When John White returned to Roanoke Island in August 1590, he found the English settlement abandoned. What he described was more specific than a colony that had vanished without a clue: dismantled houses, a defended enclosure, missing boats, damaged possessions and a carved place-name. The surviving account makes relocation a serious explanation, but does not establish where every colonist went or what happened afterwards. Later archaeological finds add context without identifying a complete community or proving one final destination.
Map location: Approximate northern Roanoke Island reference at the modern Fort Raleigh visitor centre, using National Park Service coordinates. This is a public heritage-site reference, not a surveyed location of the 1587 houses, the carved post, a burial site or the colonists’ later destination. The original settlement’s precise location remains uncertain.
A colony already under pressure
The NPS history of the 1587 expedition describes an intended settlement at Chesapeake Bay that instead remained on Roanoke after captain Simon Fernandes refused to carry the colonists farther north. White’s party reused buildings and fortifications associated with the earlier English expedition. This was an attempted colony within an inhabited Indigenous landscape, not an empty island awaiting discovery.
Relations with neighbouring communities were already damaged. After colonist George Howe was killed, White’s men attacked a community they wrongly identified as their opponents; Manteo stopped the attack on friendly Croatoan people. Food insecurity and approaching winter contributed to the decision that White should return to England for supplies. Virginia Dare was born in August 1587. Her later fame should not obscure the other women, men and children whose individual histories are also missing from the surviving English record.
A rescue voyage with its own losses
White’s 1590 voyage account, in a modern-spelling excerpt, records a difficult approach to the coast. Storms repeatedly interrupted progress. Smoke seen near Roanoke on 15 August initially suggested that colonists might still be waiting for him. Another smoke prompted a tiring search ashore that found neither people nor the fresh water the searchers needed. The smoke was a reason for hope, rather than a verified signal from the settlement.
On 17 August, the search suffered a serious accident at a dangerous inlet. White describes one boat taking water and Captain Spicer’s boat being overturned. Seven men drowned, while four survived with assistance from Captain Cooke and others. The sailors were reluctant to continue, but White and Cooke persuaded them. This loss helps explain the expedition’s practical vulnerability. It was not a disappearance of seven colonists and should not be added to the lost settlement’s population.
Nineteen men then set off in two boats. Darkness caused them to overshoot their intended landing. They approached another fire, sounded a trumpet, played familiar English tunes and called out, but received no answer. At daybreak they found burning grass and rotten trees. White records footprints along their route, without establishing who made them. Neither the fire nor those tracks independently identifies surviving English settlers.
The message White expected to recognise
In the account printed by Richard Hakluyt in 1600, White describes first seeing the letters CRO cut into a tree, then CROATOAN carved on a post at the enclosure’s entrance. He understood the markings through an earlier agreement: if the colonists moved, they would leave the name of their destination. A cross above the name was to indicate distress. He reported no such cross.
The important point is that White did not treat CROATOAN as a meaningless code. It was a known place and people, associated in his account with Manteo and friendly relations. He says the colonists had been considering movement inland when he left. Finding a destination-name was therefore consistent with a contingency already discussed, even though the place actually named differed from the inland direction he mentioned.
Absence of the agreed distress sign supported White’s hopeful reading. It cannot establish that everyone survived, that a move occurred all at once, or that conditions remained safe afterwards. A message records an intention or information available when it was made. It does not supply the missing subsequent history of every person. Treating the inscription either as no clue at all or as proof of a fully reconstructed outcome goes beyond the account.
Dismantled buildings and disturbed chests
White found houses taken down and a strong palisade with defensive features. Heavy iron and lead objects remained, partly overgrown, while he could not locate the boats and some weapons he expected. These observations distinguish deliberate dismantling from his description of storm damage elsewhere on the voyage. They also suggest why a picture of untouched meals in suddenly empty houses is a poor substitute for the actual source.
Sailors located buried chests that had been dug up and opened. Three belonged to White. He described damaged books, rain-spoiled pictures and maps, and corroded armour. He blamed people he regarded as enemies on the mainland, imagining that they had watched the colonists depart. That attribution was his inference, not an independently observed raid. The damaged possessions and the identity of whoever disturbed them are different evidential questions.
Why Croatoan was not searched
The National Humanities Center excerpt preserves White’s account of the decision to sail to Croatoan and the difficulties that prevented it. Worsening weather made even recovering water casks unsafe. When the ship tried to leave, a cable broke, an anchor was lost and another failed to hold. With very limited anchoring equipment, dwindling provisions and water left ashore, the company decided to seek supplies elsewhere.
These details explain a central gap: White did not inspect the destination he thought the inscription indicated. He could reasonably believe his family and companions had moved there while still lacking direct confirmation. Failure of this search is not proof that Croatoan was empty, and it does not establish that the colonists reached it. The later mystery partly results from an interrupted investigation rather than an absence of any proposed explanation.
What archaeology can and cannot connect
The NPS account of the 2021 excavations concerns the workshop associated with Joachim Gans and Thomas Harriot and structures from the 1585–1586 military expedition. Finds included evidence of metallurgy, an eight-foot-square watchtower, a connected workshop, changes to a doorway and a drainage ditch. Archaeologists also identified later firepits. Their possible use by the 1587 settlers or people left by Manteo is discussed as an interpretation, not a demonstrated identification of the missing colonists.
The NPS review of major theories distinguishes Croatoan, an inland Albemarle Sound destination and Chesapeake Bay proposals. English objects found at a site need not identify their English owners: trade could carry them into Indigenous communities. The review notes that neither Hatteras Island nor the inland finds have been conclusively linked to a complete lost-colony settlement, its buildings or graves. A suggestive object and an identified community represent different levels of evidence.
Sources and the remaining question
This case uses White’s own narrative through the Hakluyt excerpt and the longer modern-spelling voyage selection, alongside the cited NPS historical and archaeological summaries. The transcriptions were read; no claim is made to have examined White’s original manuscript, every surviving colonial document, excavation notebooks or the artefacts themselves. The 1590 map provides contemporary geographical context, while its modern scan cannot establish the colonists’ final location.
The unresolved question is the subsequent history of the people who left Roanoke, rather than whether White found any indication of departure. His record supplies a named destination, preparations compatible with movement and an unsuccessful attempt to follow. Archaeology keeps several explanations open. None of that requires a supernatural disappearance, and none permits a confident account of every person’s fate. Browse Other Mysteries for further source-led cases.