The Lost Colony of Roanoke

by Fullam, Brandon

  The details contained in the legend regarding the aftereffects of the tidal wave are remarkably consistent the modern NOAA surge models, and of course those models did not exist when Bullard transcribed the legend. The saline contamination of the soil is a particularly interesting and credible detail in the narrative. A storm tide of seven to thirteen feet, as illustrated in the Category 2–4 surge, would have “salted the earth” in a vast coastal area, including present-day Carteret, Pamlico, and Beaufort Counties, as well as other potential mainland settlement locations in the surrounding areas. Croatoan/Hatteras, too, would have been inundated and contaminated. Tidal surges result in increased levels of both sodium and chloride, both of which are toxic to crop plants such as corn, the main food source for both the colonists and the Indians.2 Depending on rainfall and soil factors, salt levels may gradually reduce to tolerable levels within a year’s period of time, but “on soils with high water tables, it may take several years for salt levels to drop to acceptable levels.”3

  As mentioned previously, Indian corn planting was done from April through June so that a corn harvest was available from early summer through October. A September 1589 hurricane and surge would not only have spoiled the fall harvest and the remainder of the summer stores, but also contaminated the soil for at least the next nine months, eliminating the 1590 planting season as well. Given the total impact of a storm tide in 1589, the migration westward “to reach higher land on which they could grow Indian corn” seems plausible.

  Saline floodwaters can also have a delayed effect on drinking water. The U.S. Geological Survey compared the quality of water pumped and distributed to customers by the CHWA (Cape Hatteras Water Association) before and after Hurricane Emily in 1993. Before Emily the chloride content of water distributed to customers was 40 to 45 milligrams per liter. During the three months after Emily the chloride content rose slowly but steadily to a maximum of 280 to 285 milligrams per liter, as the salt water infiltrated the production zones of the wells.4 Croatoan and the flooded coastal mainland may very well have remained barren and uninhabitable for an extended time period after the hurricane of September 1589.

  The legend also tells of an interesting secondary effect of the storm surge and the loss of crops and existing food stores. According to the legend, “Many of the colonists grew sick for lack of bread to eat with sea foods and game which were abundant.” Both the colonists and native Algonquians were dependent on the previously mentioned agricultural cycle. It is entirely conceivable that the survivors of the hurricane and surge, both colonists and native Croatoans alike, would have had little choice but to abandon the sterile areas on the Outer Banks and seek arable land farther west of the flooded coastal mainland. It seems equally credible that a sudden dietary change, presumably from mostly grain-based to exclusively shellfish and wild game, would have had a negative reaction among the colonists. As noted in the previous chapter, death can occur among hurricane surge survivors from the effects of injury, illness, and malnutrition from inadequate or spoiled food.

  The legend claims that “the remnant of the Lost Colony,” accompanied by Manteo and other Croatoan survivors, migrated west from Croatoan “about the end of the year 1588 or early in the year 1589.” This, though, would place the occurrence of the tidal wave in 1588 rather than 1589. There were three hurricanes cited in the previous chapter which could have impacted the colony between September of 1588 and September of 1589. One of these was the powerful hurricane that struck Cuba in September 1588. As mentioned, there were few details relating to this event in the Spanish archives, but it very well could have followed the same track as the hurricane in 1589, in which case the colony may have been struck by two consecutive hurricanes and surges, first in September of 1588 and then in September of 1589. The Atlantic hurricane season technically runs from June 1 until November 30, but on the Outer Banks the majority of hurricanes or tropical cyclones have historically arrived in late August or September.5 If the 1588 hurricane recorded briefly in the Archivo General actually struck present-day North Carolina, it would place the collapse of the colony only six months after the colonists left Roanoke and established their new settlement on the mainland.

