SMITHSONIAN INSTITUTIONBureau of American EthnologyBulletin 191 Anthropological Papers, No. 72ARCHEOLOGICAL NOTES ON ALMIRANTE BAY,BOCAS DEL TORO, PANAMABy MATTHEW W. and MARION STIRLING 255 682-611—64 21 CONTENTS PAGEIntroduction 259Historical background 260Ceramic studies 263Site descriptions 264Darkland (Boc-1) 264Darkland Cave (Boc-2) 267Boca del Drago (Boc-3) 275Jungle Point 277Cocos Island 278Bibliography 278Explanation of plates 279ILLUSTRATIONSPLATES(All plates follow page 284)28. Ceramic ware from Bocas del Toro, site 1.29. Ceramic ware from Bocas del Toro, site 1.30. Tripod vessels and subglobular bowls from Bocas del Toro, site 2.31. Ceramic ware from Bocas del Toro, site 2.32. Ceramic ware from Bocas del Toro, site 2.33. Ceramic ware from Bocas del Toro, site 2.34. Ceramic ware from Bocas del Toro, site 2.35. Large vessels with high collars from Bocas del Toro, site 2.36. Ceramic ware exteriors from Bocas del Toro, site 2.37. Interiors of sherds shown in plate 36.38. Nonceramic artifacts from Bocas del Toro, site 2.39. Ceramic ware from Bocas del Toro, site 3.40. Ceramic ware from Bocas del Toro, site 3.41. Stone female figure from Cocos Island, Almirante Bay; stone metates fromCricamola River and Bastimentos Key.42. Midden mound and mortars, Darkland Point.43. Stone artifacts from Escudo Island, Jungle Point, and Darkland Cave.44. Jaguar heads and double-bitted axes.TEXT FIGURES PAGE18. Sherds from Bocas del Toro, site 1 26619. Bowl with slotted pedestal base 26820. Sherds from Bocas del Toro, site 2 27021. Sherds from Bocas del Toro, site 2 27122. Sherds from Bocas del Toro, site 2 27223. Shallow bowl from Bocas del Toro, site 2 27424. Sherds from Bocas del Toro, site 3 276257 ARCHEOLOGICAL NOTES ON ALMIRANTE BAYBOCAS DEL TORO, PANAMA By Matthew W. and Marion Stirling INTRODUCTIONSince there appears to be no information on the archeology of theAlmirante Bay-Chiriqui Lagoon region, we, together with Richard H,Stewart, took occasion to visit briefly Almirante Bay during the monthof April 1953 as an extension of our work for the National GeographicSociety-Smithsonian Institution on the north coast of Panama thepreceding winter.During 10 days of rather intensive activity, we visited aU of theprincipal islands and a considerable portion of the mainland coast ofAlmirante Bay. Systematic questioning of natives elicited very littleinformation as to archeological sites. Most informants denied havingseen or heard of evidences of aboriginal occupation in the area. In-deed, on completion of our survey we concluded that archeologicalsites in the vicinity are neither numerous nor conspicuous. However,we did locate three sites worthy of excavation, and we photographedseveral carved stone objects found in the region.In the relatively level area north of Almirante, which has beenextensively cleared for banana plantations, nothing appears to havecome to light. In the remainder of the district the search for sites ishampered by the heavy growth of tropical jungle which covers most ofthe ground.The Almirante Bay-Chiriqui Lagoon region apparently followed thepattern of the rest of the north coast of Panama in that permanentsettlements were established toward the headwaters of the rivers andnot along the coast.The immediate incentive for our visit to Almirante was an invita-tion from Dr. Gustav Engler, director of the Chiriqui Land CompanyHospital at that place. Dr. Engler has long been interested in theantiquities of Central America and has a small collection of artifactsfrom Costa Rica and Panama. 