The Listing Coal Carrier off St Augustine: The Loss of the SS Cotopaxi, 1 December 1925

On 29 November 1925 the SS Cotopaxi, a steel-hulled bulk freighter of 2,351 gross register tons, cleared Charleston, South Carolina under Captain W. J. Meyer (rendered “William J. Meyers” in the JaySea Archaeology summary of the subsequent court records) with a crew of thirty-two and 2,340 long tons of bituminous coal in her holds. She was bound for Havana, a four-and-a-half to five-day run under steam. On or about 1 December, as a tropical storm crossed southern Florida, a wireless distress message reporting that the ship was listing and taking on water was received at Jacksonville, Florida. The Cotopaxi never reached Havana. She was formally listed as overdue on 31 December 1925. No survivors were recovered. No bodies were recovered. No wreckage was located in the immediate search.

The case did not stay quiet. In February 1964 the journalist Vincent Gaddis named the Cotopaxi in his Argosy Magazine article “The Deadly Bermuda Triangle” and in 1965 in his book Invisible Horizons. Charles Berlitz repeated her in The Bermuda Triangle in 1974. In 1977, Steven Spielberg’s Close Encounters of the Third Kind put her by name in the Gobi Desert sequence, where the missing vessels are returned by an alien intelligence, and that single cameo did more than any of the paranormal books to fix the ship’s name in popular memory. In February 2020 a different kind of attention reached her. Marine biologist Michael Barnette and the Association of Underwater Explorers announced, on the premiere of the Science Channel docuseries Shipwreck Secrets, that a wreck site long known to local sport divers off St Augustine was the Cotopaxi. The geographic mystery, such as it ever was, closed. The cause-of-loss question did not. We keep three things separate, as always: what is documented, what the evidence shows, and what is still only a hypothesis.

The ship and the voyage

The Cotopaxi was built in 1918 at the Great Lakes Engineering Works River Rouge Yard in Ecorse, Michigan, hull number 209. She was laid down on 29 August 1918, launched on 15 November 1918, and commissioned on 30 November 1918. She measured 253 feet in registered length, 43 feet 8 inches in beam, and 25 feet 6 inches in depth. Her registered tonnages were 2,351 gross, 1,433 net, and 4,062 deadweight. She was driven by a Great Lakes Engineering Works three-cylinder triple-expansion steam engine of 1,350 indicated horsepower, fed by two coal-fired Scotch marine boilers, for a service speed of roughly 9 knots.

She entered service under the United States Shipping Board and was sold on 23 December 1919 to the Clinchfield Navigation Company, a subsidiary of the Clinchfield Coal Corporation, for $375,000. Clinchfield put her into the standard Charleston to Havana coal trade. By late November 1925 she had been in that route for nearly six years.

Her final voyage began at Charleston on 29 November 1925 under Captain Meyer. The standard reported complement is a crew of thirty-two; one secondary account, propagated through the modern Wikipedia treatment of the 1925 Florida tropical storm, gives a crew of thirty, and the figure of thirty-two is the Lloyd’s-derived number repeated across the major press coverage of the 2020 identification. Her cargo was 2,340 long tons of bituminous coal. Her destination was Havana. Her expected run, in good weather, was a little under five days.

The hatch-cover testimony

The most consequential piece of pre-departure record is summarised by JaySea Archaeology from the civil action that followed the loss. In late November 1925 the Clinchfield carpenter C. N. Coates inspected the existing hatch covers with Captain Meyers and found them, in the language attributed to the court file, in “deplorable” condition with large holes. Coates fabricated new wooden hatch covers in response.

The Cotopaxi sailed for Havana without the new covers. The new covers were reported to have been left on the dock.

This is testimony, and it sits inside one summary of the original court filings rather than in the original filings themselves; the underlying record should be verified directly against the civil action file before any quoted detail is treated as exact. Even handled cautiously, it is load-bearing. A bulk freighter carrying 2,340 long tons of coal into a tropical storm with hatch covers documented as defective is a vessel whose loss does not require any explanation more exotic than the one in front of it.