  On the other hand it is evident that the legend, in its present transcribed form, contains a number of historical details and references which could not have been part of the original oral tradition. These historical details are associated with the general narrative, but clearly were added either by Ernest Bullard, as historian, or perhaps by someone earlier in order to augment the legend’s narrative. One of these additions is probably the name of George Howe, Sr., along with the exact date of his death at the hands of hostile Indians at Roanoke on July 28, 1587. Another is the mention of his son, “young George Howe,” and of course the names of Carteret and Pamlico Counties. All of these are historically accurate details, the Howe references coming directly from White’s account of the 1587 voyage, but these would have been later attachments, inserted into what could have been only a very general oral tradition if it had survived for centuries. It is likewise implausible that the precise “end of the year 1588 or early in the year 1589” reference could have been passed down over the centuries as part of the oral tradition.

  It is possible that the oral tradition could have referenced the effects of a great storm that occurred sometime after the colonists resettled, and that Bullard or someone earlier added that specific timeframe to the transcription as a logical estimate. It should be noted once again that the 1588 and 1589 hurricanes were not part of the historical record until their recent discovery in the Spanish archives. When Bullard first heard the legend in 1892 and transcribed it in 1950, there was no knowledge whatsoever of an Atlantic hurricane in either 1588 or 1589, and it is the unique reference to a hurricane or “tidal wave” that makes the legend intriguing. In the final analysis, whether the storm struck in 1588 or 1589, or whether storms struck in September of both years is largely inconsequential, since the outcome would have been the same. The mainland settlement would have been wiped away, the colony would have ceased to exist, and whatever survivors there were would have been forced farther inland in search of arable land.

  The legend goes on to assert that the combined group of tidal wave survivors—including Manteo, some of his native Croatoans, and some colonists—attempted to ascend the Neuse River in search of unspoiled land to grow corn. As cited above as one of the examples of Bullard’s augmentations, he added that “what is now Carteret or Pamlico County … [was] where they first settled.” Carteret County was also mentioned in connection with the Lost Colony by Hamilton McMillan (1837–1916), North Carolina assemblyman and advocate for the Lumbee tribe in Robeson County. McMillan had recorded an oral tradition among the Lumbees, which—like the Bullard legend but with far less detail—claimed a direct ancestral connection to the Hatteras/Croatoan tribe and the Lost Colonists. McMillan wrote, “After the English colony became incorporated with the [Hatteras] tribe, they began to emigrate westward…. The line of emigration extended westward from what is now Carteret County.”6

  The Bullard and Hamilton reference to either a temporary settlement or emigration point in present-day Carteret County supports the suggestion made earlier that the Croatoans, and by association the Lost Colonists, could have interacted or at least had contact with the Coree tribe. As mentioned, at the time of the Roanoke voyages the Coree/Coranine occupied the peninsula and coastal areas south of the Neuse River in present-day Carteret County. This would in turn support the belief that at least part of the tribe which eventually became known as the Coharie had ancestral links to both the Coree as well as the Neusiok, who dwelt on the south side of the lower Neuse River in present-day Carteret and Craven counties. The actual “Coharie” name did not come into existence until about a century later.

  The Coharie-Coree connection could also explain a potential linguistic question not addressed in the “Legend of the Coharie” or elsewhere, namely that of the language affiliation of the Indians in
this Lost Colony/Indian group. Manteo and his small tribe from Croatoan, the only Indians mentioned in the legend as accompanying the group of Lost Colonists traveling inland, were Algonquians. James Mooney proposed that, since the Neusiok were allied with the Algonquian Pomouik (Pamlico) in 1584, the Neusiok may also have been Algonquian,7 but Mooney’s conclusion is generally disputed. Most ethnologists now hold that the Neusioks (and possibly but not definitely the Coree), from which the present-day Coharie tribal council officially claims descent,8 were Iroquoian. Consequently, for the Coharie tribe to be linear descendants of colonists and Croatoans, as the legend claims, and also descendants of the Neusioks, as the tribal council claims, there had to have been an infusion of Iroquoian blood at some point. Present-day Carteret County would have been a logical place for that admixture to occur.

  It is not stated how far the group traveled up the Neuse River, but according to the legend, “Before they could get settled on a desirable location they were attacked by an unfriendly tribe and a few of the migrants were wounded.” This part of the narrative is also historically consistent, at least with what is known of the late 16th century tribes. The “unfriendly tribe” would almost certainly have been the Tuscarora, known to the English as Mongoaks, who occupied the upper Neuse at that time and were frequently at war with the tribes closer to the coast.