259 260 BUREAU OF AMERICAN ETHNOLOGY [Bull. 191As a guide, Dr. Engler secured for us the services of Teodor F.Machazek, a longtime resident of the region and a surveyor wellacquainted with the section. Through the courtesy of Mr. G. D.Munch, at that time superintendant of the Chiriqui Land Company,we were the guests of the company, enjoying the facilities of theirspacious and comfortable guesthouse during our entire stay. Weowe a particular debt also to Rev. Robert W. Turner III, who gen-erously gave us the use of his launch for our rather extensive survey.This courtesy was essential to the success of the work since almostaU travel in the region is by water.We are grateful to Mr. Edward G. Schumacher, artist for the Bureauof American Ethnology, for the line drawings in this report.Finally, we wish to thank Mr. and Mrs. T. W. Dunn, who ac-companied us to Boca del Drago. They gave unstintingly of theirtime and information collected regarding archeological sites duringa long residence in Central America while in the employ of the UnitedFruit Co. HISTORICAL BACKGROUNDInformation concerning the aboriginal inhabitants of AlmiranteBay and the Chiriqui Lagoon is very meager. However, there canbe little doubt that the Indians occupying the region at the beginningof the 16th century were the Guaymi, whose descendants stUl live inthe vicinity.Curiously enough, the most satisfactory descriptions are those ofthe Columbus expedition of 1502.The interpreters abducted by Columbus at Puerto Lim6n, CostaRica, could speak the language of the natives as far south as theCode del Norte River, an area which corresponds pretty well withthe hypothetical distribution of the Guaymi along this coast. Ferdi-nand Columbus, who gives the most complete account, does notdescribe the habitations of the natives encountered on Almirante Bayand the Chiriqui Lagoon, but he states that somewhat farther to theeastward they lived in single houses separated one from the other byconsiderable distances. He also states that their permanent habita-tions were not along the coast, but well up the rivers; an observationconfirmed by our archeological reconnaissance in the area.Oviedo, in describing the 1536 expedition of Felipe Gutierrez on thenorth coast of Veraguas, relates that their villages normally consistedof four or five large houses. This description would seem to correspondwith the archeological evidence at the Darkland site which we examinedon Almh'ante Bay.The scant archeological remains around Almirante Bay and theChiriqui Lagoon do not seem to bear out the rather considerable '^'^No!^^72f^^' ARCHEOLOGY, ALMIRANTE BAY, PANAMA—STIRLING 261population indicated by the Columbus accounts. It is probablethat the natives living up the rivers heard of the coming of the shipsof Columbus, the first to visit this coast, and descended to the bay forthe occasion, thus accounting for an abnormal concentration of people.Ferdinand Columbus tells us that as soon as the ships anchored inAlmirante Bay, which the natives called Carambaru, the boats wentto one of the islands where there were 20 canoes on the beach. Thenatives were described as being "as naked as the day they were born."One of these was wearing, hanging around his neck, a gold disk;another, a gold eagle.On the mainland the Spaniards were met by 10 canoes and saw con-siderably more gold worn in the form of ornaments. The menwere described as wearing narrow loincloths as their only garments,with their faces and bodies painted all over with designs in red, black,and white.It is interesting to note that the Guaymi today still use these samecolors for their face and body paintings, although now they haveadopted clothing.Pedro de Ledesma says that at one place 80 canoes gathered aroundthe ships, and that the natives wore gold ornaments consisting of "crowns," disks, eagles, frogs, tigers, etc. These, of course, were ofthe same type as the gold ornaments that have been found in suchabundance in graves in Chiriqui and Veraguas. Unfortunately, theSpaniards were not much interested in describing other aspects ofmaterial culture, although a number of stray items are mentioned.From Almirante Bay the ships of Columbus entered the ChiriquiLagoon, called Aburena by the Indians. Here they were met in amore hostile manner by even greater numbers of Indians, who blewhorns, beat drums, and created a tremendous din. This reminds usof the present-day noisemaking activities of the Guaymi during their "Balseria" ceremonies.At Guaiga, a river 12 leagues east of the Chiriqui Lagoon, theSpaniards again landed among a large concentration of Indians. Justbeyond here, at a place called Catiba, Ferdinand Columbus says:This was the first place in the Indies where they saw any sign of a structure,which was a great mass