1525. As these voyages were not followed up by permanent occupation

of the country it is doubtful if they made any lasting impression upon Indian tradition. The author has chosen to assume, with Brinton and Rafinesque, that the Walam Olum reference is to the settlement of the Dutch at New York and the English in Virginia soon after 1600. (8) De Soto's route (p. 26): On May 30, 1539, Hernando de Soto, of Spain, with 600 armed men and 213 horses, landed at Tampa bay, on the west coast of Florida, in search of gold. After more than four years of hardship and disappointed wandering from Florida to the great plains of the West and back again to the Mississippi, where De Soto died and his body was consigned to the great river, 311 men, all that were left of the expedition, arrived finally at Pánuco, in Mexico, on September 10, 1543. For the history of this expedition, the most important ever undertaken by Spain within eastern United States, we have four original authorities. First is the very brief, but evidently truthful (Spanish) report of Biedma, an officer of the expedition, presented to the King in 1544, immediately after the return to Spain. Next in order, but of first importance for detail and general appearance of reliability, is the narrative of an anonymous Portuguese cavalier of the expedition, commonly known as the Gentleman of Elvas, originally published in the Portuguese language in 1557. Next comes the (Spanish) narrative of Garcilaso, written, but not published, in 1587. Unlike the others, the author was not an eyewitness of what he describes, but made up his account chiefly from the oral recollections of an old soldier of the expedition more than forty years after the event, this information being supplemented from papers written by two other soldiers of De Soto. As might be expected, the Garcilaso narrative, although written in flowery style, abounds in exaggeration and trivial incident, and compares unfavorably with the other accounts, while probably giving more of the minor happenings. The fourth original account is an unfinished (Spanish) report by Ranjel, secretary of the expedition, written soon after reaching Mexico, and afterward incorporated with considerable change by Oviedo, in his "Historia natural y general de las Indias." As this fourth narrative remained unpublished until 1851 and has never been translated, it has hitherto been entirely overlooked by the commentators, excepting Winsor, who notes it incidentally. In general it agrees well with the Elvas narrative and throws valuable light upon the history of the expedition. The principal authorities, while preserving a general unity of narrative, differ greatly in detail, especially in estimates of numbers and distances, frequently to such an extent that it is useless to attempt to reconcile their different statements. In general the Gentleman of Elvas is most moderate in his expression, while Biedma takes a middle ground and Garcilaso exaggerates greatly. Thus the first named gives De Soto 600 men, Biedma makes the number 620, while Garcilaso says 1,000. At a certain stage of the journey the Portuguese Gentleman gives De Soto 700 Indians as escort, Biedma says 800, while Garcilaso makes it 8,000. At the battle of Mavilla the Elvas account gives 18 Spaniards and 2,500 Indians killed, Biedma says 20 Spaniards killed, without giving an estimate of the Indians, while Garcilaso has 82 Spaniards and over 11,000 Indians killed. In distances there is as great discrepancy. Thus Biedma makes the distance from Guaxule to Chiaha four days, Garcilaso has it six days, and Elvas seven days. As to the length of an average day's march we find it estimated all the way from "four leagues, more or less" (Garcilaso) to "every day seven or eight leagues" (Elvas). In another place the Elvas chronicler states that they usually made five or six leagues a day through inhabited territories, but that in crossing uninhabited regions--as that between Canasagua and Chiaha, they marched every day as far as possible for fear of running out of provisions. One of the most glaring discrepancies appears in regard to the distance between Chiaha and Coste. Both the Portuguese writer and Garcilaso put Chiaha upon an island--a statement which in itself is at variance with any present conditions,--but while the former makes the island a fraction over a league in length the latter says that it was five leagues long. The next town was Coste, which Garcilaso puts immediately at the lower end of the same island while the Portuguese Gentleman represents it as seven days distant, although he himself has given the island the shorter length. Notwithstanding a deceptive appearance of exactness, especially in the Elvas and Ranjel narratives, which have the form of a daily journal, the conclusion is irresistible that much of the record was made after dates had been forgotten, and the sequence of events had become confused. Considering all the difficulties, dangers, and uncertainties that constantly beset the expedition, it would be too much to expect the regularity of a ledger, and it is more probable that the entries were made, not from day to day, but at irregular intervals as opportunity presented at the several resting places. The story must be interpreted in the light of our later knowledge of the geography and ethnology of the country traversed. Each of the three principal narratives has passed through translations and later editions of more or less doubtful fidelity to the original, the English edition in some cases being itself a translation from an earlier French or Dutch translation. English speaking historians of the expedition have usually drawn their material from one or the other of these translations, without knowledge of the original language, of the etymologies of the Indian names or the relations of the various tribes mentioned, or of the general system of Indian geographic nomenclature. One of the greatest errors has been the attempt to give in every case a fixed local habitation to a name which in some instances is not a proper name at all, and in others is merely a descriptive term or a duplicate name occurring at several places in the