  The exact origin of the “Legend of the Coharie” is unclear. Ernest Bullard wrote that it was first told to him about 1892, and added that it had been passed down “by word of mouth for more than three hundred years before one word of it was ever put into writing.” The legend is not widely known today outside the Bullard extended family and Sampson County, but it is probable that it was passed down orally prior to 1892 in the Hall family line. Ernest Bullard wrote, “on top of a knoll overlooking the lowland of Big Swamp in the western part of what was known at the time (1779) as ‘The Territory’ of Duplin County [the western part of Duplin County became Sampson County in 1784] there stood a small log cabin belonging to Enoch Hall. According to the legend Hall was “said to have been a lineal descendant from George Howe of the ‘Lost Colony,’ the name having been changed from Howe to Haw, then to Hall.” George Howe, as mentioned, was killed shortly after White and the colonists arrived at Roanoke in July of 1587. His son, George Howe, Jr., was listed among the colonists and—according to the legend—is purported to be the ancestral link to Enoch Hall and the Hall descendants in Sampson County centuries later.

  The legend, therefore, seems to have been passed down in the Hall family line at least until 1892, when Ernest Bullard first heard it. There is a genealogical connection between the Bullards and the Halls which might explain Bullard’s access to the legend. W. Stephen Lee, grandson of Ernest Bullard, reports that one of Ernest Bullard’s great aunts, Lucy Bullard Hall, was married to Everett Hall, son of Enoch Hall, who himself was the son of another Enoch Hall.9 It has always been common practice to recount old stories and traditions when extended families occasionally gather from near and far for such events as wakes and weddings. Lucy Bullard Hall died in 1892, the same year the legend was related to Ernest Bullard. It is possible that young Ernest Bullard first heard the legend from one of his Hall relatives when the Bullard and Hall families assembled for Lucy’s funeral.

  Ernest Bullard was a member of the Sampson County Historical Society, an indication of his interest in local history. W. Stephen Lee believes his grandfather transcribed the “Legend of the Coharie” after his retirement, and that it was published at that time by the historical society. The story was reprinted in “Pitch ’n’ Tar,” a series of publications for a high school oral history project conducted by Matilda West and her students in Roseboro in the late 1960s or early 1970s.10 Most recently the legend was reprinted in the January 2014 issue of “Huckleberry Historian,” the quarterly journal of the Sampson County Historical Society.

  No discussion of the “Legend of the Coharie” would be complete without further mention of Hamilton McMillan and the Lumbee tribe. In 1892, when Ernest Bullard first heard the Coharie Legend, a great deal of attention was being paid to the above-referenced booklet which had been published by Hamilton McMillan just four years earlier. McMillan had recorded the Lumbee oral tradition which—like Bullard’s legend—claimed a direct ancestral connection to the Hatteras/Croatoan tribe and the Lost Colonists. McMillan wrote that he first learned of the Lumbee connection to the Lost Colony in 1864, when he heard an old Indian named George Lowrie state that his tribal ancestors included English colonists from Roanoke. McMillan’s subsequent investigations and interviews in Robeson confirmed that there was a tradition connecting some portion of the Lumbee tribe and the 1587 colony at Roanoke. McMillan’s arguments, in fact, led directly to North Carolina’s recognition of the Lumbee tribe, initially called “Croatan Indians.”

  That same oral tradition is found among the closely related Coharie tribe in Sampson County. Coharie tribal member Enoch Emanuel wrote the following about 1913 when he was approximately 70 years old:

  The mixed race of people living in Sampson County are sure that the statements given to us by our ancestors concerning our origin are true. We have only asked for Indian prestige, while we know in our veins also flows the blood of our white ancestors. We have always been told by our fathers and mothers that we were mixed with the lost colony of the Roanoke. We therefore are a mixture of Governor White’s colony and the original Indians.11

  More than a half century later another analysis of Coharie traditions was done by Don A. Grady for his 1981 master’s thesis on the oral folk history of the Coharie tribe. Grady interviewed ten descendants of the tribe and wrote, “Perhaps the most important tradition is that of the tribe’s origin. They believe very strongly that they are the descendants of a coastal tribe of Indians and the Lost Colony.”12