of wall or imagry, that to them seemed to be of lime andstone; the admiral ordered a piece of it to be brought away as a memorial ofthat antiquity.Since Columbus speaks of the structure as "an antiquity," itevidently was not constructed by the contemporary natives. Whenwe were on our reconnaissance of the Panama north coast, we madediligent inquiries of many natives who professed to know the regionwell but all denied seeing or hearing of such a structure. We areinclined to believe that the men with Columbus saw some natural 262 BUREAU OF AMERICAN ETHNOLOGY [Bull. 191formation, possibly a limestone dike, which they mistook for a wall.There is a great deal of limestone in this region, such as the cliff andridge where our cave site Boc-2, at the east end of Almirante Bay,is located.In 1951 when we were on the Rio Indio we heard from severalnatives of a "casa de laja" on one of the tributaries farther up theriver. It was described as a stone masonry ruin, definitely not ofSpanish construction. With the stone waU of Columbus in mind, wewent to considerable trouble to visit the ruin, eventually succeeding.The "casa de laja" turned out to be a curious natural tunnel about 50yards long through a horizontally stratified sandstone formation. Theonly evidence of occupation, past or present, was a huge cluster ofbats hanging from the ceiling.At some time before 1540 an Aztec colony was established in theSixaola Valley, on the mainland back of Almirante Bay. Before themiddle of the 17th century this group, known as the Sigua, had movedto Bastimentos Island in Almirante Bay, where they eventuallysettled in four towns. Here they maintained themselves with varyingdegrees of fortune until about the year 1760. The interesting historyof this band, the southernmost outpost of the Aztecs, has been out-lined by Lothrop.Rev. Ephraim Alphonse, who has lived among the Valiente Guaymlfor more than 25 years and speaks the language fluently, has foundsome interesting traditions that evidently refer to this group. TheValiente Guaymi still retain the term "Montezuma," meaning ruler.A tradition which is frequently recounted at the present time re-lates to the "Dekos," a group of conquerors who came from thenorth in large canoes having paddles studded with pearls. Theirleader was called Siri lOave. The Dekos are always pictured as beingmore civilized and smarter than the Guaymi. Some of the talesrefer to contests in wits between Guaymi and Dekos leaders. Inthese contests the Guaymi were always defeated, as they were inmilitary combat (Alphonse, 1956).We spent considerable time on Bastimentos in a futile search forremains that might indicate a settlement of the Aztec Colony, whichall the evidence seems to indicate was located here. Not only didour personal search result negatively, excepting for a few nondescriptsherds near the mouth of a small stream by the present village ofBastimentos, but the old inhabitants who had spent their lives farmingon the island insisted that they had seen no sherds nor other evidenceof occupation, except a metate and a mano (pi. 41). We visited thespots where these were found, but saw no evidence of a site.After the voyage of Columbus, the Almirante region acquired areputation for being rich in gold. It was frequently visited thereafter '^'^No!'^72r^^' ARCHEOLOGY, ALMIRANTE BAY, PANAMA—STIRLING 263by gold seekers and slavers with the result that native culture waspretty well broken down. There is little in the records of these moreor less transient visits that is of help to the archeologist.Modern accounts of the Guaymi are unsatisfactory, even for areasconsiderably removed from the Northwest Panama Coast.During the last quarter of the 19th century A. L. Pinart madesome observations on the Guaymi of the Chiriqui Lagoon region,and more recently they have been studied by Ephraim Alphonse.Practically all of the available early source material has beencollected and published by Lothrop (1950) in his excellent Veraguasreport. CEKAMIC STUDIESA considerable amount of time was spent in examining the temperof the various wares from the several sites herein described. Theonly instruments used in this