rae sieht re r 3 x an > » ® , { ® ° , t 7 by _ SMITHSONIAN MISCELLANEOUS COLLECTIONS VOLUME 113 (WHOLE VOLUME) ‘ARCHEOLOGY OF’ THE FLORIDA GUILF ‘COAST (Wit 60 Priates) BY ; GORDON R. WILLEY Bureau of American Ethnology Smithsonian Institution (Pustication 3988) CITY OF WASHINGTON PUBLISHED BY THE SMITHSONIAN INSTITUTION DECEMBER 29, 1949 rit NM LA cps > ob SMITHSONIAN MISCELLANEOUS COLLECTIONS VOLUME 113 (WHOLE VOLUME) PacliEOLOGY OF THE FRORIDA GULF COAST (WitTH 60 PLAtEs) BY GORDON R. WILLEY Bureau of American Ethnology Smithsonian Institution (PuBLICATION 3988) Re’ Votanat Mush gy Core Ee oe" CITY OF WASHINGTON PUBLISHED BY THE SMITHSONIAN INSTITUTION DECEMBER 29, 1949 v4 The Lord Waltimore Press BALTIMORE, MD., U. 8. A. CONTENTS PAGE TED Ete aay EONS Bed eee ie ee oS CIO RY Ree ag ee xix iPrepataon alid Writings .r. .c.-c\cecn bem RR ne eees eve secdss es xix PA CHUIOW IEC SINIEMESs che «itis 6.6 ds 0:2 on siecee CO ae Saeed ees xx LEAPY SIG) tary accycr| ic 010) | eat a nee cs Ab 3 2 ct Nee ee ee XXiil AES CE PIONE omicVe ie mia eiavele is rey: we: 600s 4 eco ng TA HUE o cists I GHOSE AOE. SHUR oR ere om a <2 UNE RCE In eae Rete Ce dew'd I Definition ‘of the Florida.Gulf Coast area. ........ctohs.s.oees 2 PROCCGNCES 1 OF CIASSINCATIONS <...«.. \ura sae RoR eis ce ee eid ws 3 fae formulation, ot culture’ periodsa,ascemens ieee a eee ane: 3 Bote ery elassincatian, 5< )s:.:. cv ge oe aaa eo te sic as 5 MESeriplive: Classiications, ... 0:1 Cc ice eae Pees «skin 7 J) SPAS s@ Celia 14 ce a Vg 2 Hn a Si ae Ua oe a 9 General (COATACteTIZattOn, 2. \.ic\s Laden octamer Cee Moe Pia ase 9 Geolozy, soils, vegetation, and. climate seis. 25) ek. es dec coe. 9 MEN TOMAMOLETIPIALIEIES ogc 6 osc. v win ersievs tute Somme Re te A a Sib eR os is wis 12 II. History of archeological research: 1846-1946.............0..0000 15 BripleritivestioatiGliSs ' co cisia «ace «1s ceca ree hie eer k? aos erg => sors 15 mye surveys or Clarence, B.: Mooresieace. tama kxtex O28 re Oreo sit« 21 She) classificatory studies of W. Hi Holmess. 6... 2 co... ees en<- se 26 he problem, of Early man in Gulf Florida..... 660.24 ho sce en's 29 SOGMCHE ¢ VOLE. ccc aici. 2.6 00s hus a Seatuie er RIS CIE Es a Os es wis 30 The ethnohistorical studies of John R. Swanton.................. 34 III. Excavations on the northwest coast: 1040...........cceeeecceeees 37 Des TESA ER GTEC ees 0 7 Lon ous & as sane Soava ree reremTeT eM RcIe ate tetaleee Etat Miao ake 37 Garrapelle, Franklin ‘County «CP r-2)ic< veccslena ssc). 22 celtic nase. 38 Weseription: OF the SitGs\ 5. . cv c/ed cence etn maton a aa teleteie s 38 EEX CAVALIONS Mey ic teecsrcis clcig cio 6: chand OPCTOR Rene tees mise aatt ote cay Sie aie 42 SES ELIONU I PIRY, vig er soi Sic v cvs, ove Fisca. Seleomtoce, seolsratare ee eta eteah oral NE Seals a 47 Mound Field, Wakulla County (Wa-8)ies seen ees ods ee wee 55 BPESETBUIOH OL w Te) SILC ss a/s-<)s1e cae ereiaieiatate sisielon sc AIR si, s's 55 Bosca ttONIR EG oc ot iiece'e nos. o oy one RRR tert cc ete ts ete 56 Seas ley gz 0 an eRe eo eR RP ge a hea ee 60 Bawelly say “Goumty « (By=3)). nce este ate nten e «wie ecto erdets es lns 64 Darerapiion (of the (Site: «). +.4s\ue «coe ye since elniolss viele Cleteies = se 64 Bececarya tions bare yy sitar telcos els See eee clas sieeert cictae te ciers cc 66 RI ReaieCi PLE ett ee cies 2s 5. ae eee Rea ts hia sheets as 67 Fort Walton, Okaloosa County (Ok-6)...............cceeeeeees 72 Descrinhan ier ihe SiO. . .