Found Under Full Sail: The Carroll A. Deering, 1921

At about eight in the morning on 31 January 1921, a lookout at the Cape Hatteras lifesaving station on the North Carolina coast saw a five-masted schooner hard aground on the outer Diamond Shoals, every sail set, sitting where no ship under command would put herself. Heavy surf kept any boat from reaching her for days. When a wrecking captain finally climbed aboard, around 4 February, he found a ship with no one on her. Both her lifeboats were gone. Her steering gear was smashed. Her log, her papers, her navigation instruments, and the crew’s belongings were missing. Food sat in the pots in the galley. Three cats were aboard. About eleven men had sailed in her, and not one of them, nor a body, nor a boat, was ever found.

That much is documented, and it is where the Carroll A. Deering differs from the more famous ghost ship. The Mary Celeste’s eeriest details, the warm breakfast and the untouched coffee, are fiction laid over a thin record. The Deering’s eeriest details are mostly real. They were recorded by the Coast Guard and a salvage crew, supported by a lightship keeper’s sworn account, and examined by a joint investigation across five federal departments whose file still survives. The trap with this case is not a single great myth. It is two distortions stacked on a solid record: a 1921 message in a bottle that briefly seemed to prove a pirate capture and turned out to be a confessed hoax, and a Bermuda Triangle framing invented in the 1960s and 1970s and fitted backward onto a 1921 event. This is an account of what the record holds, what the evidence can and cannot show, and what is still open once both distortions are pulled away. We keep three things separate, as always: what is documented, what the evidence shows, and what is still only a hypothesis.

The documented account

The Carroll A. Deering was new. She was a five-masted commercial schooner launched at the G.G. Deering Company yard in Bath, Maine, on 4 April 1919, named for the owner’s son, and reported at roughly 255 feet in length, about 1,879 gross tons. She was less than two years old when she was lost. (The dimensions and tonnage come from a secondary specifications table and should be confirmed against a marine registry.)

Her final voyage ran from late 1920 into 1921. She sailed from the U.S. East Coast for Rio de Janeiro carrying a cargo of coal, discharged it at Rio, and began the return leg northbound toward Norfolk, Virginia. On the way home she stopped at Barbados to resupply, leaving around 9 January 1921. The coal was the outbound cargo. On the return leg, when she was lost, she appears to have been in ballast and empty, not carrying coal, a distinction the popular tellings tend to blur.

She did not finish the voyage under her original master. The trip had begun under Captain William H. Merritt, with his son Sewall Merritt as first mate. Merritt fell ill, reportedly at Lewes, Delaware, early in the voyage, and both father and son left the ship. The owners replaced Merritt with a retired master, Captain Willis B. Wormell, reported to have been about sixty-six, and brought in a new first mate, Charles B. McLellan. The crew was about eleven men, described in most accounts as largely Scandinavian and mostly Danish, with an engineer named Herbert Bates. Wormell is reported, through the investigation and family accounts, to have distrusted the replacement crew, with the exception of Bates. That report has not been confirmed against the investigation file.

The last credible contact with the ship came three days before she grounded. On 28 January 1921 the Deering passed the Cape Lookout lightship. The keeper, Captain Jacobson, was hailed through a megaphone by a man he described, in accounts later attributed to him, as tall and thin with reddish hair and a foreign accent, and as someone who neither spoke nor acted like an officer. The man reportedly said the ship had lost her anchors in a storm off Cape Fear. Jacobson is also reported to have noticed the crew gathered on the quarterdeck, an area where ordinary crew were not customarily allowed. He could not relay a warning to the owners because the lightship’s radio was out of service. The exact wording of these phrases is widely repeated but needs primary confirmation against Jacobson’s actual testimony in the investigation file, and is given here as attributed paraphrase rather than as a verified verbatim quote.

