Exploration and Trade in the Deep Frontier of Spanish Florida:  
Possible Sources for 16th-Century Spanish Artifacts in Western North Carolina  
John E. Worth  
Fernbank Museum of Natural History, Atlanta  
Abstract  
Recent archaeological investigations in western North Carolina have revealed the presence of 16th century Spanish artifacts in association with contemporaneous aboriginal occupation. This paper  examines the available Spanish historical evidence regarding the various mechanisms by which  such items may have been disseminated among aboriginal populations in the deep interior  Southeast. Possible sources include direct or indirect contact with the well-known Hernando de  Soto or Juan Pardo expeditions between 1540 and 1568, dispersal as a result of several minor  Spanish entradas into the northern interior between 1597 and 1628, and ongoing long-distance trade  with aboriginal and Spanish populations along the Atlantic coastline.  
Paper presented in the symposium "A Consideration of Archaeological and Documentary Evidence  for 16th- and 17th-Century Spanish and Native Contact in Western North Carolina" at the 51st  annual meeting of the Southeastern Archaeological Conference, Lexington.
 The recent identification of a variety of 16th-century Spanish artifacts at the Berry site in  western North Carolina provides tantalizing evidence for some form of direct or indirect contact  between Spanish explorers or colonists and aboriginal groups living in the deep frontier of Spanish  Florida. Although rare, Spanish artifacts dating to this period have nevertheless previously been  found in a number of contemporaneous aboriginal sites in western North Carolina and eastern  Tennessee, making the finds at the Berry site only the most recent discovery in this region of the  Appalachian highlands. The nature of the Spanish artifact assemblage at the Berry site, however,  seems to be largely unique for this region, prompting questions as to the origin of these so-called  "trade goods." Sixteenth-century Spanish artifacts identified to date at the Berry site include an  iron knife recovered from an aboriginal burial, a blue glass bead, a rolled brass bead, nearly a dozen  sherds of Spanish ceramics, a lump of molten lead shot, and a forged iron nail (Moore and Beck  1994). As I will discuss in this paper, an examination of Spanish documentary sources dating to  the 16th and early 17th centuries, combined with a comparison with other assemblages of Spanish  artifacts from this region of the Appalachian highlands, strongly suggests that most of the Berry site  assemblage was not the result of direct or indirect trade with Spanish sources, but rather was the  result of a short-term occupation of the Berry site by Spanish soldiers during the late 16th century.  Specifically, based on the written accounts of the Juan Pardo expeditions of 1566-1568, the Berry  site may be tentatively identified as the capital town of the late 16th-century aboriginal chiefdom of  Joara, and thus the site of Fort San Juan, the first Spanish military fortification constructed in the  deep interior of the southeastern United States.  
 The presence of 16th-century Spanish artifacts within the archaeological context of  contemporaneous aboriginal villages in the interior Southeast can be explained by several  mechanisms. Such artifacts may be broken down into two broad categories--trade goods and 
occupation debris. Specifically, these artifacts were either exchanged with or given to Indians by  Spaniards (either as a result of direct or indirect contact), or they represent debris left by Spaniards  as a result of temporary occupation within or adjacent to established Indian settlements. Although  many artifacts in this second category were undoubtedly scavenged and re-used by Indians in the  interior after their deposition at a site, the functional distinction between these two categories of  artifacts relates more to the original intended function of these artifacts, and not to their eventual  fate in the archaeological record.  
 In this connection, it is instructive to review the historical and archaeological evidence for  both late 16th-century trade and occupation in western North Carolina resulting from the  contemporary Spanish colonial presence to the south. Only two Spanish exploratory expeditions  ever reached the deep northern interior region now known as western North Carolina, and these  therefore represent the only possibility of direct, or face-to-face, trade between 16th-century  Spaniards and aboriginal groups in this area. The first of these expeditions was that of Hernando de  Soto, whose 600-man army swung eastward and northward from the Gulf coast region of present day Florida to make contact with several powerful and populus aboriginal chiefdoms in the  Carolinas during 1540 (see Hudson et al. 1984). The second and last expedition to reach this area  was that of Juan Pardo, who set out twice from the newly-established Spanish colonial town of  Santa Elena (on modern Parris Island), ultimately retracing portions of Soto's earlier route into the  northern interior between 1566 and 1568 (see Hudson 1990).  