connection were small hand lenses.Samples of the wares were shown to several professional archeologists.Their conclusions as to the nature of the tempering material differedso widely among themselves and with our own tentative determina-tions that we decided to follow the advice of Miss Anna O. Shepherdof the Ceramic Technological Laboratory, Carnegie Institution ofWashington, D.C., and describe them in general terms rather thanto guess at specific descriptions, a practice likely to lead to falseassumptions.All clays are originally derived from decomposition of igneousrocks and contain a certain amount of grit and fragments of un-altered rock. With the naked eye or a hand lens it is impossible toteU in most instances whether the grit is natural or hand ground.In general, the temper of the pottery from Almirante Bay is roundedsand mixed with angular particles.Similar difficulties were experienced in studying the nature ofslips. It was not always possible to distinguish between self slips,applied slips, and effects obtained by polishing. Here again we havetried to be not too categorical in our designations.Hardness was determined by scratch tests in accordance with theMohs scale.Of the pottery excavated in the three Almirante Bay sites, thesignificant pieces are illustrated and complete data are given with thedescriptions of each plate.Since the excavations were small, we felt that it would be pre-mature to give names to the pottery types encountered.Colors have been identified in the MunseU color system. Asevery ceramist knows, the colors of primitive pottery vary considerablybecause of uneven firing conditions, as well as other factors. The 264 BUREAU OF AMERICAN ETHNOLOGY [Bull. 191same pot may contain several colors, blending, for example, fromweak yellowish orange, to moderate orange, light brown, and black.In describing the pottery this fact has been taken into considerationand the best personal opinion given of each specimen illustrated inorder to convey to the reader a proper understanding of the wares.The laboratory studies were made under fluorescent light. Ininstances where the buff has definitely fired moderate orange, it is sodescribed. In Boc-1 and Boc-3, the buffs generally fired orange.In Boc-2, the buffs frequently fired brown.The equivalents of MunseU color symbols are given according toI.S.C.C. (Inter-Society Color Council) Standards. When a specimenis unique, the MunseU equivalent is given in the text (Judd andKelly, 1939). The following tabulation gives the general colorequivalents used in this paper:MunseU I.S.C.C.Light buff 1Y8.5/4.5 Pale yellowish orange.Buff 10YR7/4 Weak yellowish orange.10YR8/6 Weak yellowish orange.10YR6/4 Light yellowish brown.Dark buff 5YR5/6 Light brown.Dark brown 5YR2/2 Dusky brown.Orange 5YR6/8 Moderate orange.10R5/8 Moderate reddish orange.Red 6R5/8 Moderate red.10R4/8 Moderate reddish brown.Dusky red 5R3/4 Dusky red.Deep red 5R3/10 Deep red.SITE DESCRIPTIONSDARKLAND (BOC-1)In the southern part of the baj^, south of Cristobal Island, a longnarrow point known as Darkland projects into the bay. This hadbeen partially cleared for cultivation and the grazing of stock. Thepeninsula is formed by a rather flat-topped ridge which graduallyincreases in elevation from the point. Scattered along the top ofthis ridge were four large mortars made from unshaped stones (pi.42, b). Just above this area and about 300 yards from the point, wediscovered four midden mounds, each about 40 feet in diameter and5 feet in height; apparently each was the refuse mound for a singlehouse. Having no facilities for carrying much material at the time ofour visit, we made in one of these mounds a single small excavationabout 4 feet square and 2 feet deep. The mound proved to be rich insherds and other materials. Among the more abundant forms weresubglobular bowls, with incurving rims, of buff-colored ware with a *°No.