same tribal territory. Thus Tali is simply the Creek word talua, town, and not a definite place name as represented by a mistake natural in dealing through interpreters with an unknown Indian language. Tallise and Tallimuchase are respectively "Old town" and "New town" in Creek, and there can be no certainty that the same names were applied to the same places a century later. Canasagua is a corruption of a Cherokee name which occurs in at least three other places in the old Cherokee country in addition to the one mentioned in the narrative, and almost every old Indian local name was thus repeated several times, as in the case of such common names as Short creek, Whitewater, Richmond, or Lexington among ourselves. The fact that only one name of the set has been retained on the map does not prove its identity with the town of the old chronicle. Again such loose terms as "a large river," "a beautiful valley," have been assumed to mean something more definitely localized than the wording warrants. The most common error in translation has been the rendering of the Spanish "despoblado" as "desert." There are no deserts in the Gulf states, and the word means simply an uninhabited region, usually the debatable strip between two tribes. There have been many attempts to trace De Soto's route. As nearly every historian who has written of the southern states has given attention to this subject it is unnecessary to enumerate them all. Of some thirty works consulted by the author, in addition to the original narratives already mentioned, not more than two or three can be considered as speaking with any authority, the rest simply copying from these without investigation. The first attempt to locate the route definitely was made by Meek (Romantic Passages, etc.) in 1839 (reprinted in 1857), his conclusions being based upon his general knowledge of the geography of the region. In 1851 Pickett tried to locate the route, chiefly, he asserts, from Indian tradition as related by mixed-bloods. How much dependence can be placed upon Indian tradition as thus interpreted three centuries after the event it is unnecessary to say. Both these writers have brought De Soto down the Coosa river, in which they have been followed without investigation by Irving, Shea and others, but none of these was aware of the existence of a Suwali tribe, or correctly acquainted with the Indian nomenclature of the upper country, or of the Creek country as so well summarized by Gatschet in his Creek Migration Legend. They are also mistaken in assuming that only De Soto passed through the country, whereas we now know that several Spanish explorers and numerous French adventurers traversed the same territory, the latest expeditions of course being freshest in Indian memory. Jones in his "De Soto's March Through Georgia" simply dresses up the earlier statements in more literary style, sometimes changing surmises to positive assertions, without mentioning his authorities. Maps of the supposed route, all bringing De Soto down the Coosa instead of the Chattahoochee, have been published in Irving's Conquest of Florida, the Hakluyt Society's edition of the Gentleman of Elva's account, and in Buckingham Smith's translation of the same narrative, as well as in several other works. For the eastern portion, with which we have to deal, all of these are practically duplicates of one another. On several old Spanish and French maps the names mentioned in the narrative seem to have been set down merely to fill space, without much reference to the text of the chronicle. For a list and notices of principal writers who have touched upon this subject see the appendix to Shea's chapter on "Ancient Florida" in Winsor's Narrative and Critical History of America, II; Boston, 1886. We shall speak only of that part of the route which lay near the Cherokee mountains. The first location which concerns us in the narrative is Cofitachiqui, the town from which De Soto set out for the Cherokee country. The name appears variously as Cofitachequi (Ranjel), Cofitachique (Biedma), Cofachiqui (Garcilaso), Cutifa-Chiqui (by transposition, Elvas), Cofetaçque (Vandera), Catafachique (Williams) and Cosatachiqui (misprint, Brooks MSS), and the Spaniards first heard of the region as Yupaha from a tribe farther to the south. The correct form appears to be that first given, which Gatschet, from later information than that quoted in his Creek Migration Legend, makes a Hitchitee word about equivalent to "Dogwood town," from cofi, "dogwood," cofita, "dogwood thicket," and chiki, "house," or collectively "town." McCulloch puts the town upon the headwaters of the Ocmulgee; Williams locates it on the Chattahoochee; Gallatin on the Oconee or the Savannah; Meek and Monette, following him, probably in the fork of the Savannah and the Broad; Pickett, with Jones and others following him, at Silver bluff on the east (north) bank of the Savannah, in Barnwell county, South Carolina, about 25 miles by water below the present Augusta. It will thus be seen that at the very outset of our inquiry the commentators differ by a distance equal to more than half the width of the state of Georgia. It will suffice here to say, without going into the argument, that the author is inclined to believe that the Indian town was on or near Silver bluff, which was noted for its extensive ancient remains as far back as Bartram's time (Travels, 313), and where the noted George Galphin established a trading post in 1736. The original site has since been almost entirely worn away by the river. According to the Indians of Cofitachiqui, the town, which was on the farther (north) bank of the stream, was two day's journey from the sea, probably by canoe, and the sailors with the expedition believed the river to be the same one that entered at St. Helena, which was a very close guess. The Spaniards were shown here European articles which they were told had been obtained from white men who had entered the river's mouth many years before. These they conjectured to have been the men with Ayllon, who had landed on that coast in 1520 and again in 1524. The town was probably the ancient capital of the Uchee Indians, who, before