  After McMillan’s success with the Lumbee tribe, efforts were undertaken in Sampson County to do for the Coharie what had been done for the Lumbee in Robeson. In 1914 Special Indian Agent O.M. McPherson was directed to investigate the Indians of Robeson and the surrounding counties, and in early 1915 he presented his “Report on the Condition and Tribal Rights of the Indians of Robeson and Adjoining Counties.”13 In his report McPherson included 213 Sampson County Indians as part of the Croatan tribe (which by 1913 was called “Cherokee Indians of Robeson County”). McPherson’s report relied heavily on the previous work done by McMillan. In 1916 Coharie advocate George Edwin Butler (1868–1941) used the same arguments in his successful appeal for separate schools for the Coharie in Sampson County.14

  It might seem from all of this that Ernest Bullard, like his contemporaries McPherson and Butler, may have been directly influenced by McMillan’s earlier work, and that the “Legend of the Coharie” could essentially be an adaptation of McMillan’s 1888 publication. That does not seem to be the case. Other than its thematic link to the Lost Colony, Bullard’s legend bears virtually no resemblance whatsoever to the narrative in McMillan’s publication.

  In the first place the Bullard Legend begins with the tidal wave and its aftereffects, events which were unknown to McMillan, McPherson, Butler, or anyone else. The legend also provides a far more detailed and coherent chronology of events from the occurrence of the tidal wave to about 1690–1700, when the descendants of the mixed Colonist-Indian group supposedly reached Sampson County, as will be seen later. The Lumbee oral tradition, as recorded by McMillan, recalls a direct connection to the Croatoan [Hatteras] tribe and the Lost Colony, but provides virtually no information on the century that transpired afterwards, when the Lumbees were said to have eventually settled in Robeson County.

  Secondly, the “Legend of the Coharie” consists essentially of two separate, interrelated narratives. The first is the aforementioned chronological and geographical account about survivors of John White’s English colony and a group of Croatoan Indians. Woven into that broader narrative, however, is the male lineage of two specific individuals: George Howe, Jr., son of the English colonist kil
led at Roanoke on July 28, 1587, and an unnamed daughter of Manteo, leader of the tribe at Croatoan. McMillan’s version, as stated, is a much more general account of Lumbee origins, reconstructed from interviews he conducted long after the tribe had developed into the heterogeneous group that it remains today. By McMillan’s time the “Lumbees” were already comprised of a blend of tribal elements and traditions including Hatteras, Cherokee, Cheraw, Tuscarora, Mattamuskeet, and perhaps others. McMillan offers no explicit narrative whatsoever about specific individuals or events between 1587 and 1711.

  In addition, from what can be determined it seems that the “Legend of the Coharie” was passed down along Enoch Hall’s ancestral bloodline and was not known outside the Hall family until it was finally recorded by Ernest Bullard, “…as ’twas told to me.” This may account for the fact that McMillan, McPherson, and Butler were apparently unaware of the Howe-Hall oral family history. As mentioned, McMillan’s investigations were focused on Robeson County and the Lumbee tribe. Both McPherson and Butler borrowed liberally from McMillan and, since McMillan’s arguments had been successful, possibly did not see the need to explore alternate oral histories in Sampson County.

  Thirdly, it should be noted that, unlike the Bullard Legend, the McMillan, McPherson, and Butler accounts all had a political agenda, i.e., tribal recognition of the Lumbee and the Coharie by the state of North Carolina. These unclassified admixed “Indians” in Robeson, Sampson, and surrounding counties had been designated “free persons of color” in 1835, and were thereby prevented from attending schools for whites, a prohibition they would resent until McMillan entered the scene. McMillan spent a number of years compiling enough evidence to persuade the state that the Lumbee ancestry could be traced back to the 1587 colony at Roanoke and the Croatoan (Hatteras) tribe. His arguments were aimed at—and directly brought about—North Carolina’s official recognition of the “Croatan” tribe and its right to a separate school system. The Bullard Legend is a family oral history and had no such political agenda.