°^72f*^' ARCHEOLOGY, ALMIRANTE BAY, PANAMA—STIRLING 265 coarse temper. This ware is characterized by a gray to black core, withmargins ranging from weak yellowish orange to moderate orange.These were decorated with elongate punctations in rows, betweenincised parallel lines. The designs are geometric and tend towardtriangular forms. Some of these bowls were further decorated withsmall nodes or bosses connected with incised lines or roughenedbands. One specimen was of much thicker ware and had an excurvaterim (pi. 28, a-i).Also abundant were sherds of large vessels with modeled appliquedesigns on the body. One had two leaping porpoises (pi. 29, a).Another was an octopus (pi. 28, I), while still another fragmentdisplayed what seems to be the tail of a fish (pi. 28, n).One curious sherd consists of what appears to be a thick beveledrim with rectangular openings cut below the rim. It is possible thatit is a part of a slotted pedestal base. If so, it is unusually thick andheavy (pi. 29, /; fig. 18, e).Other pieces were from large pots with slightly outcurving rims.Some of these have red paint on the lip, others on either the exterioror interior. The majority are polished light buff on the interior(pis. 28; 29, e, i).Excurvate, buff-colored rims with combed decoration on the interiorand smooth on the exterior were almost identical with similar piecesfrom the site at Boca del Drago (Boc-3). This is a buff-colored warewith coarse tempering material (pi. 28, k).One of the most individual wares is rather thin with finely groundtemper and fine line red painting on a light buff slip (pi. 28, m).Two sherds appear to be longitudinal sections of hoUow vesselsupports. The fact that one of these has horizontal red line paintingstrengthens this hypothesis since this is a normal type of decorationfor the typical conical base tripod vessel characteristic of Chiriqui, ex-amples of which were found in the nearby cave site (Boc-2).Several small sherds were of very thin fine paste ware with polishedsurface painted red on buff. Because of the small size of the sherds itwas impossible to deduce the vessel forms, but they were probably ofsmall size.Thick body sherds from large pots were quite abundant. These hada coarse sand temper and were usually fired buff on the exterior andblack on the interior. Some of these had combed decoration on theexterior, some were smooth. This is apparently the same ware as thatwith modeled animals on the body, and is analogous to the large urnsfrom Boc-3.One flat awl with sharp point and sharp edges, made apparentlyfrom the leg bone of a deer, was the only bone implement found. 266 BUREAU OF AMERICAN ETHNOLOGY [Bull. 191 Figure 18.—Sherds from Bocas del Toro, site 1. a, Large rim sherd with applique porpoises,smooth orange interior, pi. 29, a; b, buff punctate bowl, smooth rim, triangular design,pi. 28, a; c, buff interior, exterior blackened with charcoal, pi. 29, b; d, smooth buffexterior and interior, pi. 29, i; e, heavy slotted rim (or base }), pi. 29, /; /, light buffexterior and interior, red painted rim, pi. 29, d; g, smooth buff exterior and interior,punctate ware with incising, pi. 28, b; h, buff surfaces, interior of rim is combed, pi. 28,k; i, light buff exterior and interior with red painted lip, pi. 29, c; ;', rough buff punctateware with incising, pi. 28, c; k, smooth buff interior with red painted lip, pi. 28, o; /,smooth buff with inside lip painted red, pi. 29, k; m, smooth light buff interior with lippainted red, pi. 29, j; n, smooth light buff interior, charcoal on exterior, lip paintedred, pi. 29, e.Scattered throughout the excavation were a number of flmt flakes andmany coral branches.The above-described material, it should be remembered, came froma small test pit and represents at best a most incomplete sample. ^°No.°'72f^^' ARCHEOLOGY, ALMIRANTE BAY, PANAMA—STIRLING 267Boc-1 is by far the most promising site that we located on our srn-veyof Almirante Bay. The small mounds are rich in content and a fullscale excavation would without doubt give key results for this virtuallyunknown area. DARKLAND CAVE (BOC-2)On a high jungle-covered ridge about 2 miles southwest of the Dark-land peninsula there is a limestone cliff about 100 feet high. Alongthe base of this cliff are a number of caves containing the remains ofhuman burials and accompanying offerings. The native who guidedus to this spot had discovered the caves a few years before and