Then came the morning of 31 January, the lookout, and the schooner aground on the shoals with all her sails set. A motor lifeboat launched, but the surfmen found her, in the account of the day, fast in a nest of boiling breakers, and could get no closer than about a quarter-mile. Heavy surf prevented boarding for several days. The first man aboard, around 4 February, was Captain James Carlson of the wrecking tug Rescue.

What Carlson and the others found, as recorded in the Coast Guard accounts, the investigation summaries, and Bland Simpson’s scholarly retelling, was this. The sails were set (one account specifies all set except the flying jib). Both lifeboats were gone, with the falls, the lines that lower the boats, trailing over the side. The steering gear was destroyed: the wheel shattered, the binnacle box stove in, the rudder knocked out of use. The ship’s log, the ship’s papers, the chronometer, the navigation instruments, and the crew’s personal effects and clothing were all missing, with some accounts adding that heavy trunks of gear were gone. The ordinary stores, signal flags, and general gear were left untouched. In the galley there was food in the pots on the stove, described as soup, spareribs, and coffee, apparently being prepared for a meal, with unwashed dishes and unmade beds. Three half-starved cats were aboard, and the captain’s cabin was in disarray.

One detail contradicted the lightship hail directly. Both anchors were still aboard when she grounded, the kedge anchor lashed under the bowsprit. This finding is attributed to Shipping Board Commissioner Edward Plummer and should be confirmed against the official record, but if it stands, the man who hailed Jacobson said the ship had lost anchors that she had not in fact lost. The schooner later broke up on the shoals and was eventually demolished as a hazard, with portions of the wreck washing ashore on the Outer Banks.

The disappearance drew an unusually heavy official response, pressed in part by the captain’s family, including his daughter Lula Wormell. It became a joint federal investigation across five cabinet departments, reported as Commerce, Treasury, Justice, Navy, and State. Secretary of Commerce Herbert Hoover took it up through his department, which then included the Bureau of Lighthouses, troubled that other vessels had vanished in the same waters in the same period. He put his assistant Lawrence Richey, a former Secret Service agent, in charge, and the inquiry drew on the Bureau of Investigation, the forerunner of the FBI, and on the Coast Guard.

For a time, in the spring of 1921, the case seemed to break. On 11 April, near Buxton, North Carolina, a local man named Christopher Columbus Gray produced a message in a bottle. As commonly transcribed, it read: “DEERING CAPTURED BY OIL BURNING BOAT SOMETHING LIKE CHASER. TAKING OFF EVERYTHING HANDCUFFING CREW. CREW HIDING ALL OVER SHIP NO CHANCE TO MAKE ESCAPE.” (The exact punctuation and capitalization should be confirmed against the note as preserved in the investigation file.) For a while the note was taken seriously and pushed the inquiry toward piracy. Investigators and family compared the handwriting, including against the hand of the engineer Herbert Bates. By September 1921, Richey had established that the note was forged, and Gray confessed. He had faked it, reportedly, hoping the publicity would help him land a job at the Cape Hatteras light or Coast Guard station. The case’s one apparent piece of direct evidence dissolved.

The investigation reached no formal conclusion. The National Archives summary of the Hoover Library file states that Richey and Hoover concluded a mutiny was the most likely scenario but kept the case open in case more evidence surfaced. No crew member was ever found, alive or dead, and the cause was never established.

The evidence

The strength of the Deering case is its official record, which is rare for a vanishing of this kind. There is a real wreck examined by the Coast Guard and a salvage crew, a contemporaneous lightship keeper’s testimony, and a multi-department federal file preserved in the Lawrence Richey Papers at the Hoover Presidential Library. Its weakness is the same as every vanishing: no survivor, no body, no witness to the moment the crew left, and a key apparent clue that turned out to be a fraud. With that stated, here is what each channel actually shows, and where it stops.

The physical state of the ship. Read together, the set sails, the two missing lifeboats with their falls trailing, and the smashed steering gear describe a ship left under sail by people who departed in or with her boats, amid or after damage to her steering. That is the physical core of the case, and it is solid. The limit is that it shows a crewless ship abandoned by her boats with damaged steering, and nothing more. It does not show why the crew left, who damaged the steering, or whether the departure was voluntary, coerced, or forced.