 Now, inasmuch as a precise and accurate reconstruction of the actual route of Hernando de  Soto and Juan Pardo is pivotal in understanding the possible origin of Spanish artifacts at the Berry  site, I will first discuss recently identified--and what I believe to be conclusive--documentary proof  that the Berry site was actually situated along the general track of the routes of both Soto and 
Pardo, and thus may have been visited by the members of both expeditions. The Soto and Pardo  routes have been and continue to be the subject of considerable controversy. Despite the  conclusions reached by the United States DeSoto Commission in the 1930s (Swanton 1939), recent  research by Charles Hudson, Chester DePratter, and Marvin Smith, in concert with investigations  by a number of other scholars across the Southeast, has resulted in a substantially revised  reconstruction of the Soto and Pardo routes (Hudson et al. 1984). While the specific details of their  evidence will not be reviewed here, it is important to note that what may be referred to as the  Hudson route is based on the combined use of modern archaeological data on the geographic  distribution of mid-16th-century archaeological sites and a detailed evaluation of all available  primary historical sources regarding these expeditions.  
 Much of the eastern portion of the 1540 Soto route was later retraced by the Juan Pardo  expeditions, which are somewhat better documented with respect to individual town locations. It is  this fact that helped Hudson and his colleagues initially revise the route of the earlier Hernando de  Soto expedition, which visited several of the same towns documented during the Pardo expeditions  (see discussion by Hudson et al. 1984). Specifically, the most detailed of the four contemporary  accounts of the Pardo expeditions, written by notary Juan de la Bandera, helped demonstrate that  the location of the important chiefdom of Cofitachequi (also known as Canos), and thus the river  corridor taken by Soto and Pardo northward toward the Appalachian summit, was not on the  Savannah River between Georgia and South Carolina, but instead along the Wateree River in  central South Carolina.  
 Despite a considerable amount of evidence supporting this revised route, including  archaeological data indicating that the middle Savannah River was largely uninhabited during the  16th century, there remains some controversy over the Wateree River location for Cofitachequi. 
Since the Berry site is situated along the Hudson route, and far to the east of the Savannah River  route, interpretations regarding the origin of Spanish artifacts at the Berry site hinge upon which  route is correct. Fortunately, new documentary evidence has recently come to light which strongly  supports the Hudson route.  
 Four major documentary accounts describing the Hernando de Soto expedition have been  utilized in the past for reconstructing Soto's route, including a relation written by the King's factor  Luys Hernández de Biedma, Oviedo's summary of the diary of de Soto's private secretary Rodrigo  Ranjel, an anonymous Portuguese account written by a participant in the expedition, and a  secondary account compiled at a much later date by Garcilaso de la Vega (translations in Clayton et  al. 1993). Another four major documentary accounts relative to the later Juan Pardo expeditions  have been employed for reconstructing Pardo's route. These include an overview written by Juan  Pardo himself, a letter written by soldier Francisco Martínez and witnessed by several other  soldiers on the expedition, and two accounts by Pardo's notary Juan de la Bandera--one  representing the day by day affairs of Pardo's second expedition, and a summary based on this  relation (translations in Hudson 1990).  
 More recently, I have identified yet another account of the Juan Pardo expeditions, written  in 1584 by soldier, notary, and interpreter Domingo de León, who not only participated in the  Pardo expeditions but also served as an important interpreter with the Indians around Santa Elena  during the years following the entradas (Worth n.d.). The León account represents a recollection  penned some 16 years after the fact, but contains a variety of details not mentioned in other  accounts of the Pardo expeditions, including the fate of some of its other participants and survivors.  Indeed, the fact that Domingo de León incorporated not only the details of his own personal  experience in the interior, but also information gleaned from Indians around Santa Elena, 
apparently in contact with groups in the interior, makes his account unique among all extant  accounts in the sense that León provided a broad overview of the physical and social geography of  the Carolina interior during the late 16th century. No other account, including those relative to the  Soto expedition, provides such a revealingly broad overview of the locations of 16th-century  aboriginal chiefdoms in the interior of the Carolinas with respect to major river drainages and  mountain ranges. In several instances, León's insightful comments regarding settlement  distributions and sociopolitical relationships provide tantalizing clues that archaeologists could  pursue for years to come.  