saidthat he and his father had removed several complete pots, of which hewas unable to give an adequate description, and a stone metatecarved in the form of a jaguar.We selected the most promising looking of the caves and excavatedit completely. The cave, whose opening was about 20 feet across, ex-tended about 30 feet into the cliff, with a ceiling of irregular height.The last 20 feet of the cave was filled to within about a foot of theceiling and could not be entered until the fill was removed. The bulkof the artifacts recovered and a few fragments of human bone were inthe upper 12 inches of the deposit. No evidence of human occupationwas found below this surface level, although the fill averaged from 3 to4 feet in depth.Objects of stone consisted of a well-carved jaguar head broken froma metate of Costa Rica-Chiriqui type, and eight celts of mediocrefinish (pi. 43, g-^). There were none of the type with flaring edge socommon in Veraguas. One small triangular polished celt of a blueslatelike stone is 10 cm. long and 5 cm. wide at the blade. It is flaton one side and convex on the other. A chisel of limestone was 19.5cm. long. There was also a small polishing stone of bright red jasperabout 5 cm. in diameter (pi. 38, e,i, g). Scattered throughout thesurface layer were many large marine conchs, principally Stromhusgigas Linne and Charonia iritonis nobilis Conrad, and about 20 arti-facts of unknown use made from the spire of the conch with a circularhole 5 cm. in diameter in the center (pi. 38, a-d).Although five or six nearly complete vessels were recovered, thebulk of the pottery was in the form of scattered sherds, representing aconsiderable variety of forms and wares.Apparently, the great majority of the vessels originally had abuff-colored surface, although some examples were dark buff andbrown. In addition, various sherds are blackened by carbon, discolor-ing the surface as well as the paint.All the sherds are coated to a certain extent with a white depositcaused by the dust in the limestone cave, a condition making it 268 BUREAU OF AMERICAN ETHNOLOGY [Bull. 191difficult to decide the original color of the pottery. In some instances,the red paint has turned orange and in others, dusky brown; it isdifficult to decide the original hue. One example of this wide range inhue can be seen on the horizontal red line striping on the legs of thetripod vessel (pi. 30, b).There were five examples of tripod vessels, two of which are fairlycomplete. One of these consists of smooth buff ware well coated withlime on both interior and exterior (pi. 30, a). The outflaring rim is4.7 cm. wide. The interior is decorated with four groups of seven redperpendicular lines 0.3 cm. wide, spaced quadrilaterally. Eachgrouping is approximately 3.8 cm. wide. The edge of the rim is alsopainted red.The other fairly complete specimen has the rim missing, but thiswas probably a bowl-shaped form (pi. 30, b). The body is roughenedby horizontal striating. The legs are smooth, decorated with hori-zontal red line painting and three perpendicular slots 5.0 cm. long,0.5 cm. wide. Two slots are on either side of the leg, near the top;the other in the center front near the foot, which is broken off. Themost nearly complete leg has a modeled animal figure seated on theupper edge. The interior of the vessel is smooth and buff in color.Most of the exterior is blackened by carbon. This specimen is almostidentical with two figured by Wassen, 1949 (figs., 9, a; 10) found nearBoquete, Chiriqui.Of the remaining three examples, two were of the buff striated typeand the other, consisting of a conical base, is of thick buff ware. The REO BL A CK BUFFFigure 19.