What was taken against what was left. The crew, or whoever was aboard, took the log, the ship’s papers, the chronometer, the navigation instruments, and personal effects and clothing, while leaving the stores, the signal flags, and the ordinary gear. Taking the navigation gear, the papers, and one’s own belongings is consistent with a deliberate departure by the ship’s own people rather than a panicked flight, and the absence of any plunder of the stores weakens a simple “robbed by pirates” reading. The limit is that the same pattern fits an orderly hijacking as well as a voluntary abandonment. It narrows the field without choosing within it.

The galley food. This is the detail the legend leans on hardest, and it needs care. Unlike the Mary Celeste’s invented warm breakfast, the soup, spareribs, and coffee in the galley pots do appear in the Deering record. But the ship had been aground and pounding in the surf for up to four days before anyone could board, around 4 February. So what the wrecking crew saw was the state of the galley on boarding, not a freeze-frame of the moment the crew left. The popular framing of a meal interrupted mid-bite overstates it badly. The food should not be called warm, fresh, or half-eaten. What the record supports is that food had been in preparation at some point, no more than that, and even the exact foodstuffs should be confirmed against the boarding report.

The lightship testimony. Jacobson’s account is the single most suggestive contemporaneous data point, because it places something abnormal three days before the grounding: the wrong man speaking for the ship, and the crew standing where crew did not customarily stand. But it cannot be taken at face value, and the anchor contradiction is why. The man hailed that the ship had lost her anchors, yet both anchors were aboard when she grounded. Either the hailer lied or was mistaken, or the statement is being over-read. That contradiction is itself a clue, suggesting the man at the rail was not a straight-dealing officer, but it also means the lightship account is not a clean window onto the ship. The exact wording, again, awaits primary confirmation.

The federal file. The Richey Papers are the documentary spine of the case and the reason it is unusually well recorded for a 1921 loss. They are also where the unproven “mutiny most likely” assessment lives. The limit here is one of access rather than substance: the file itself is the prime confirmation target, and the summaries available are secondary.

The bottle note. The one apparent direct clue was fabricated. As evidence of what happened to the Deering, it is worth nothing. Its value is negative and historical: it shows that the piracy-and-capture certainty of 1921 rested substantially on a hoax, and it is the reason the case stayed muddy. A forged note removes a false lead. It proves nothing about the truth.

The theories

Everything in this section is a hypothesis. The investigation itself reached no determination, and none of what follows is proven. The most defensible modern framing is that the credible explanations fall into three families, mutiny, piracy or hijacking, and abandonment in heavy weather, with the cause never established. They are set out best-supported first, and where the record contains historical suspicions of crime against named men, those are included only because the file contains them, marked strictly as suspicions that were raised and never sustained.

Mutiny. This was the investigators’ private leaning, not a finding. On this reading, the replacement crew rose against the elderly Captain Wormell and his mate and then abandoned ship. The supporting points are circumstantial: Wormell’s reported distrust of his crew, the lightship anomaly of a non-officer hailing while the crew stood on the quarterdeck, and the smashed steering. The file also contains a reported remark by first mate McLellan, said to have been overheard by another captain, Hugh Norton, at Barbados, that he would “get the captain before we get to Norfolk.” That remark is second-hand and uncorroborated, and it imputes intent to harm to a named, never-charged, long-dead man. It is recorded here only as an unconfirmed reported statement that fed the mutiny hypothesis, never as fact and never as an assertion that McLellan or anyone else committed a crime. The case against mutiny is that it leaves obvious questions open: if the crew rose, where did they go, and why was the ship left under sail to ground on the shoals rather than sailed off or scuttled? There were no bodies and no confession. The only “confession” the case ever produced, the bottle note, was a hoax. Hoover and Richey leaned this way and could prove nothing.