their absorption by the Creeks, held or claimed most of the territory on both banks of Savannah river from the Cherokee border to within about forty miles of Savannah and westward to the Ogeechee and Cannouchee rivers (see Gatschet, Creek Migration Legend, I, 17-24). The country was already on the decline in 1540 from a recent fatal epidemic, but was yet populous and wealthy, and was ruled by a woman chief whose authority extended for a considerable distance. The town was visited also by Pardo in 1567 and again by Torres in 1628, when it was still a principal settlement, as rich in pearls as in De Soto's time (Brooks MSS, in the archives of the Bureau of American Ethnology). Somewhere in southern Georgia De Soto had been told of a rich province called Coça (Coosa, the Creek country) toward the northwest. At Cofitachiqui he again heard of it and of one of its principal towns called Chiaha (Chehaw) as being twelve days inland. Although on first hearing of it he had kept on in the other direction in order to reach Cofitachiqui, he now determined to go there, and made the queen a prisoner to compel her to accompany him a part of the way as guide. Coça province was, though he did not know it, almost due west, and he was in haste to reach it in order to obtain corn, as his men and horses were almost worn out from hunger. It is apparent, however, that the unwilling queen, afraid of being carried beyond her own territories, led the Spaniards by a roundabout route in the hope of making her escape, as she finally did, or perhaps of leaving them to starve and die in the mountains, precisely the trick attempted by the Indians upon another Spanish adventurer, Coronado, entering the great plains from the Pacific coast in search of golden treasure in the same year. Instead therefore of recrossing the river to the westward, the Spaniards, guided by the captive queen, took the direction of the north ("la vuelta del norte"--Biedma), and, after passing through several towns subject to the queen, came in seven days to "the province of Chalaque" (Elvas). Elvas, Garcilaso, and Ranjel agree upon the spelling, but the last named makes the distance only two days from Cofitachiqui. Biedma does not mention the country at all. The trifling difference in statement of five days in seven need not trouble us, as Biedma makes the whole distance from Cofitachiqui to Xuala eight days, and from Guaxule to Chiaha four days, where Elvas makes it, respectively, twelve and seven days. Chalaque is, of course, Cherokee, as all writers agree, and De Soto was now probably on the waters of Keowee river, the eastern head stream of Savannah river, where the Lower Cherokee had their towns. Finding the country bare of corn, he made no stay. Proceeding six days farther they came next to Guaquili, where they were kindly received. This name occurs only in the Ranjel narrative, the other three being entirely silent in regard to such a halting place. The name has a Cherokee sound (Wakili), but if we allow for a dialectic substitution of l for r it may be connected with such Catawba names as Congaree, Wateree, and Sugeree. It was probably a village of minor importance. They came next to the province of Xuala, or Xualla, as the Elvas narrative more often has it. In a French edition it appears as Chouala. Ranjel makes it three days from Guaquili or five from Chalaque. Elvas also makes it five days from Chalaque, while Biedma makes it eight days from Cofitachiqui, a total discrepancy of four days from the last-named place. Biedma describes it as a rough mountain country, thinly populated, but with a few Indian houses, and thinks that in these mountains the great river of Espiritu Santo (the Mississippi) had its birth. Ranjel describes the town as situated in a plain in the vicinity of rivers and in a country with greater appearance of gold mines than any they had yet seen. The Portuguese gentleman describes it as having very little corn, and says that they reached it from Cofitachiqui over a hilly country. In his final chapter he states that the course from Cofitachiqui to this place was from south to north, thus agreeing with Biedma. According to Garcilaso (pp. 136-137) it was fifty leagues by the road along which the Spaniards had come from Cofitachiqui to the first valley of the province of Xuala, with but few mountains on the way, and the town itself was situated close under a mountain ("a la falda de una sierra") beside a small but rapid stream which formed the boundary of the territory of Cofitachiqui in this direction. From Ranjel we learn that on the same day after leaving this place for the next "province" the Spaniards crossed a very high mountain ridge ("una sierra muy alta"). Without mentioning the name, Pickett (1851) refers to Xuala as "a town in the present Habersham county, Georgia," but gives no reason for this opinion. Rye and Irving, of the same date, arguing from a slight similarity of name, think it may have been on the site of a former Cherokee town, Qualatchee, on the head of Chattahoochee river in Georgia. The resemblance, however, is rather farfetched, and moreover this same name is found on Keowee river in South Carolina. Jones (De Soto in Georgia, 1880) interprets Garcilaso's description to refer to "Nacoochee valley, Habersham county"--which should be White county--and the neighboring Mount Yonah, overlooking the fact that the same description of mountain, valley, and swift flowing stream might apply equally well to any one of twenty other localities in this southern mountain country. With direct contradiction Garcilaso says that the Spaniards rested here fifteen days because they found provisions plentiful, while the Portuguese Gentleman says that they stopped but two days because they found so little corn! Ranjel makes them stop four days and says they found abundant provisions and assistance. However that may have been, there can be no question of the identity of the name. As the province of Chalaque is the country of the Cherokee, so the province of Xuala is the territory of the Suwali or Sara Indians, better known later as Cheraw, who lived in early times in the piedmont country about the head of Broad