The Flannan Isles

At about noon on 26 December 1900 the lighthouse tender Hesperus came up to Eilean Mòr, a rock in the Atlantic twenty miles west of Lewis. She was six days late; the weather had held her in harbour. Aboard was Joseph Moore, one of the four keepers of the Flannan Isles light, going back out to start his turn.

Everyone aboard knew what a relief day looked like. A flag on the flagstaff. The empty provision boxes stacked ready on the landing. A keeper coming down the steps to take the ropes.

There was no flag. There were no boxes. Nobody came down. The ship fired a rocket and nothing moved. Captain Jim Harvie’s telegram to Edinburgh that day opens “A dreadful accident has happened at Flannans.” It records what he did when the rocket brought no one out: “as no response was made, managed to land Moore, who went up to the station but found no Keepers there.”

Three men were gone: James Ducat, the principal keeper; Thomas Marshall, the second assistant; and Donald McArthur, an occasional keeper filling in for William Ross, who was on sick leave. None of them was ever found.

What the records actually say

Almost everything people believe about the Flannan Isles was written by people who were never on the island. The documents are shorter, plainer and far more useful.

Moore went up alone. He found the compound gate closed and the main door closed; the kitchen door was open. He looked at the fireplace and “saw that the fire was not lighted for some days.” He went through to where the keepers slept and found “the beds empty just as they left them in the early morning.” Then he stopped: “I did not take time to search further, for I only too well knew that something serious had occurred.”

Men from the ship searched the station and the island with him. There was nobody and no note. In the lightroom, Moore wrote, “The lamp was cleaned. The fountain full. Blinds on the windows.” The morning’s work had been done properly, and then nobody had come back to light the lamp that evening.

Two other things matter. The clocks had stopped, which Harvie read, with the rest, as meaning the accident had happened about a week before. And the kitchen was tidy: “The kitchen utensils were all very clean, which is a sign that it must be after dinner some time they left.” The men had eaten, cleared up, and walked out.

The written logbook stops on 13 December. That is less strange than it sounds — readings were chalked up in rough first and copied into the book afterwards, and a day or two’s lag was normal. The slate held the particulars for the 14th, the hour the lamp was put out on the morning of the 15th, and the barometer, the thermometer and the wind taken at nine o’clock that same morning — the ordinary work of an ordinary morning. That is the last thing any of the three men wrote. There is no storm in it and no distress in it.

The west landing

Superintendent Robert Muirhead reached the island on 29 December. He knew these keepers. He went through the station, then down the cliff to the two landings.

The east landing was in order. The west landing was not.

“The iron railings round the crane platform and from the terminus of the tramway to the concrete steps up from the West landing were displaced and twisted.”

“A large block of stone, weighing upwards of 20 cwt, had been dislodged from its position higher up and carried down to and left on the concrete path.”

Twenty hundredweight is over a ton. Higher up again, in a crevice about a hundred and ten feet above the sea, the keepers kept the box holding the mooring ropes and tackle. It had been washed away, and its ropes were “strewn in the crevices of the rocks near the crane platform.”

A lifebuoy fastened to the railings had gone too, and Muirhead looked closely at how. The ropes that held it had not been touched; scraps of canvas were still clinging to them. The sea had torn the buoy off them “even at this great height (110 feet above sea level).”

The west landing sits at the head of a narrow inlet — a geo. Water running into a channel like that cannot spread sideways, so it piles up and climbs. The box was put up there precisely because the sea was not supposed to reach it.

Back in the station Muirhead went through the outdoor clothing with Moore. On a rock like that one, what a man put on told you where he was going. “Ducat was wearing sea boots and a waterproof, and Marshall sea boots and oilskins.” Those are what you wear to work at a landing, close to the water, in weather. McArthur’s coat was left behind. Moore drew the obvious conclusion: “McArthur has his wearing coat left behind him which shows … that he went out in shirt sleeves.”

Muirhead’s report of 8 January 1901 is carefully worded: he calls it “the most likely explanation”, not the answer. The men had gone down on the afternoon of the 15th to secure the box with the mooring ropes, and “an unexpectedly large roller had come up on the Island, and a large body of water going up higher than where they were and coming down upon them had swept them away with resistless force.”

The story you have probably heard

In the familiar version, the logbook records a storm worse than anything in twenty years. Ducat has gone quiet. McArthur is weeping. The men pray. A last entry says the storm is over, the sea is calm, and God is over all. The relief party walks in to find a meal laid and untouched and a chair knocked over.

None of that is in any record.

The meal and the chair come from a poem: Wilfrid Wilson Gibson’s Flannan Isle, published in 1912, twelve years after the event, by a man writing a ballad who never pretended otherwise. The logbook entries are worse, because they look like evidence. Mike Dash went back to the Board’s papers in the 1990s and published what he found in Fortean Studies in 1998: the entries are not there. Nobody knows who first wrote them. What is certain is that no keeper did.

That matters beyond tidiness. Every strange explanation the case has collected — a sea serpent, a ghost ship, foreign agents, three men driven mad on the cliff edge — needs the storm nobody recorded and the fear nobody wrote down. Take the poem and the invented entries away and they have nothing to stand on.

What the evidence does and does not prove

It does not prove when. Muirhead saw the damage on the 29th, fourteen days after the men vanished, and some of it could have been done by later seas. What it proves is that a sea of that size could and did reach those heights at that landing — which is the thing worth knowing.

It does not prove why the third man went out. The coat proves haste, not its reason. The rule was that the light was never left unattended; one man should have stayed inside. Whatever McArthur saw from the tower or the doorway, he judged it mattered more than the rule, and he did not stop to dress for it. Walter Aldebert, who kept this light from 1953 to 1957, believed one man was washed in and the other two taken by the next sea trying to reach him. That fits everything, and it is still an inference.

And it does not prove a storm. The Flannans face three thousand miles of open Atlantic, and a swell raised by distant weather arrives long after the wind that made it. The sea can be deadly there on a quiet afternoon. The real evidence needs no great storm; only the legend does.

What is left is smaller than the legend and sadder. Two men went down to save a box of rope — ordinary work, done many times. A third ran out after them without his coat. In the Board’s register of lightkeepers someone later wrote the same line in red ink against Ducat and Marshall — “Disappeared on or about 15 December 1900” — and in the register of deaths the cause for all three is given as “probably drowning.” McArthur, being an occasional keeper, was not in the register at all.

The light was relit and is still shining. In 1925 Flannan became one of the first Scottish lights to receive messages from the shore by wireless telegraphy; it was automated on 28 September 1971, and no keeper has lived on Eilean Mòr since. In December 2025 plans were approved to restore the keepers’ old shore station at Breasclete on Lewis, with community and interpretation space, an archive and holiday lets — the first place a visitor could stand and look out toward the Seven Hunters with the documented story, rather than the poem, in front of them.