|.c/ccea se aeieces we sun as molG aWEING oe 2 0s 72 ECA VATIONS ie ies reece cclo's 6 “Sarre eMC e tein cae a taretere pens Ens arait, ose 75 RETA TS RT AIS an Gh oo aaa: e « 25:0 win = MORO RNER. adie Sh tee Seidl 5 ae Saye 77 Golf. Breeze, Santa Rosa County ((Sa-8)dt ics. cciccicaemwnces see 89 Wescrinerdeh Of fhe SUES a5. .c0oy GN ent A oh vues (os uae wee ss qeetecers lees eee eee eee eee 113 EXCAVATIONS’ . bin b:avisibis bi si sero svete oR Oe taloioie mate Glo lore eee arene 114 Mound! stricture <)52.ceeceaacsoe cee cae osm cone tle eee es eee II5 BS EA ANS ag oro- sa, o. Scale: & a ales on; Sys wialins eal oleae. RIOTS PRG Ene 116 Physical: anthropology \...2004..s6. sleckied wae 2s. seietek eee eee 117 POUEERY ya id e 1oiase: 5 a fo.o me ass wlevecchace’@ bata e elesdienis a os he Oe else CT eee 117 PTAIEACES © sere iw na eid a nie ete Solel aie deka oy eee ee ee 121 SUMMA Py} eie:sieraere. sas dois sciase areiereniok ole on MERE ROE eee 12 Englewood, Sarasota County ((S0-1) <.< sc .2 es sellocteee eee 126 Deseription of the ‘sites. 4/56... aie la ss siete g ode cicero eee uW0: EExCavatiOns: isis. o-0:é-ceie setae oie lw eOOiielaeeime eee ine ee eee 126 Mound “structure i... Joc. gaa lnce owe vase e oe Eee 127 Burials: 3. 5do iiwdisietise eee oe Se eee oe ee Ree eee 130 POLECEY:. c:e:asisla's 0-6) sieie sida a wie dale AG alee ei bieleaiee lade e cee eee 131 VAPLITACES Wesiays. ccc valoooard igs Rind ynieienorsitte alate Glen ae ee ee 134 SSCA y,. as a: aes odie +.» Shave Slevmuatn eo era ovstiode etelere aletalore = ele) eee 134 Safety Harbor; Pinellas ‘County (Pi-2) =. 5 oh eites tele itae nanan 146 Parrish mound 2, Manatee County (Ma-2)..................... 146 Description of the. Site... sishsoreiert:«sjeiesmie sisreie sa erie nis «nee 200 Lop ‘pencils marks (Bs-2) 6 ooo Ge te . cc: Apet anion, Jae eee 200 mite 5 miles west of Navarre’ (Es-3) «2/5. .....thsaseee een 201 First site’opposite Woodlawn, (Es-4))..< sc Yeu sckeecemnee seem 202 Second’ site opposite. Woodlawn (Es-5).........:0 sane aremewiaie siete asta ete 264 Other sites in Liberty County..00055. rire aie ae cine cite eee olen aie ee 332 Boca Cieca Island (CPirGys.. eee eee ee eel rr aie te nee 333 Bayview? (CRIl7)\ co iiad sate etnaots 0 ane oe clears picuels[ahs 0c ieee 333 Seven. Oaks (G@P1-8) 522082 Sea ae Sone eee 334 Other: isites-an {Pinellas Contity.c0. 0.0. .0- +5: yee eee 335 Hillsborough: (Courity...<.ce6 ccc. oe 6 oes sci een oe 335 "Dhomas (CENT) 2. bo baceta rac s cine as os ore saree ae oe 335 Gockctoach: (CHiR-2)) So Fria le owe noe ec e aetna ee 335 Piclatick \@H129) - 2 she hactecs sos os es wok oes orca c eee 335 Jones )CHIR4) oo, hrc Se Cah Sono aro tse eign 337 Snavely «(Hi-§)) oA. diet oeuc ) occa eae ss oe ee aca eae ee 337 Buck ‘stand: (CHi-rG) 00. ee ee ee eee ee rere cin 337 Roeky “Point “(Hi-7 eos ae eas ae Oe tie 338 Other sites in Hillsborough County... .---...:.- 1 -- eee 339 Manatee Connity® oc iss 2 ole ose a cleasiew ects se 0) © opiate ra 340 Parrish mounds 1-5 (Ma-t, 2, 3, 4, 5)------2++esseeeresensces 340 Perico Island ((Ma=6)§ vice. css tors 1c steis ls siete + 9 oe sara a 340 Shaws: Paint. CMa )iiscd «oie eee epeunrsyelele oere ns cele) ies ie ee 340 ©ther sites in) Manatee County... ccmeme 25 reeiein reneneerenste 342 Sarasota“ COUMLY. seni cele io0 ele eieieinieys wie alae ome 3 piaya erika 342 Bnglewood (S0-1 )i sti-u6, cosh er oe oalte yy aes 5 og 342 Osprey: (S022). eaesitcc ot cies oeeye ie m ee cisinle see aia coe 342 Pool Hammock (80-3) p50 3 seen oe 343 Whittaker or Whitfield estate (S0-4)..........2. cc cece eeeeeee 344 Trise. Site, CS0-5) sires sche Salas orposneranrels cafe sin pi siegl ci rae 344 Other sites in Sarasota County. (cco. .e.50-- +e) 6 seeeeeee 344 Charlotte Gounty .o5)< wacccaexs sais sce ts uae 02s oki oe ne 344 Cayo Pal (Chat) isd sas. oie, cite ore op Wepaterniee ose ois, oe 344 Gasparilla Sound (Ch=2), 2). 