river in North Carolina, adjoining the Cherokee, who still remember them under the name of Ani'-Suwa'li. A principal trail to their country from the west led up Swannanoa river and across the gap which, for this reason, was known to the Cherokee as Suwa'li-nuñnâ, "Suwali trail," corrupted by the whites to Swannanoa. Lederer, who found them in the same general region in 1670, calls this gap the "Suala pass" and the neighboring mountains the Sara mountains, "which," he says, "The Spaniards make Suala." They afterward shifted to the north and finally returned and were incorporated with the Catawba (see Mooney, Siouan Tribes of the East, bulletin of the Bureau of Ethnology, 1894). Up to this point the Spaniards had followed a north course from Cofitachiqui (Biedma and Elvas), but they now turned to the west (Elvas, final chapter). On the same day on which they left Xuala they crossed "a very high mountain ridge," and descended the next day to a wide meadow bottom ("savana"), through which flowed a river which they concluded was a part of the Espiritu Santo, the Mississippi (Ranjel). Biedma speaks of crossing a mountain country and mentions the river, which he also says they thought to be a tributary of the Mississippi. Garcilaso says that this portion of their route was through a mountain country without inhabitants ("despoblado") and the Portuguese gentleman describes it as being over "very rough and high ridges." In five days of such travel--for here, for a wonder, all the narratives agree--they came to Guaxule. This is the form given by Garcilaso and the Gentleman of Elvas; Biedma has Guasula, and Ranjel Guasili or Guasuli. The translators and commentators have given us such forms as Guachoule, Quaxule, Quaxulla, and Quexale. According to the Spanish method of writing Indian words the name was pronounced Washulé or Wasuli, which has a Cherokee sound, although it can not be translated. Buckingham Smith (Narratives, p. 222) hints that the Spaniards may have changed Guasili to Guasule, because of the similarity of the latter form to a town name in southern Spain. Such corruptions of Indian names are of frequent occurrence. Garcilaso speaks of it as a "province and town," while Biedma and Ranjel call it simply a town ("pueblo"). Before reaching this place the Indian queen had managed to make her escape. All the chroniclers tell of the kind reception which the Spaniards met here, but the only description of the town itself is from Garcilaso, who says that it was situated in the midst of many small streams which came down from the mountains round about, that it consisted of three hundred houses, which is probably an exaggeration, though it goes to show that the village was of considerable size, and that the chief's house, in which the principal officers were lodged, was upon a high hill ("un cerro alto"), around which was a roadway ("paseadero") wide enough for six men to walk abreast. By the "chief's house" we are to understand the town-house, while from various similar references in other parts of the narrative there can be no doubt that the "hill" upon which it stood was an artificial mound. In modern Spanish writing such artificial elevations are more often called lomas, but these early adventurers may be excused for not noting the distinction. Issuing from the mountains round about the town were numerous small streams, which united to form the river which the Spaniards henceforth followed from here down to Chiaha, where it was as large as the Guadalquivir at Sevilla (Garcilaso). Deceived by the occurrence, in the Portuguese narrative, of the name Canasagua, which they assumed could belong in but one place, earlier commentators have identified this river with the Coosa, Pickett putting Guaxule somewhere upon its upper waters, while Jones improves upon this by making the site "identical, or very nearly so, with Coosawattee Old town, in the southeastern corner of Murray county," Georgia. As we shall show, however, the name in question was duplicated in several states, and a careful study of the narratives, in the light of present knowledge of the country, makes it evident that the river was not the Coosa, but the Chattahoochee. Turning our attention once more to Xuala, the most northern point reached by De Soto, we have seen that this was the territory of the Suwala or Sara Indians, in the eastern foothills of the Alleghenies, about the head waters of Broad and Catawba rivers, in North Carolina. As the Spaniards turned here to the west they probably did not penetrate far beyond the present South Carolina boundary. The "very high mountain ridge" which they crossed immediately after leaving the town was in all probability the main chain of the Blue ridge, while the river which they found after descending to the savanna on the other side, and which they guessed to be a branch of the Mississippi, was almost as certainly the upper part of the French Broad, the first stream flowing in an opposite direction from those which they had previously encountered. They may have struck it in the neighborhood of Hendersonville or Brevard, there being two gaps, passable for vehicles, in the main ridge eastward from the first-named town. The uninhabited mountains through which they struggled for several days on their way to Chiaha and Coça (the Creek country) in the southwest were the broken ridges in which the Savannah and the Little Tennessee have their sources, and if they followed an Indian trail they may have passed through the Rabun gap, near the present Clayton, Georgia. Guaxule, and not Xuala, as Jones supposes, was in Nacoochee valley, in the present White county, Georgia, and the small streams which united to form the river down which the Spaniards proceeded to Chiaha were the headwaters of the Chattahoochee. The hill upon which the townhouse was built must have been the great Nacoochee mound, the most prominent landmark in the valley, on the east bank of Sautee creek, in White county, about twelve miles northwest of Clarkesville. This is the largest mound in upper Georgia, with the exception of the noted Etowah mound near Cartersville, and is the only one which can fill the requirements of the