 While I do not wish to overstate the importance of the León account, it is important to  discuss here the implications of his description of the interior regarding the general route of Juan  Pardo into the interior. Specifically, a detailed examination of the text of Domingo de León's  account makes it possible to extrapolate a sort of "mental map" of the interior, representing León's  understanding of the physical and social geography of the regions traversed by Pardo, focusing on  the region of Canos, or Cofitachequi. Closer examination of the map derived from León's textual  account reveals considerable information regarding the location of the town and province called  Canos with respect to the river system on which it was situated, and with respect to other physical  and social features of the interior landscape. León describes Canos, or Cofitachequi, as being 40  leagues north of Santa Elena. Furthermore, he locates Canos on a river system that comprised two  major branches that joined 30 leagues inland from the sea. An unnamed town at the junction of  these rivers marked the beginning of population along the northwestern branch of the river, and  towns were spaced at between half a league and a league apart. Canos was said to be 12 leagues  upriver from this first town, and comprised a district stretching for 20 leagues in which the towns  were more closely spaced, and on the eastern bank of the river. Farther into the interior along this 
northwestern river, yet another cacique was said to be located at the foot of the mountains, and the  settlements of this populus province were said to be more dispersed as a result of warfare.   Interestingly, León's description of the western branch of this river system seems more  fanciful in retrospect, since León noted that although it was not populated along its lower reaches,  his Indian informants attested to the existance of a large, long lake at the river's source, supposedly  well-populated. This area was known as the "River of the Shells." On the other side of the  mountains, estimated to be 50 leagues away, was said to be yet another large, populated lake, this  time round in shape, from which a tremendous river ran toward the west. Examination of León's  text and "mental map" makes it clear that he had never personally visited these mythical regions, in  contrast to his more detailed description of the northwestern branch where Canos was located.  Other evidence in the León account suggests that he may have been stationed at Pardo's first fort  named San Juan, located in the town of Joara, and thus never crossed the mountains to the west.   Significantly, León's description of these river systems and the aboriginal chiefdoms on  them corresponds remarkably well to the Pardo route reconstructed by Charles Hudson and his  colleages. Indeed, the similarity between Hudson's map, constructed without the benefit of the  León account, and Domingo de León's "mental map," is uncanny. The locations of river systems,  population centers, and even the locations of uninhabited stretches of major rivers, match all other  historical and archaeological data regarding the interior of 16th-century South Carolina. What is  perhaps most telling, however, is a further comparison of the river system described by Domingo  de León, with its two major branches 30 leagues from the coast (both of which reached to the  mountains), with the course of the Savannah River, long argued to be the site of Cofitachequi. The  Savannah lacks any major branches so close to the coast, and none of the feeder creeks along the  western bank of the river extend even close to the mountains. 
 Consequently, I believe the newly-discovered Domingo de León account to represent  substantial proof for Charles Hudson's reconstruction of Juan Pardo's route, and thus also for his  Hernando de Soto route. There is effectively no reasonable way to fit a Savannah River location  for Pardo's Canos and Soto's Cofitachequi into León's description of the regions he viewed. Unless  the León account is dismissed outright (which would be difficult to justify, considering his lengthy  record of service as a trusted and able interpreter among the Carolina Indians), the general course  of the Hudson reconstruction of Pardo's route must be concluded to be accurate.   Now, given that the Hudson's reconstruction of Juan Pardo's route is accurate in its general  course, we may conclude that the expeditions of both Hernando de Soto and Juan Pardo passed  very close to the Berry site. Indeed, the Berry site is only 20 miles from Hudson's predicted  location of the important interior capital town of Joara, visited by both the Soto and Pardo  expeditions (Hudson 1990). This aboriginal town was the site of Fort San Juan, the first and most  important of a string of Spanish fortifications constructed and manned by Pardo's soldiers during  his two expeditions. Consequently, the inhabitants of the Berry site had several opportunities for  face-to-face contact with Spanish soldiers between 1540 and 1568, making it at least possible that  some of the Spanish artifacts found there resulted from direct trade or gift-giving, and even as a  result of short-term Spanish occupation at Fort San Juan.  