0. 361 PP ANOTER Gren a. thee s asciava 2, a.51.4 6 BREE RIOR este Coe RICE 362 erat zation. Of \SOCIEEY a... 75 0.01 nc RS aca aU eee eRe eo eec es 363 Dsposalmors they dead acc c.as sicssmce so Se Re eee 363 SUR ALG SANT S ota ois wis kracarsiols: Crs a atv-at va hae Rae oO aah 363 Oihecc arts and: technologies: a/sicgcGuiecls Wide fees eed ae twee 8 365 Speculations on population and period duration................ 366 tEhe- Santa. Rosa-Switt. Creek. Period!s foe s.necas Dok Benes bes 306 I CLIOULGEHNNTIONy «orc crete wicks unk a eee tee eee eres cies 366 EG EN ERE er eh oats a rg ave, Wk; a») ae aco ado alee RTI RR ne oe ee 367 PIPIPICHIORE “PALE sic:0,3,0:s = tis s's(e! s «snd 9 ARIS ee Rie aes 368 PE RERTENOAR Y= ses oa coins di s'w'n:p:B%6 eis cn, c.ohoiinrehnxuxe: wcayerb lal he Rae cre 368 iE GAUIZatlGtt OE -SOCICLY «<5 « . 2c s1sudcieeudh vee ate a eRe 368 DEE POSAI COL EMC LGEAG 2's snare a 004.5 6 = ad hiahe genio eee eta fee 369 CORLISS Ty Ee AR ee aera en Serie et ees 371 Giher atts and technolovies ..<..<. oi 0 e-snshamenincaeeesicken secs 303 Speculations on population and period duration................ 395 me” WV eeiien NGlatil PeriOGS. «ais : :\s cs soncnaae se ene eee ne 424 44. Weeden Island Period rim forms, Carrabelle Incised............ 424 45. Weeden Island Period rim forms, Carrabelle Punctated......... 426 masevy ecden Jslatit Period rim f6fis’ «....02 5 cce=sene tov es etn aeeels 427 47. Weeden Island Period rim forms, Keith Incised................. 428 Be ma Vected Dstt, PCTIOd) VESSELS con xc. Pits I and JI, Mound’ Field (Wa-S)ici 616.25 2's cease cceiies 1, datas JC eral TE Soyoil ((eiiee) Migcenoc6oa00 dacs JunopOoUDOOINWon ar ~ LEGS UMN erate IA, Sioa (CBM Sar codconcanco5GsbounsDOCCUaNOC , IDE I mode WWeliterr (ON Aeseacnocoo such nooeognuneDUnGOO0E Weis tioanderll Hore Walton COG) ta sncites sissies se cces , IBAA Narre Welt (OKO) Sonauoocns baucccoudggudnAdUuBGodKc BETES Vg ONT VV ALON, CO OK=O) 6 5s ice seca mate eile efile tehislalelua givd'e’o.s ate PPE VERN Teakiont., VWalton CONK-0): oars diercucsae nie cena ea im 3) ohetaissicierare Mae 1d Ort VV alton (K=O), .« 2.0 = casas meme ere mic analy: eta feiciaverang PePitspleandelles Gules reeze. (Saco) ice ceil eieietee delete laters leraneiet= Merits Ghbana eve Gule Breeze” (Sa-8)\..20: 20d gs seree ba acs os 00s mPits) Veande VlenGult Breeze (Sa-c))icje eccrine len eters eras rns take Jackson: ((Mhe-1')\sicg sctcaers see een tans bam wers te 3 re oe XVii om > ot ; o- , mire ror a — & ant vane tee get 7 ving » Cl a? 140 ak ne rays — ane ane ee ths sonigt dake ae a | ae ie Jt oll?) ai) es ae | ¥ 9 a Tea" : ’ Ses é ; ahs 7 vas yt poke eal Oh pared Pain va" “Shea aw - y Ay i O's Ati Rig if : wr " 2 ie bai Pair ah ji fie ’ ant ‘et aaa Oy tt! ae Lee) Quyatll™ heey) ry nid “7% 3 Pa. ae week Can) at fg , 1 $2 , i - 4 * PREFACE PREPARATION AND WRITING Interest in the archeology of the Florida Gulf Coast began some 10 years ago while I was serving as an archeological field and labo- ratory assistant to Dr. A. R. Kelly at the Ocmulgee site near Macon, Ga. In the spring of 1938 I had occasion to make a short survey trip to Panama City, Fla. The midden sites in the vicinity of St. Andrews Bay attracted my attention, and I was struck by the similarity of much of the stamped pottery to that from Central Georgia. I was also interested to see the frequent associations of this stamped material with the distinctive incised and punctated ware of the Florida Gulf. As it happened, the