case. There are but two considerable mounds in western North Carolina, that at Franklin and a smaller one on Oconaluftee river, on the present East Cherokee reservation, and as both of these are on streams flowing away from the Creek country, this fact alone would bar them from consideration. The only large mounds in upper Georgia are this one at Nacoochee and the group on the Etowah river, near Cartersville. The largest of the Etowah group is some fifty feet in height and is ascended on one side by means of a roadway about fifty feet wide at the base and narrowing gradually to the top. Had this been the mound of the narrative it is hardly possible that the chronicler would have failed to notice also the two other mounds of the group or the other one on the opposite side of the river, each of these being from twenty to twenty-five feet in height, to say nothing of the great ditch a quarter of a mile in length which encircles the group. Moreover, Cartersville is at some distance from the mountains, and the Etowah river at this point does not answer the description of a small rushing mountain stream. There is no considerable mound at Coosawatee or in any of the three counties adjoining. The Nacoochee mound has been cleared and cultivated for many years and does not now show any appearance of a roadway up the side, but from its great height we may be reasonably sure that some such means of easy ascent existed in ancient times. In other respects it is the only mound in the whole upper country which fills the conditions. The valley is one of the most fertile spots in Georgia and numerous ancient remains give evidence that it was a favorite center of settlement in early days. At the beginning of the modern historic period it was held by the Cherokee, who had there a town called Nacoochee, but their claim was disputed by the Creeks. The Gentleman of Elvas states that Guaxule was subject to the queen of Cofitachiqui, but this may mean only that the people of the two towns or tribes were in friendly alliance. The modern name is pronounced Nagu`tsi' by the Cherokee, who say, however, that it is not of their language. The terminal may be the Creek udshi, "small," or it may have a connection with the name of the Uchee Indians. From Guaxule the Spaniards advanced to Canasoga (Ranjel) or Canasagua (Elvas), one or two days' march from Guaxule, according to one or the other authority. Garcilaso and Biedma do not mention the name. As Garcilaso states that from Guaxule to Chiaha the march was down the bank of the same river, which we identify with the Chattahoochee, the town may have been in the neighborhood of the present Gainesville. As we have seen, however, it is unsafe to trust the estimates of distance. Arguing from the name, Meek infers that the town was about Conasauga river in Murray county, and that the river down which they marched to reach it was "no doubt the Etowah," although to reach the first named river from the Etowah it would be necessary to make another sharp turn to the north. From the same coincidence Pickett puts it on the Conasauga, "in the modern county of Murray, Georgia," while Jones, on the same theory, locates it "at or near the junction of the Connasauga and Coosawattee rivers, in originally Cass, now Gordon county." Here his modern geography as well as his ancient is at fault, as the original Cass county is now Bartow, the name having been changed in consequence of a local dislike for General Cass. The whole theory of a march down the Coosa river rests upon this coincidence of the name. The same name however, pronounced Gansâ'gi by the Cherokee, was applied by them to at least three different locations within their old territory, while the one mentioned in the narrative would make the fourth. The others were (1) on Oostanaula river, opposite the mouth of the Conasauga, where afterward was New Echota, in Gordon county, Georgia; (2) on Canasauga creek, in McMinn county, Tennessee; (3) on Tuckasegee river, about two miles above Webster, in Jackson county, North Carolina. At each of these places are remains of ancient settlement. It is possible that the name of Kenesaw mountain, near Marietta, in Cobb county, Georgia, may be a corruption of Gansâ'gi, and if so, the Canasagua of the narrative may have been somewhere in this vicinity on the Chattahoochee. The meaning of the name is lost. On leaving Canasagua they continued down the same river which they had followed from Guaxule (Garcilaso), and after traveling several days through an uninhabited ("despoblado") country (Elvas) arrived at Chiaha, which was subject to the great chief of Coça (Elvas). The name is spelled Chiaha by Ranjel and the Gentleman of Elvas, Chiha by Biedma in the Documentos, China by a misprint in an English rendering, and Ychiaha by Garcilaso. It appears as Chiha on an English map of 1762 reproduced in Winsor, Westward Movement, page 31, 1897. Gallatin spells it Ichiaha, while Williams and Fairbanks, by misprint, make it Chiapa. According to both Ranjel and Elvas the army entered it on the 5th of June, although the former makes it four days from Canasagua, while the other makes it five. Biedma says it was four days from Guaxule, and, finally, Garcilaso says it was six days and thirty leagues from Guaxule and on the same river, which was, here at Chiaha, as large as the Guadalquivir at Sevilla. As we have seen, there is a great discrepancy in the statements of the distance from Cofitachiqui to this point. All four authorities agree that the town was on an island in the river, along which they had been marching for some time (Garcilaso, Ranjel), but while the Elvas narrative makes the island "two crossbow shot" in length above the town and one league in length below it, Garcilaso calls it a "great island more than five leagues long." On both sides of the island the stream was very broad and easily waded (Elvas). Finding welcome and food for men and horses the Spaniards rested here nearly a month (June 5-28, Ranjel; twenty-six or twenty-seven days, Biedma; thirty days, Elvas). In spite of the danger from attack De Soto allowed his men to sleep under trees in the open air, "because it was very hot and the