The wireless distress and the loss

On or about 1 December 1925 the Cotopaxi transmitted a wireless distress message. The standard reported content is that she was listing and taking on water in a tropical storm. The St Augustine Lighthouse & Maritime Museum, in its press release announcing the Shipwreck Secrets identification, states that the distress signals were “picked up in Jacksonville, Florida” and that contemporaneous wireless distress data was located in the archives of Lloyd’s of London during the research for the program.

The form of the message is disputed in the secondary record. The SS Cotopaxi Wikipedia article and the major 2020 press coverage describe a formal distress with a position transmitted. The Wikipedia article on the 1925 Florida tropical storm states, in direct contradiction, that “no actual SOS was ever issued nor was Cotopaxi’s position telegraphed with these final messages.” The two readings cannot both be right. The authoritative resolution path is the Jacksonville shore-station wireless log and the Lloyd’s of London distress archive, which the Science Channel research team is reported to have consulted. Until those originals are read directly, the safest summary is that a wireless distress was received in Jacksonville on or about 1 December, that it reported a list and water ingress, and that whether it was a coded SOS with a fixed position is genuinely open in the secondary literature.

The ship never arrived in Havana. She was formally listed as overdue on 31 December 1925. No survivors were recovered. No bodies were recovered. The immediate search produced no wreckage and no debris field.

The 1925 tropical storm

The weather is the one part of the picture that the modern record reconstructs cleanly. A tropical storm was active in the western Caribbean and southern Florida region from late November 1925, first identified on 27 November southeast of the Yucatán Peninsula, reaching peak sustained winds of roughly 65 miles per hour, and making landfall just south of Fort Myers, Florida on 1 December 1925. Modern reconstructions of the storm track place the Cotopaxi loss inside the system. The storm is not a competing hypothesis for the loss. It is the weather the ship was in when she sent her distress message.

The Bermuda Triangle attribution and the Spielberg cameo

The Cotopaxi entered popular paranormal literature decades after the loss. Vincent Gaddis published “The Deadly Bermuda Triangle” in Argosy Magazine in February 1964 and folded it into his 1965 book Invisible Horizons. Charles Berlitz built on Gaddis in The Bermuda Triangle in 1974. Larry Kusche’s 1975 The Bermuda Triangle Mystery Solved (revised 1995) was the first systematic skeptical examination of the Triangle case list and argued that most entries had conventional explanations within the ordinary loss probability for ocean shipping of the period. Gian J. Quasar returned to a paranormal-leaning treatment in Into the Bermuda Triangle in 2003.

The geography of the Triangle attribution is not as clean as the literature suggests. The reported position of the Cotopaxi’s distress, off the northeast Florida coast, sits at the western edge of, or just outside, the canonical Miami to San Juan to Bermuda triangle drawn by Gaddis. Barnette has publicly emphasised this point.

The single piece of culture that put the ship’s name into general circulation was not a book. Steven Spielberg’s Close Encounters of the Third Kind, released in November 1977, includes a sequence in which a search team in the Gobi Desert finds a row of ships that an alien intelligence has returned. One of them is named onscreen as the SS Cotopaxi. This article reports Gaddis 1964, Berlitz 1974, Quasar 2003, and the Spielberg cameo as documented popular-culture artifacts. It does not adopt any of them as framing. Kusche’s 1975 skeptical reading and Barnette’s 2020 public characterisation of the paranormal framing as “total rubbish” are reported as the documented institutional and expert positions they are.

The Bear Wreck

A wreck site roughly 35 nautical miles off St Augustine, Florida, in approximately 100 to 105 feet of water, had been known to local sport divers since the mid-1980s. It was nicknamed the “Bear Wreck.” Michael Barnette, a marine biologist with NOAA’s National Marine Fisheries Service and the founder, in 1996, of the Association of Underwater Explorers, worked the identification across roughly fifteen years.

The dating of the identification is best read as a working identification preceding a public one. The Michael C. Barnette Wikipedia entry reports the wreck as “positively identified in 2014 by the Association of Underwater Explorers” at 36 nautical miles off St Augustine in 105 feet of water. The SS Cotopaxi Wikipedia article and the major press coverage in 2020 date the public identification to the late January and early February window around the Science Channel announcement. The two should not be read as a contradiction. The 2014 figure is the AUE working identification; the 2020 figure is the public announcement carried by the premiere of Shipwreck Secrets, season one episode one, on 9 February 2020. The original AUE publications and Barnette’s Wreckdiving Magazine write-ups should be consulted to resolve the precise sequence.

