Four later texts are still cited as evidence of survivors. They are of unequal weight and only one of them is independent of everything above.
Bland, 1650. Edward Bland, a Virginia merchant, led a party in 1650 towards the Tuscarora, the Iroquoian-speaking people of the interior of what is now North Carolina, carrying orders from the governor of Virginia to speak with "an Englishman amongst them" and "to enquire for an English woman cast away long since, and was amongst those Nations" (The Discovery of New Brittaine, 1651, p. 3). They wrote to the Englishman in five languages, were told he was "a great way off at the further Tuskarood Towne", and never met him (pp. 5 to 6, 13). A woman "cast away" among the Tuscarora sixty-three years after 1587 is as likely a shipwreck survivor as a colonist. The passage shows only that Virginia still expected to find English among the southern nations in 1650.
Yeardley, 1654. In September 1653 four traders reached Roanoke Island, where "the great commander of those parts" received them "and shewed them the ruins of Sir Walter Ralegh's fort" (Francis Yeardley, a Virginia planter and councillor, to John Ferrar in England, formerly of the Virginia Company, 8 May 1654; Salley, Narratives of Early Carolina, pp. 25 to 26). The chief asked for his son to be taught "to speak out of the book". The letter records no tradition of the colonists; so far as it tells, the island's people in 1653 knew the fort only as a ruin.
Lawson, 1709. John Lawson, an English surveyor who became surveyor-general of North Carolina, wrote from his travels and residence in Carolina from 1700:
"A farther Confirmation of this we have from the Hatteras Indians, who either then lived on Ronoak-Island, or much frequented it. These tell us, that several of their Ancestors were white People, and could talk in a Book, as we do; the Truth of which is confirm'd by gray Eyes being found frequently amongst these Indians, and no others. They value themselves extremely for their Affinity to the English, and are ready to do them all friendly Offices. It is probable, that this Settlement miscarry'd for want of timely Supplies from England; or thro' the Treachery of the Natives, for we may reasonably suppose that the English were forced to cohabit with them, for Relief and Conversation; and that in process of Time, they conform'd themselves to the Manners of their Indian Relations."
Lawson, A New Voyage to Carolina, 1709, p. 62. His census on p. 234 gives "Hatteras. Town 1, Sand Banks, Fighting Men 16."
This is the one later testimony that comes from the Indians' side, from the people of the island the colonists named, and it is independent of the Jamestown tradition, which was in manuscript or in Purchas's margins where Lawson never saw it. It is also 120 years after the event, about a community of sixteen men, on a coast that had seen wrecks, traders and Lane's own men. Lawson's explanation, supplies failed and the English "were forced to cohabit with them", is the first printed statement of assimilation, and it is his inference. Robert Beverley, a Virginia planter, in his history of that colony four years earlier, gave the Virginian view: the colonists "were remov'd to Croatan" and the Indians "cut them off: For to this Day they were never more heard of" (History and Present State of Virginia, 1705, pp. 10 to 11). Lawson's Carolina tradition and Beverley's Virginian one both put the colonists at Croatoan; they differ on whether the colonists were absorbed or killed.
McMillan, 1888. Hamilton McMillan, a North Carolina state legislator who had the Indians of Robeson County, in the south of the state, recognised as "Croatan Indians" in 1885 (they have been the Lumbee since 1953), published their traditions three years later: a tribe that "once lived in Roanoke in Virginia", "took the English to live with us", moved inland, and bore colonists' surnames (Sir Walter Raleigh's Lost Colony, 1888, pp. 16 to 24). His evidence is a speech of 1864 reported "in substance" twenty-four years later; traditions gathered from 1875 in a community he had already named Croatan; surnames that are common English names; and a pronunciation "Darr, Durr and Dorr" elicited after he had told the story of Virginia Dare (p. 18). The Lumbee historian Malinda Maynor Lowery (2019) and the anthropologist Robert K. Thomas, in a report of 1976 for the Lumbee themselves, reject the surname argument. McMillan builds openly on Lawson and adds nothing that can be checked. Alexander Brown's endorsement in 1890 put the Lumbee link into the scholarly literature, in the same volume as his transcription of the Zúñiga chart, and it has been cited from there since.