people should have suffered great extremity if it had not been so" (Elvas). This in itself is evidence that the place was pretty far to the south, as it was yet only the first week in June. The town was subject to the chief of the great province of Coça, farther to the west. From here onward they began to meet palisaded towns. On the theory that the march was down Coosa river, every commentator hitherto has located Chiaha at some point upon this stream, either in Alabama or Georgia. Gallatin (1836) says that it "must have been on the Coosa, probably some distance below the site of New Echota." He notes a similarity of sound between Ichiaha and "Echoy" (Itseyi), a Cherokee town name. Williams (1837) says that it was on Mobile (i. e., the Alabama or lower Coosa river). Meek (1839) says "there can be little doubt that Chiaha was situated but a short distance above the junction of the Coosa and Chattooga rivers," i. e., not far within the Alabama line. He notes the occurrence of a "Chiaha" (Chehawhaw) creek near Talladega, Alabama. In regard to the island upon which the town was said to have been situated he says: "There is no such island now in the Coosa. It is probable that the Spaniards either mistook the peninsula formed by the junction of two rivers, the Coosa and Chattooga, for an island, or that those two rivers were originally united so as to form an island near their present confluence. We have heard this latter supposition asserted by persons well acquainted with the country."--Romantic Passages, p. 222, 1857. Monette (1846) puts it on Etowah branch of the Coosa, probably in Floyd county, Georgia. Pickett (1851), followed in turn by Irving, Jones, and Shea, locates it at "the site of the modern Rome." The "island" is interpreted to mean the space between the two streams above the confluence. Pickett, as has been stated, bases his statements chiefly or entirely upon Indian traditions as obtained from half breeds or traders. How much information can be gathered from such sources in regard to events that transpired three centuries before may be estimated by considering how much an illiterate mountaineer of the same region might be able to tell concerning the founding of the Georgia colony. Pickett himself seems to have been entirely unaware of the later Spanish expeditions of Pardo and De Luna through the same country, as he makes no mention of them in his history of Alabama, but ascribes everything to De Soto. Concerning Chiaha he says: "The most ancient Cherokee Indians, whose tradition has been handed down to us through old Indian traders, disagree as to the precise place [!] where De Soto crossed the Oostanaula to get over into the town of Chiaha--some asserting that he passed over that river seven miles above its junction with the Etowah, and that he marched from thence down to Chiaha, which, all contend, lay immediately at the confluence of the two rivers; while other ancient Indians asserted that he crossed, with his army, immediately opposite the town. But this is not very important. Coupling the Indian traditions with the account by Garcellasso and that by the Portuguese eyewitness, we are inclined to believe the latter tradition that the expedition continued to advance down the western side of the Oostanaula until they halted in view of the mouth of the Etowah. De Soto, having arrived immediately opposite the great town of Chiaha, now the site of Rome, crossed the Oostanaula," etc. (History of Alabama, p. 23, reprint, 1896). He overlooks the fact that Chiaha was not a Cherokee town, but belonged to the province of Coça--i. e., the territory of the Creek Indians. A careful study of the four original narratives makes it plain that the expedition did not descend either the Oostanaula or the Etowah, and that consequently Chiaha could not have been at their junction, the present site of Rome. On the other hand the conclusion is irresistible that the march was down the Chattahoochee from its extreme head springs in the mountains, and that the Chiaha of the narrative was the Lower Creek town of the same name, more commonly known as Chehaw, formerly on this river in the neighborhood of the modern city of Columbus, Georgia, while Coste, in the narrative the next adjacent town, was Kasi`ta, or Cusseta, of the same group of villages. The falls at this point mark the geologic break line where the river changes from a clear, swift current to a broad, slow-moving stream of the lower country. Attracted by the fisheries and the fertile bottom lands the Lower Creeks established here their settlement nucleus, and here, up to the beginning of the present century, they had within easy distance of each other on both sides of the river some fifteen towns, among which were Chiaha (Chehaw), Chiahudshi (Little Chehaw), and Kasi`ta (Cusseta). Most of these settlements were within what are now Muscogee and Chattahoochee counties, Georgia, and Lee and Russell counties, Alabama (see town list and map in Gatschet, Creek Migration Legend). Large mounds and other earthworks on both sides of the river in the vicinity of Columbus attest the importance of the site in ancient days, while the general appearance indicates that at times the adjacent low grounds were submerged or cut off by overflows from the main stream. A principal trail crossed here from the Ocmulgee, passing by Tuskegee to the Upper Creek towns about the junction of the Coosa and Tallapoosa in Alabama. At the beginning of the present century this trail was known to the traders as "De Soto's trace" (Woodward, Reminiscences, p. 76). As the Indian towns frequently shift their position within a limited range on account of epidemics, freshets, or impoverishment of the soil, it is not necessary to assume that they occupied exactly the same sites in 1540 as in 1800, but only that as a group they were in the same general vicinity. Thus Kasi`ta itself was at one period above the falls and at a later period some eight miles below them. Both Kasi`ta and Chiaha were principal towns, with several branch villages. The time given as occupied on the march from Canasagua to Chiaha would seem too little for the actual distance, but as we have seen, the