—Bowl with slotted pedestal base. Black core, dark buff margins; fine temper;base and interior of bowl polished dark buff; deep red slip on exterior of bowl on whichwere painted black perpendicular stripes extending from the rim to the juncture withthe base; see pi. 31,/. ^^Nj^T^r*^^' ARCHEOLOGY, ALMIRANTE BAY, PANAMA STIRLING 269 exterior is polished and partially dark brown from carbon. All of theseare of hard fine-tempered ware, with black core and buff margins.There was one small pedestal base bowl (pi. 31,/; fig. 19) with threerectangular horizontal slots in the base. The base and the interiorof the bowl are polished dark buff ware. The exterior of the bowlhas a deep red slip on which were painted black perpendicular stripesextending from the rim to the juncture with the base, 6.8 cm. long.Vestiges of eight stripes can be seen on the portion of the bowl whichwas found. Whether the stripes encircled the entire bowl or werepainted in groups cannot be ascertained. The ware has a fine temper,is rather thin, and is fired to a dark buff on the surfaces with a blackcore. This was the only pedestal base found.Frying-pan incensarios were represented by one specimen, a trans-verse piece broken off from the base of the handle. It is made from ahard fine-tempered ware with a buff polished surface. It is paintedred along the edges and has one transverse red line, 0.6 cm. wide(pi. 33, c). Similar but complete examples of this tj^pe were found byus at the Code site of Mojara in Herrera Province.Small plain subglobular bowls were represented by seven speci-mens, two of which were more or less complete (pi. 30, c, d). Theseare rather crudely made from a buff-colored ware. The surfacetypically is carelessly smoothed. Plate 30, c, has a body diameter of8.4 cm. and a body height of 8.0 cm. Plate 30, d, has a body diameterof 11.0 cm. and a body height of 7 cm.Two of the subglobular bowls were polished brown, one polisheddark buff, and another buff with a red lip. Most had small outflaringrims, some with rolled lips. Some were carbonized on the exteriorfrom having been on an open fire. The rim variations can best beseen from the illustrations (pi. 33, a, b, e, /). One variation consistedof a combed exterior with a crude rectangular pattern and smallnodes on the shoulder (pi. 31, b).There were eight examples of bowls with vertical strap handles.Judging from the fragments found, these had been low hemisphericalbowls with outflaring rims to which were attached a pair of opposingstrap handles (pi. 31, a, c, d, ^; fig. 21, e, f, g, b).Plate 31, a, is made of polished dark buff ware weathered brownin places. The lip of the rim is painted red and extends 1 cm. onto thetop of the strap handle. Directly in front of the strap handle on theinterior of the outflaring rim are six vertical red lines, 0.2 cm. wide, 3cm. long. The strap handle was modeled and stuck on the pot care-lessly. It is 3.5 cm. wide.Plate 31, c, is buff ware with a rim painted red extending 0.4 cm.down the exterior of the bowl and 0.4 cm. onto the strap handle,where it joins the rim. There is an irregular spot of red paint on the 270 BUREAU OF AMERICAN ETHNOLOGY [Bull. 191 CM. Figure 20.—Sherds from Bocas del Toro, site 2. a, Shallow bowl, pedestal base (?), redpainted lip, dark buff interior and exterior, pi. 34, g; b, buff ware with smooth interiorand exterior and red lip; c, buff ware, lip painted red with traces of red on interior; d,buff ware, red painted lip and convex curving collar, exterior carbon coated, pi. 36, b; e,shallow bowl or pedestal base (?), lip painted red, dark buff interior, red design oninterior;/, polished buff interior, exterior rough and carbonized, red lip, pi. 32, i; g, buffware with a rough exterior, smooth, interior, and fine temper, pi. 36, d; h, buff ware,carbonized black, polished interior, rough exterior, pi. 36, /; i, red painted lip, roughexterior and smooth buff interior; j, rough carbonized exterior and smooth buff interior;k, rough exterior and smooth buff interior.body of the vessel. The temper is coarse. The exterior of the vesseland the rim are smooth, while the interior is slightly rough. Thestrap handle is 2.7 cm. wide.Plate 31, d, is buff ware weathered brown. The rim probablyhad been painted red. There appears to have been a strip of pottery ^°No.'*72r^^' ARCHEOLOGY, ALMIRANTE BAY, PANAMA—STIRLING 271 CM.Figure 21.