 Since the Juan Pardo expeditions are more well-documented with respect to trade goods  distributed, and since the late 16th-century date of several of the artifacts at the Berry site  correspond with this later expedition, it is informative to review the Spanish artifacts known to  have been distributed in the deep interior during Pardo's journey. Trade goods appearing on  Pardo's official records as having been given away to Indian leaders include the following items: 35  iron axes, 126 iron chisels and wedges, 32 iron knives, 29 bead necklaces, 6 mirrors, several gilded 
buttons, and an assortment of pieces of Spanish cloth (Hudson 1990: 134-38). In the town of Joara  itself, the aboriginal leader Joara Mico was noted to have been given a small battle axe and another  axe, and 8 "long wedges like chisels" and 8 large knives for him and other caciques subject to him  (Hudson 1990: 265).  
 Of these items, at least three of the Spanish artifacts in the Berry site assemblage correspond  directly to the list of trade goods distributed in the interior, namely the blue glass bead, the rolled  brass bead, and the iron knife found buried with a member of Joara's aboriginal elite. Indeed, the  iron knife could easily be one of the knives distributed to a cacique subject to Joara Mico, who was  subsequently interred with the gift. None of the other items found at the Berry site, however, match  any of the categories of objects known to have been distributed as gifts or trade goods. In this  sense, the Berry site departs from other contemporaneous assemblages of Spanish artifacts in  western North Carolina, in that other sites have generally produced what might be considered  typical "trade goods," including principally glass beads and iron tools (i.e. Rogers and Brown  1994). Items such as those listed for the Pardo expeditions, along with brass bells probably specific  to the Soto expedition, effectively match Marvin Smith's (1987) list of trade goods generally  associated with early to mid-16th century archaeological contexts in the interior Southeast,  suggesting that early Spanish exploratory expeditions generally carried a similar range of objects  for use in direct trade or gift-giving.  
 An examination of documentary and archaeological evidence for long-distance trade in the  late 16th and early 17th centuries suggests that although indirect trade might also be responsible for  the iron knife and beads found at the Berry site, this mechanism cannot reasonably explain the  other items either. The long-distance exchange networks among aboriginal leaders or groups  during this period undoubtedly resulted in the dispersal of items of Spanish manufacture far beyond 
their original source, and in the absence of face-to-face contact between Spaniards and Indians.  Spaniards did enter the interior after the Pardo expeditions, and trade goods may have entered  exchange networks during such episodes. Nevertheless, for the period immediately following the  Pardo expeditions, I have been able to document only seven small military exploratory expeditions  which penetrated relatively short distances into the northern interior between 1597 and 1628, and  only the final of these pushed even as far north as Cofitachequi (Worth 1994). Trade goods carried  during these expeditions were limited in number, but definitely included blankets, knives, fish hooks, scissors, glass beads, sickes, axes, and chisels. Other items listed as planned trade goods for  interior expeditions during this period included hoes and mirrors. However, even if trade goods or  supplies from these minor entradas did manage to reach the Berry site, accounting for the knife, the  beads, and perhaps even the lead shot, there is still no way to effectively account for the Spanish  ceramics or the forged iron nail.  
 The only remaining avenue for such indirect trade would be an extremely long-distance  trade system originating either in Santa Elena or St. Augustine during the late 16th century, or in  the 17th-century Spanish mission provinces west and north of St. Augustine. The deep interior  town of Joara was indeed involved in long-distance aboriginal trade, for in 1605, two Cayagua  Indians from coastal South Carolina made mention of Joara in this context, noting it to be thirty  days into the interior (Hann 1986). Marvin Smith (1987) and Gregory Waselkov (1989) have  compiled archaeological evidence for the existance of such an interior trade network during the late  16th and 17th centuries, involving trade goods such as glass beads, iron chisels, wedges, spikes,  hoes, and eyed axes, brass bells, and a wide assortment of sheet brass ornaments such as conical  bangles, disc gorgets, collars, armbands, and animal effigies. This trade network seems to have  been specifically directed toward the regions far to the west of the Berry site, however, across 
northern Alabama and eastern Tennessee. Nevertheless, while the Spanish knife and beads  recovered at the Berry site fall within the range of items involved in such long-distance trade, there  is no hint that Spanish ceramics, lead shot, or iron nails ever entered into such a trade network,  effectively eliminating such indirect trade as a probable mechanism by which these items arrived at  the Berry site.  
