signed a co-management memorandum for the wrecks, formalising Inuit oral history as primary evidence on equal footing with the documentary record. Dive seasons through 2026 have recovered scientific instruments, leather book bindings, and a daguerreotype case (firm published confirmation of the daguerreotype recovery should be checked against the most recent Parks Canada season report). In 2024 a team led by Stephen Fratpietro and Douglas Stenton (Lakehead and Waterloo) identified by Y-chromosome match a mandible from the NgLj-2 Erebus Bay site as Captain James Fitzjames, the first officer and second man identified by DNA. The same mandible bears cut marks consistent with peri-mortem dismemberment.

The evidence

The documentary record is the Victory Point note, Rae’s 1854 Admiralty report, McClintock’s 1859 narrative, the Schwatka accounts of Klutschak and Gilder, and the Admiralty papers at Kew (ADM and BJ series).

The physical record is the Beechey Island graves of Torrington, Hartnell, and Braine, exhumed in 1984 and 1986 and reinterred with samples retained; multiple skeletal sites on King William Island and the Adelaide Peninsula, including the Erebus Bay site NgLj-2; the wrecks of Erebus in Wilmot and Crampton Bay and Terror in Terror Bay; and artifacts held at the National Maritime Museum, the Royal Geographical Society, the Polar Museum at Cambridge, Parks Canada, and the Canadian Museum of History, including silver cutlery, Royal Navy buttons, the Erebus bell, scientific instruments, and Goldner tins.

Inuit oral history is the third evidentiary strand, and it is primary evidence, not folklore. The chain runs from In-nook-poo-zhe-jook to Rae in 1854; through McClintock’s Inuit interviews of 1857 to 1859, including Oonalee, who drew the map directing the search to Victory Point; through Hall in the 1860s; through Schwatka and Klutschak in 1878 to 1880; into the modern oral-history collection compiled by Louie Kamookak over thirty years; and into the 2018 Inuit Heritage Trust co-management agreement. The 2014 and 2016 wreck discoveries confirmed this oral record had been geographically correct all along.

The skeletal evidence on cannibalism rests on three peer-reviewed studies. Keenleyside, Bertulli, and Fricke (Arctic, 1997) documented cut marks on approximately one quarter of 400 bones recovered at Erebus Bay, in patterns consistent with systematic dismemberment for marrow extraction. Simon Mays and Owen Beattie (International Journal of Osteoarchaeology, 2016) identified additional pot-polishing marks consistent with end-stage cannibalism. The 2024 Fitzjames identification confirms the practice was not confined to the lower ranks. The evidence documents a final stage of cumulative starvation on men who had marched for months, not a moral failure.

The theories

Everything in this section is a hypothesis.

A. Lead poisoning, primary cause (Beattie and Geiger 1987). For: elevated bone, hair, and nail lead at Beechey Island; Goldner contract irregularities; poor decisions in the surviving record. Against: the bone-lead re-analyses (Millar et al. 2014 to 2015; Christensen et al. 2017 to 2018) find the levels insufficient to be acutely fatal and consistent with Victorian baseline; the Beechey three died after only nine months in the ice.

B. Cumulative scurvy and starvation (Cookman 2000). For: documented provision depletion; consistency with Inuit descriptions of the survivors’ physical state; the classical Arctic pattern. Against: does not by itself explain the speed of mass mortality after the march began.

C. Botulism from canned food (Horowitz 2003). For: documented under-processing of Goldner cans; clinical match to some Inuit-reported symptoms. Against: no surviving documentary or skeletal evidence directly supports it; no Goldner can has been shown to contain Clostridium botulinum residue.

D. Lead from the ships’ water system (Battersby 2008). For: technical analysis of the lead-soldered pipework of the desalination system is consistent with the lead profile observed; later isotopic work is broadly compatible. Against: still requires lead to do mortality work the bone-chemistry re-analyses say it cannot do alone.

E. Tuberculosis with adrenal insufficiency (Taichman et al. 2017). For: tuberculosis documented in Torrington at autopsy; endemic in Royal Navy crews of the period; clinical course fits. Against: does not on its own account for the skeletal cannibalism evidence or the geographical distribution of the deaths.

F. The combined-cause reading. For: the multi-modal evidence (Keenleyside 1997, Mays and Beattie 2016, Stenton 2024) is consistent with starvation, lead exposure from any source, tuberculosis, hypothermia, and end-stage cannibalism acting together. Against: an over-determined answer does not isolate a primary mechanism, and so does not close the question.

G. The ships’ southward movement. For: the 2014 and 2016 wreck positions lie far south of the abandonment; Inuit oral history from Rae through Kamookak consistently describes one ship sinking off King William Island and another boarded by Inuit before it sank further south; Parks Canada dive seasons (2021 to 2024) document features of Erebus consistent with active use after April 1848. Against: no documentary record of any re-manning survives, and ice drift is an alternative.

What remains unknown

178 years after the abandonment, five things are still open. The precise cumulative cause of the deaths after 22 April 1848. Whether one or both ships were re-manned or drifted south unmanned. The precise route and final fate of the Crozier and Fitzjames party. Which expedition member was the last to die, and where. The location of Sir John Franklin’s burial. The 2024 Fitzjames identification raises the question whether Franklin himself may yet be identified from the existing skeletal record.

The ships have been found. The Inuit oral history that guided the search has been formally restored to the evidentiary record. The cause of death of 129 men has not.

Victoria Strait, 22 April 1848: the lost Franklin Expedition

The documented account

The evidence

The theories

What remains unknown

Sources

Primary

Secondary