chroniclers do not agree among themselves. We can easily believe that the Spaniards, buoyed up by the certainty of finding food and rest at their next halting place, made better progress along the smooth river trail than while blundering helplessly through the mountains at the direction of a most unwilling guide. If Canasagua was anywhere in the neighborhood of Kenesaw, in Cobb county, the time mentioned in the Elvas or Garcilaso narrative would probably have been sufficient for reaching Chiaha at the falls. The uninhabited country between the two towns was the neutral ground between the two hostile tribes, the Cherokee and the Creeks, and it is worth noting that Kenesaw mountain was made a point on the boundary line afterward established between the two tribes through the mediation of the United States government. There is no large island in either the Coosa or the Chattahoochee, and we are forced to the conclusion that what the chronicle describes as an island was really a portion of the bottom land temporarily cut off by back water from a freshet. In a similar way "The Slue," east of Flint river in Mitchell county, may have been formed by a shifting of the river channel. Two months later, in Alabama, the Spaniards reached a river so swollen by rains that they were obliged to wait six days before they could cross (Elvas). Lederer, while crossing South Carolina in 1670, found his farther progress barred by a "great lake," which he puts on his map as "Ushery lake," although there is no such lake in the state; but the mystery is explained by Lawson, who, in going over the same ground thirty years later, found all the bottom lands under water from a great flood, the Santee in particular being 36 feet above its normal level. As Lawson was a surveyor his figures may be considered reliable. The "Ushery lake" of Lederer was simply an overflow of Catawba river. Flood water in the streams of upper Georgia and Alabama would quickly be carried off, but would be apt to remain for some time on the more level country below the falls. According to information supplied by Mr Thomas Robinson, an expert engineering authority familiar with the lower Chattahoochee, there was formerly a large mound, now almost entirely washed away, on the eastern bank of the river, about nine miles below Columbus, while on the western or Alabama bank, a mile or two farther down, there is still to be seen another of nearly equal size. "At extreme freshets both of these mounds were partly submerged. To the east of the former, known as the Indian mound, the flood plain is a mile or two wide, and along the eastern side of the plain stretches a series of swamps or wooded sloughs, indicating an old river bed. All the plain between the present river and the sloughs is river-made land. The river bluff along by the mound on the Georgia side is from twenty to thirty feet above the present low-water surface of the stream. About a mile above the mound are the remains of what was known as Jennies island. At ordinary stages of the river no island is there. The eastern channel was blocked by government works some years ago, and the whole is filled up and now used as a cornfield. The island remains can be traced now, I think, for a length of half a mile, with a possible extreme width of 300 feet.... This whole country, on both sides of the river, is full of Indian lore. I have mentioned both mounds simply to indicate that this portion of the river was an Indian locality, and have also stated the facts about the remains of Jennies island in order to give a possible clew to a professional who might study the ground."--Letter, April 22, 1900. Chiaha was the first town of the "province of Coça," the territory of the Coosa or Creek Indians. The next town mentioned, Coste (Elvas and Ranjel), Costehe (Biedma) or Acoste (Garcilaso), was Kasi`ta, or Cusseta, as it was afterward known to the whites. While Garcilaso puts it at the lower end of the same island upon which Chiaha was situated, the Elvas narrative makes it seven days distant! The modern towns of Chehaw and Cusseta were within a few miles of each other on the Chattahoochee, the former being on the western or Alabama side, while Cusseta, in 1799, was on the east or Georgia side about eight miles below the falls at Columbus, and in Chattahoochee county, which has given its capital the same name, Cusseta. From the general tone of the narrative it is evident that the two towns were near together in De Soto's time, and it may be that the Elvas chronicle confounded Kasi`ta with Koasati, a principal Upper Creek town, a short distance below the junction of the Coosa and Tallapoosa. At Coste they crossed the river and continued westward "through many towns subject to the cacique of Coça" (Elvas) until they came to the great town of Coça itself. This was Kusa or Coosa, the ancient capital of the Upper Creeks. There were two towns of this name at different periods. One, described by Adair in 1775 as "the great and old beloved town of refuge, Koosah," was on the east bank of Coosa river, a few miles southwest of the present Talladega, Alabama. The other, known as "Old Coosa," and probably of more ancient origin, was on the west side of Alabama river, near the present site of Montgomery (see Gatschet, Creek Migration Legend). It was probably the latter which was visited by De Soto, and later on by De Luna, in 1559. Beyond Coca they passed through another Creek town, apparently lower down on the Alabama, the name of which is variously spelled Ytaua (Elvas, Force translation), Ytava (Elvas, Hakluyt Society translation), or Itaba (Ranjel), and which may be connected with I'tawa', Etowah or "Hightower," the name of a former Cherokee settlement near the head of Etowah river in Georgia. The Cherokee regard this as a foreign name, and its occurrence in upper Georgia, as well as in central Alabama, may help to support the tradition that the southern Cherokee border was formerly held by the Creeks. De Soto's route beyond the Cherokee country does not concern us except as it throws light upon his previous progress. In the seventeenth chapter the Elvas narrative summarizes that portion from the landing at