The identification rests on a stack of cumulative evidence. The wreck’s location aligns with the Cotopaxi’s last reported wireless position to within roughly twenty nautical miles, per the JaySea Archaeology summary. The wreck’s measured dimensions match the Cotopaxi’s 253-foot length and 43-foot-8-inch beam. The boiler arrangement matches the Great Lakes Engineering Works plans for hull 209: two Scotch boilers, one propeller. Markings on a recovered valve component identify it as a Scott Valve Manufacturing item from Michigan, consistent with a Great Lakes-built ship of 1918. The general bulk-freighter configuration is the “stemwinder” type. Chuck Meide and Brendan Burke of the St Augustine Lighthouse Archaeological Maritime Program, the museum’s archaeology arm, dove the site and corroborated the identification independently.

What the identification does not yet rest on is a single name-bearing artifact. No ship’s bell, builder’s plate, or registry plate carrying the name Cotopaxi has been recovered from the site in any reporting reviewed for this account. The identification is cumulative and, in the working maritime-archaeology sense, very strong. It is not yet absolute. AUE has conventionally withheld the exact coordinates of the site to protect it from looting, which is standard practice for protected wreck sites.

Why the ship sank

Everything in this section is a hypothesis. Even with the wreck identified, no forensic dive survey has reconstructed the failure mode. The hypotheses are set out with the most credible reading given the most weight, and the paranormal framing is included only because the published record contains it, marked strictly as unsupported.

A. Cargo shift in heavy weather causing a fatal list. For: bulk bituminous coal in 1925-era holds was prone to shifting; the distress message reported listing. Against: a pure cargo-shift loss is normally rapid and may not have left the wireless operator the time to compose and send a coherent distress.

B. Hatch-cover failure under heavy seas allowing progressive flooding. For: the JaySea summary of the court file places defective hatch covers in front of Captain Meyers before sailing and reports the replacement covers were left on the dock; the listing reported in the distress is consistent with progressive water ingress through the holds; the timing of the wireless transmission fits a progressive rather than instantaneous loss. Against: a single failure mode cannot be cleanly separated from C without a forensic survey. Strongly supported.

C. Structural failure from overloading or fatigue. For: 1918 emergency wartime construction, seven years of hard coal-trade service, the standing maritime-engineering record of WWI-era hull problems. Against: not directly supported by the surviving documentary record and difficult to disentangle from B.

D. Boiler explosion or catastrophic engine failure. For: coal-fired Scotch boilers of the period are known for failures. Against: a boiler explosion of the magnitude required to sink the ship would normally leave no time for a wireless distress transmission, which conflicts with the reported message. Weakly supported.

E. Bermuda Triangle paranormal hypothesis. For: the case is canonical in the paranormal literature and was further fixed by the 1977 Spielberg cameo. Against: the distress position is at the western edge of, or outside, the canonical Triangle; the documented hatch-cover negligence is sufficient to account for the loss; Barnette has publicly called the framing “total rubbish”; the 2020 identification of the wreck in conventional waters with conventional damage patterns removes any residual geographic mystery. Not supported by the evidentiary record.

F. A combination of B and C: hatch-cover failure plus structural weakness under storm loading. For: this is broadly consistent with Barnette’s public attribution of the loss to “negligence, shoddy maintenance and bad weather” and with the standard maritime-engineering reading of a worn 1918 hull, defective covers, a heavy bituminous load, and a tropical storm. Against: it remains a reconstruction rather than a forensically demonstrated cause. Most defensible reading.

What remains unknown

The precise cause of the Cotopaxi’s loss on or about 1 December 1925 has not been established. Whether the final wireless message was a coded SOS with a transmitted position or a plain non-coded distress, and which the Jacksonville shore station actually received, is contradicted in the secondary literature and stands to the original logs. The identity of the wireless operator who sent the message is not in the record reviewed here. The names of the thirty-two (or thirty) crew should be recovered from contemporary press lists and from the Clinchfield civil action file rather than recited speculatively. Whether a name-bearing artifact survives on the seabed to convert the cumulative 2014 to 2020 identification into an absolute one is not yet answered.