 Since neither direct nor indirect trade seems a likely explanation for a number of Spanish  items found at the Berry site, we must examine the possibility that such objects represent  occupational debris. While the Soto expedition only stayed in Xuala over a single weekend,  members of the Pardo expedition remained in Joara for many months. The list of general supplies  known to have been brought into the interior by Juan Pardo for the use of his troops includes the  following: 639 pounds of gunpowder, 428 pounds of matchcord, 502 pounds of lead shot, 2 iron  picks, 7 iron shovels, 5 maddocks, 5 crossbows with 240 bolts, 77 linen sacks, 178 pair of shoes,  265 pair of fiber sandals, 878 pounds of biscuit, 72 liters of wine, and 10 cheeses (Hudson 1990).  This list, of course, does not include weapons and personal gear carried by each soldier.  Importantly, it should be noted that this list of supplies does not mention containers used for items  such as gunpowder and wine. Powder was probably carried in small wooden barrels or kegs, and  the wine was most likely carried in the ceramic containers typically used by colonial Spaniards for  storage and transport, known as botijas at the time, and more commonly referred to as "olive jars"  by modern archaeologists.  
 Given that the Berry site is in the immediate vicinity of Hudson's predicted location for Juan  Pardo's town of Joara, a review of all the items of Spanish manufacture known to have been left  specifically in Joara is also informative. During Pardo's first and second expeditions, Fort San Juan  was provisioned a wide array of supplies. In addition to the weapons and equipment carried by 
each soldier stationed at the fort, Pardo is known to have issued the following military supplies:  235 pounds of gunpowder, 201 pounds of matchcord for arquebuses, 235 pounds of lead shot, and  4 crossbows along with 240 crossbow bolts. A considerable amount of construction supplies were  also issued, including 6 iron shovels, 4 mattocks, 4 picks, 42 chisels, 2 socketed axes, 4 iron  wedges, and 34 pounds of nails (Hudson 1990: 148, 150).  
 Interestingly, and of considerable significance for the present paper, Fort San Juan at Joara  is the only site in the interior specifically noted to have been left with a supply of nails during the  Juan Pardo expeditions. Based on Juan de la Bandera's account, none of the other interior forts was  left with nails, making Joara the principal location in the interior where archaeologists might expect  to find 16th-century Spanish nails. This is not surprising, particularly considering the often  extreme scarcity of forged iron nails in colonial Spanish Florida, even for Spaniards. Documentary  references to the lack of nails in Florida are not uncommon, and the problem was emphasized in a  1630 letter by Fray Francisco Alonso de Jesús, who specifically petitioned the King of Spain for a  supply of nails for the more remote Franciscan missions in Florida. Given, therefore, that  Spaniards themselves frequently had difficulty obtaining supplies of iron nails, it is no wonder that  such objects are never mentioned as items of trade with Indians.  
 It would therefore seem to be no coincidence that the Berry site is so close to the predicted  location of the aboriginal town of Joara and its Fort San Juan. The recovery of an unmodified 16th century forged iron nail at the Berry site is fortunately supplemented by 9 sherds from at least three  and possibly four Spanish olive jars, 1 sherd of Caparra Blue Spanish majolica tableware, and a  sherd of what seems to be a Spanish ceramic type referred to as "greyware," otherwise unique to  the 16th-century town of Santa Elena from which Juan Pardo departed (Moore and Beck 1994).  All of these objects together make a strong case for the tentative identification of the Berry site as 
the location of Juan Pardo's Fort San Juan, and the important aboriginal town of Joara. The  recovery of so many unmodified sherds of at least three distinct Spanish ceramic vessels used for  transport and storage, and the fact that between 2 and 4 sherds were recovered from each vessel,  strongly implies that these represent breakage as the result of a comparatively extended stay by  Spaniards at the site. Indeed, these could be fragments of several of the botijas, or olive jars, used  to carry the 72 liters of wine brought on the Pardo expeditions. The majolica may well be personal  tableware used by soldiers stationed at the fort. The lead shot is easily explained as military  supplies. The final piece of evidence is the forged iron nail, which provides a strong link not only  with Pardo's chain of forts, but specifically with Joara and Fort San Juan, thus serving literally to  nail down the argument.  
