The Lost Colony of Roanoke

The record of the colony is almost entirely John White's own, printed by Richard Hakluyt. It gives the voyage, the landing, a list of the men, women and children who stayed (including two children born on the island and two Algonquian men, Manteo and Towaye, who had been to England), the colonists' formal request that White go home, and his account of 1590⁠[2]⁠[8]. The National Park Service counts the colonists at 117⁠[9].

Three facts from 1590 frame every explanation. First, the colonists had planned before White left to move fifty miles inland, and had agreed with him to carve their destination, with a cross if they were in distress. Second, White found "CRO" on a tree and "CROATOAN" on a palisade post, with no cross, and took it as "a certaine token of their safe being at Croatoan, which is the place where Manteo was borne". Third, the houses had been dismantled and the heavy goods left, which suggests an orderly departure rather than an attack, though chests the colonists had buried had been dug up and plundered⁠[2].

The colony also arrived in a drought. Bald cypress rings from south-east Virginia make 1587–1589 the most extreme drought in 800 years, and 1606–1612, when Jamestown nearly failed, the driest seven years in 770⁠[5].

On Roanoke Island, excavation at Fort Raleigh National Historic Site began in 1895; J. C. Harrington located the small earthwork in 1947, and it was reconstructed in 1950⁠[4]. Ivor Noël Hume's excavations of 1991–1993 reinterpreted a nearby structure as the workshop of the scientist Thomas Harriot and the metallurgist Joachim Gans from the 1585–1586 colony⁠[10]. The 1587 settlement itself has not been found; the Park Service notes that erosion has removed almost half a mile of the island and the site may be under water⁠[4]. In 2024 the First Colony Foundation reported that a copper earring from the Elizabethan Gardens was of European copper, which it ties to the Algonquian village of Roanoac and the 1584–1586 contacts, not to the 1587 colony⁠[1].

On Hatteras Island, East Carolina University's David Phelps found a signet ring in 1998 that was long said to be gold and linked to an Elizabethan family; X-ray fluorescence in 2017 showed it was brass, of a common trade type, and it had been found among artifacts of 1670–1720⁠[11]. Mark Horton's Croatoan Archaeological Project dug near Buxton from 2009 to 2018 with Scott Dawson, who argues the colonists moved to Croatoan⁠[12]. In May 2025 the two announced hammerscale from a blacksmith's forge in a layer whose animal teeth were dated by a laboratory at UC Irvine to the late 1500s, with finds including reworked iron tools, a bronze Tudor rose and a Nuremberg counter; no report or paper has been published⁠[6].

On the mainland, the British Museum revealed in 2012, at the First Colony Foundation's request, a fort symbol under a patch on White's map at the head of Albemarle Sound⁠[3]⁠[13]. The Foundation then dug Site X, at a creek mouth beside the Algonquian town of Mettaquem in Bertie County, and later Site Y upstream; it reports English pottery of possible late 16th-century date, a lead cloth seal from Augsburg, and at Site Y more Elizabethan artifacts and no Native ones⁠[3]⁠[7]. A Museum of London specialist judged the pottery could be late 16th century and unlikely to date after 1650; no sealed 16th-century context or structure has been found, the soil has been ploughed, and later Jamestown-era traders could have left the sherds⁠[7].

To Croatoan. The colonists, or some of them, went to Croatoan (Hatteras Island) as the carved word says, and in time merged with the Croatoan people⁠[2]⁠[12]. It has White's own reading of the sign and Manteo's kinship ties. The archaeological support is contested: the ring is a later brass trade item, the Hatteras finds have been dated by critics to the mid to late 17th century, and the 2025 hammerscale and dates are unpublished⁠[11]⁠[7]⁠[6]. Luccketti doubts many colonists could have moved to a barrier island during a drought⁠[7]. The Foundation's own map study traces Croatoan families taking over the lands the Roanoke people abandoned in 1587⁠[14].

Fifty miles inland. The colonists moved west up Albemarle Sound to the Chowan River, as White said they planned and as the hidden fort symbol suggests, and settled in dispersed farms near friendly towns⁠[2]⁠[3]. Its archaeology has the same weakness: the material at Sites X and Y could be from later traders, and it lacks a sealed, datable 16th-century context⁠[7].

To the Chesapeake. The colony's original destination was Chesapeake Bay, and some historians have argued the main body went north there⁠[2]⁠[12]. Later Jamestown reports of Englishmen killed by the Powhatans have been read in this light; this project has not yet sourced those reports.

Died in place or at sea. The colonists could have died of hunger and disease in the drought, been killed by hostile neighbours, or been lost trying to sail home in their pinnace⁠[5]. The absence of a cross of distress and the orderly removal of the houses argue against a sudden end on Roanoke⁠[2].

Several accounts combine the first two: a small party at Croatoan to watch for ships and the main body inland, which fits both White's statements⁠[2]⁠[3].

A securely dated English domestic context, sealed and undisturbed, at Hatteras or on the Chowan would establish presence; publication of the excavations with full context records, radiocarbon dates with their uncertainties, and artifact catalogues would let others judge the claims already made. Ancient DNA from burials of the right age in either area, compared with English and Algonquian reference genomes, could show whether colonists' descendants lived there. Lead isotopes and XRF on the metal finds could separate 16th-century English material from later trade goods. On Roanoke Island, a survey of the drowned shoreline might find the settlement itself.