—Sherds from Bocas del Toro, site 2. a. Buff ware, exterior blackened bycarbon, had strap handle, incised design, roughened zones, bosses, smooth interior, pi.31, e; b, buff with loop handle, covered with limestone dust inside and out, pi. 31, g; thinbuff tripod jar, carbonized exterior and smooth interior; d, polished dark buff interiorand exterior, pi. 33, b; e, polished dark buff, lip painted red and vertical lines on interiorof rim, strap handle, pi. 31, a;f, buff interior and exterior with red rim, pi. 31, c; g, buffweathered brown with smooth interior and exterior and strap handle, pi. 31, d; h, buff,polished interior, exterior of rim roughened, pi. 33, /; i, buff weathered brown withpolished interior, pi. 33, a.appliqued on top of the strap handle as decoration. The handleis 1.7 cm. wide.One specimen has a plain incurving rim. This had a roughenedzone below the rim, decorated with incised lines connecting smallhemispherical nodes. Another roughened zone runs around the middleof the bowl. The edge of the rim is painted dark red (pi. 31, e; fig.21, a).682-611—64 ^22 272 BUREAU OF AMERICAN ETHNOLOGY [Bull. 191 CM. Figure 22.Fori egend, see opposite page.) ^°No."^'2p^' ARCHEOLOGY, ALMIRANTE BAY, PANAMA—STIRLING 273Plate 31, g, shows buff ware with a rolled everted run and a loop handle.The most abundant form consisted of fragments of 17 large sub-spherical pots with medium outfiaring rims. Of these, three weresufficiently complete to give an idea of the body shape. There isconsiderable individual variation in the rim forms (pi. 34, a, b, c;fig. 22, a, h, e).One broken, but almost complete, example had the entire exteriorsurface roughened as though by rolling it with a corncob or a roughfabric. The interior is well smoothed. The lip is painted red.A series of rather undefined vertical ridges runs around the middleof the body as an additional decorative feature, and a series of indenta-tions runs around the base of the collar (pi. 34, a; fig. 22, a).The remaining examples are carelessly smoothed on the exteriorand carefully smoothed on the interior. Some had red-painted lipsand some were coated with carbon from being over an open fire.The ware is coarse, fired to a black core, with moderately reddishorange margins and dark-buff surfaces. Most have plain round lips,but on two examples the lips are beveled.There were seven fragments (six illustrated) from large vesselswith high collars (pi. 35; fig. 22, c, /). The collars are roughened onthe exterior by combing. One has a smooth zone below the pointwhere the collar joins the body; below this the combing begins again(pi. 35, d). In all examples the interior is carefully smoothed. Twohad a red-painted lip. On the rim, one sherd has part of an appliquedoctopus, suggestive of similar designs from Boc-3. This sherd hasa dark-buff polished interior and a red-painted lip. The exterior israther rough (pi. 35, a, fig. 22, d).Large shallow bowls were represented by two specimens. Theseare similar in form to the pedestal base "fruteras" from Code andVeraguas, but probably did not have such bases since none were foundin Boc-2. One of the specimens has a beveled lip, painted red, and Figure 22.—Sherds from Bocas del Toro, site 2. a. Buff ware, red painted lip, smoothinterior, roughened exterior, pi. 34, a; b, buff ware, red lip, exterior rough and somewhatblackened, pi. 34, c; c, buff ware, smooth interior, exterior of collar combed, smooth onexterior below collar, pi. 35, c; d, dark buff polished interior, exterior blackened, rough,applique design, lip painted red, pi. 35, a; e, dark buff ware, roughly combed collar,smooth interior, pi. 34, h; f, polished dark buff interior, combed exterior, pi. 35, b; g,polished buff ware, fine combing on collar, smooth below collar, deep incised grooveseparating collar from body, pi. 9, h; h, smooth buff slip on interior, exterior rough, pi. 36,i; i, smooth buff interior, exterior rough, deep groove separates collar from body, redlip, pi. 36, k; j, polished buff on interior, roughened exterior, red lip, pi. 36, • t" '* tV~ - -"T,-^ .' ' --'* .u, -«v •"' •'^ Midden mound and mortars, Darkland Point.(For explanation, see p. 283.) BUREAU OF AMERICAN ETHNOLOGY BULLETIN 191 PLATE 43 ^\ IStone artifacts from Escudo Island, Jungle Point, and Darkland Cave.(For explanation, see p. 284.) BUREAU OF AMERICAN ETHNOLOGY BULLETIN 191 PLATE 44 Jaguar heads and double-bitted axes.(For explanation, see p. 284.)