 In conclusion, the assemblage of 16th-century Spanish artifacts at the Berry site most likely  has its origin in a combination of direct trade and short-term Spanish occupation at the site, both of  which probably relate to the Juan Pardo expeditions of the late 1560s. A review of documentary  evidence for trade goods and military supplies brought on the expedition indicates that the Berry  site uniquely satisfies what would be expectated at the site of Joara, standing alone among other  contemporaneous assemblages in other sites in western North Carolina, which presumably reflect  trade goods distributed to neighboring Indian leaders during the expedition. Nevertheless, further  archaeological testing at the Berry site is clearly needed to test this proposition. Final proof would  presumably come in the form of a small Spanish fortification located at the site, and if such a  discovery is made, the Berry site will instantly become an extremely important archaeological site.  Not only was Joara the site of the first Spanish military fortification constructed in the deep interior  of the southeastern United States, but it was also the center of an emerging aboriginal chiefdom in  western North Carolina, and its identification would provide archaeologists with a unique 
opportunity to combine ethnohistorical and archaeological evidence in the examination of an  indigenous society at the dawn of the European colonial era. 
References Cited  
Clayton, Lawrence A., Vernon James Knight, Jr., Edward C. Moore (eds.)  
 1993 The DeSoto Chronicles: The Expedition of Hernando de Soto to North America in 1539- 1543. University of Alabama Press, Tuscaloosa.  
Hann, John H.  
 1986 Translation of the Ecija Voyages of 1605 and 1609 and the González Derrotero of 1609.  Florida Archaeology 2: 1-79.  
Hudson, Charles  
 1990 The Juan Pardo Expeditions: Exploration of the Carolinas and Tennessee, 1566-1568.  Smithsonian Institution Press, Washington.  
Hudson, Charles, Marvin Smith, and Chester DePratter  
 1984 The Hernando de Soto Expedition: From Apalachee to Chiaha. Southeastern Archaeology 3: 65-77.  
Jesús, Francisco Alonso de  
 1630 Letter to the crown, March 2, 1630. Archivo General de Indias, Santo Domingo 235.  
Moore, David, and Robin Beck  
 1994 New Evidence of 16th-Century Spanish Artifacts in the Catawba River Valley, North  Carolina. Paper presented in the symposium "A Consideration of Archaeological and  Documentary Evidence for 16th- and 17th-Century Spanish and Native Contact in Western  North Carolina" at the 51st annual meeting of the Southeastern Archaeological Conference,  Lexington, November 9-12.  
Rogers, Anne, and Jane Brown  
 1994 Artifacts from the Hiawassee River Valley in North Carolina. Paper presented in the  symposium "A Consideration of Archaeological and Documentary Evidence for 16th- and  17th-Century Spanish and Native Contact in Western North Carolina" at the 51st annual  meeting of the Southeastern Archaeological Conference, Lexington, November 9-12.  
Smith, Marvin T.  
 1987 Archaeology of Aboriginal Culture Change in the Interior Southeast: Depopulation during  the Early Historic Period. University of Florida Press, Gainesville.  
Swanton, John R.  
 1939 Final Report of the United States De Soto Expedition Commission. Smithsonian Institution  Press, Washington, D.C. (reprinted 1985).  
Waselkov, Gregory  
 1989 Seventeenth-Century Trade in the Colonial Southeast. Southeastern Archaeology 8(2): 117- 33. 
Worth, John E.  
 1994 Late Spanish Military Expeditions in the Interior Southeast, 1597-1628. In The Forgotten  Centuries: Indians and Europeans in the American South, 1521-1704, ed. by Charles  Hudson and Carmen Chaves Tesser, pp. 104-22. University of Georgia Press, Athens.  
 n.d. Recollections of the Juan Pardo Expeditions: The 1584 Domingo de León Account. Draft  manuscript in possession of the author.