Tampa bay to a point in southern Alabama as follows: "From the Port de Spirito Santo to Apalache, which is about an hundred leagues, the governor went from east to west; and from Apalache to Cutifachiqui, which are 430 leagues, from the southwest to the northeast; and from Cutifachiqui to Xualla, which are about 250 leagues, from the south to the north; and from Xualla to Tascaluca, which are 250 leagues more, an hundred and ninety of them he traveled from east to west, to wit, to the province of Coça; and the other 60, from Coça to Tascaluca, from the north to the south." Chisca (Elvas and Ranjel), the mountainous northern region in search of which men were sent from Chiaha to look for copper and gold, was somewhere in the Cherokee country of upper Georgia or Alabama. The precise location is not material, as it is now known that native copper, in such condition as to have been easily workable by the Indians, occurs throughout the whole southern Allegheny region from about Anniston, Alabama, into Virginia. Notable finds of native copper have been made on the upper Tallapoosa, in Cleburne county, Alabama; about Ducktown, in Polk county, Tennessee, and in southwestern Virginia, one nugget from Virginia weighing several pounds. From the appearance of ancient soapstone vessels which have been found in the same region there is even a possibility that the Indians had some knowledge of smelting, as the Spanish explorers surmised (oral information from Mr W. H. Weed, U. S. Geological Survey). We hear again of this "province" after De Soto had reached the Mississippi, and in one place Garcilaso seems to confound it with another province called Quizqui (Ranjel) or Quizquiz (Elvas and Biedma). The name has some resemblance to the Cherokee word tsiskwa, "bird." (9) De Luna and Rogel (p. 27): Jones, in his De Soto's March through Georgia, incorrectly ascribes certain traces of ancient mining operations in the Cherokee country, particularly on Valley river in North Carolina, to the followers of De Luna, "who, in 1560 ... came with 300 Spanish soldiers into this region, and spent the summer in eager and laborious search for gold." Don Tristan de Luna, with fifteen hundred men, landed somewhere about Mobile bay in 1559 with the design of establishing a permanent Spanish settlement in the interior, but owing to a succession of unfortunate happenings the attempt was abandoned the next year. In the course of his wanderings he traversed the country of the Choctaw, Chickasaw, and Upper Creeks, as is shown by the names and other data in the narrative, but returned without entering the mountains or doing any digging (see Barcia, Ensayo Cronologico, pp. 32-41, 1723; Winsor, Narrative and Critical History, II, pp. 257-259). In 1569 the Jesuit Rogel--called Father John Roger by Shea--began mission work among the South Carolina tribes inland from Santa Elena (about Port Royal). The mission, which at first promised well, was abandoned next year, owing to the unwillingness of the Indians to give up their old habits and beliefs. Shea, in his "Catholic Missions," supposes that these Indians were probably a part of the Cherokee, but a study of the Spanish record in Barcia (Ensayo, pp. 138-141) shows that Rogel penetrated only a short distance from the coast. (10) Davies' History of the Carribby Islands (p. 29): The fraudulent character of this work, which is itself an altered translation of a fictitious history by Rochefort, is noted by Buckingham Smith (Letter of Hernando de Soto, p. 36, 1854), Winsor (Narrative and Critical History, II, p. 289), and Field (Indian Bibliography, p. 95). Says Field: "This book is an example of the most unblushing effrontery. The pseudo author assumes the credit of the performance, with but the faintest allusion to its previous existence. It is a nearly faithful translation of Rochefort's 'Histoire des Antilles.' There is, however, a gratifying retribution in Davies' treatment of Rochefort, for the work of the latter was fictitious in every part which was not purloined from authors whose knowledge furnished him with all in his treatise which was true." (11) Ancient Spanish Mines (pp. 29, 31): As the existence of the precious metals in the southern Alleghenies was known to the Spaniards from a very early period, it is probable that more thorough exploration of that region will bring to light many evidences of their mining operations. In his "Antiquities of the Southern Indians," Jones describes a sort of subterranean village discovered in 1834 on Dukes creek, White county, Georgia, consisting of a row of small log cabins extending along the creek, but imbedded several feet below the surface of the ground, upon which large trees were growing, the inference being that the houses had been thus covered by successive freshets. The logs had been notched and shaped apparently with sharp metallic tools. Shafts have been discovered on "Valley river, North Carolina, at the bottom of one of which was found, in 1854, a well-preserved windlass of hewn oak timbers, showing traces of having once been banded with iron. Another shaft, passing through hard rock, showed the marks of sharp tools used in the boring. The casing and other timbers were still sound (Jones, pp. 48, 49). Similar ancient shafts have been found in other places in upper Georgia and western North Carolina, together with some remarkable stone-built fortifications or corrals, notably at Fort mountain, in Murray county, Georgia, and on Silver creek, a few miles from Rome, Georgia. Very recently remains of an early white settlement, traditionally ascribed to the Spaniards, have been reported from Lincolnton, North Carolina, on the edge of the ancient country of the Sara, among whom the Spaniards built a fort in 1566. The works include a dam of cut stone, a series of low pillars of cut stone, arranged in squares as though intended for foundations, a stone-walled well, a quarry from which the stone had been procured, a fire pit, and a series of sinks, extending along the stream, in which were found remains of timbers suggesting the subterranean cabins on Dukes creek. All these antedated the first settlement of that region, about the year