Piracy or hijacking, including rum-running. Prohibition came into force in 1920, and a five-masted schooner could carry a large liquor cargo, so the idea that the ship was boarded and her crew taken, possibly by rum-runners who valued the hull, had a real motive behind it. One contemporaneous voice, Captain O.W. Parker, insisted that piracy still existed as it had since antiquity. Before it was exposed, the bottle note pointed this way. Against it: the stores and ordinary gear were left untouched, which is strange for plunderers, the note that animated the theory was a fraud, and no pirate vessel, ransom, or surviving crew ever surfaced. Plausible, and unsupported.

Abandonment in or after heavy weather. On this reading, storm damage, the lost anchors as hailed, the smashed steering, trouble off Cape Fear, convinced the crew to abandon a ship they feared was failing, and they were lost at sea in the boats while she drifted under sail onto the shoals. Real winter-Atlantic conditions and the damaged steering support it, and the National Weather Bureau is reported to have favored a storm explanation. Against it: both anchors were in fact still aboard, undercutting the “lost her anchors” premise, and the Deering, like the steamer Hewitt lost in the same period, was reportedly sailing clear of the documented storm tracks, which is exactly why the investigators were not satisfied with weather alone.

Two further threads belong here as context rather than as live contenders. In the Red Scare climate after 1919, the State Department entertained a theory that the Deering’s crew, and the crew of the Hewitt, had been seized by Soviet or Bolshevik agents commandeering American ships, a notion tied to documents allegedly found in a police raid on a radical headquarters in New York. No evidence of any such capture ever materialized, and the conjecture is best understood as a product of the period’s anxieties. The investigation considered it and dropped it. Separately, the loss of the Hewitt in the same waters around the same time struck Hoover as a cluster worth investigating, and it is sometimes raised as a common cause. But a cluster is a reason to look, not an explanation. The Hewitt’s own fate is unresolved, and one vanishing cannot account for another.

That leaves the framing most people now attach to the Deering, the Bermuda Triangle, which is the clearest example of a later legend fitted onto an older event. The Triangle as a concept was popularized decades after 1921, by Vincent Gaddis in 1964 and by Charles Berlitz in his 1974 book The Bermuda Triangle, and the Deering was absorbed into it retroactively. It was never a contemporaneous finding. The Triangle framing was challenged in 1975 by Larry Kusche, an Arizona State University research librarian, whose book The Bermuda Triangle Mystery: Solved argued that the Triangle was a manufactured mystery built on poor research, careless reporting, and occasional deliberate falsification. It is worth being precise about what that debunking does and does not do for the Deering. Kusche’s case was against the Triangle as a supernatural or unified phenomenon, not a solution to the Deering in particular. Rejecting the Triangle does not solve this case. It only removes a label that was never evidence.

What remains unknown

A near-new five-master was found hard aground on Diamond Shoals with every sail set, both boats gone, her steering smashed, her log and papers and instruments and the crew’s belongings taken, food in the galley pots, and three cats aboard. Three days earlier a man who did not act like an officer had hailed a lightship to say she had lost anchors that were still on the ship, while the crew stood where crew did not stand. Five federal departments looked into it. The one apparent clue was a hoax, and the lead investigators privately leaned toward mutiny without being able to prove it.

What the evidence supports is narrow and consistent: the crew left, or were taken, by way of the ship’s boats, and whoever departed took what was needed to navigate and to identify the ship along with their own belongings, which points to a deliberate departure rather than a panicked one. Beyond that, the record does not choose. The credible explanations remain mutiny, piracy, and storm-driven abandonment, and the case for each runs into the same wall: about eleven men, the captain among them, left a sound and brand-new ship under full sail, and none of them was ever seen again. The cause was never established, and the fate of the crew was never learned. The file was left open in 1921, and it has stayed open since.

Found Under Full Sail: The Carroll A. Deering, 1921

The documented account

The evidence

The theories

What remains unknown

Sources

Primary / documentary

Secondary / contextual