previous year I had visited the ceramic laboratory at Baton Rouge, La., and examined collections there from southern Louisiana and Mississippi. Much of this Lower Mississippi Valley pottery bore a close resemblance to these incised and punctated types from the Florida Gulf that were found in the same sites with the Georgia-like complicated stamped wares. Here, on the Florida Gulf, were the means of relating the Louisiana and central Georgia culture sequences, in 1938 the only ones known for the lower Southeast. After 2 years at Macon I spent another in Louisiana, with J. A. Ford, on the Louisiana State archeological survey. During this year I was able to familiarize myself with Lower Mississippi Valley archeology as it was first being worked out. A year later, after my first year of graduate work at Columbia University, Dr. W. D. Strong arranged to send me for a summer season of work in Florida in 1940. I was accompanied by R. B. Woodbury, another graduate student, and we were aided in our survey by the National Park Service. The course of operations and the results of this 1940 survey are given in detail in the body of this report. In the ensuing academic year at Columbia my interest in the Florida Gulf archeological problems was broadened considerably by further study of the literature and museum collections in the East. In visits to Washington during the fall of 1940 I discussed Florida archeology with Dr. M. W. Stirling, Chief of the Bureau of American Ethnology, who had directed a series of Federal Relief-sponsored excavations there in the mid-thirties. Specimens and notes from this and earlier work of Stirling’s had never been xix xx PREFACE studied or published. It was at his suggestion and order that all these data were turned over to me with the idea in mind that I should incorporate this material with my own in an over-all Gulf Coast report. In response to this generous offer I formulated a plan to do such a general paper and applied for a research grant for the academic year 1941-42. A Julius Rosenwald Fund Fellow- ship was awarded me for this project in April of 1941, but, owing to other commitments that had developed in the interim, I was unable to proceed as planned. On this account, the directors of the Rosenwald Fund very kindly set aside my award for the succeed- ing year. Subsequently, in the summer of 1942, I began preparation of the present monograph. The study of collections in Washington and perhaps a third of the writing were complete when I was forced to lay the project aside again. This time the interval was a 4-year one, and it was not until 1947 that I could resume the task. From then until the present, with only one major interruption, I have been able to devote most of my research time to finishing the job. These delays have held many disadvantages for the prosecution of a work of this type. It is always burdensome to pick up the con- tinuity of thought and effort on a particular project after it has been dropped and forgotten for weeks or months, to say nothing of a few years. The many ramifications and details of a lengthy syn- thesis are easily lost or slighted when data are allowed to become “cold.” With each renewed attack it is always necessary to go over ground once covered to achieve again the proper frame of mind. But, in some ways, the delays have helped. Recent advances made by others in the Southeast and in Florida have aided in the under- standing of many of the Gulf Coast problems, and more useful data have been brought to my attention in the intervening years that I would have missed if the manuscript had been completed in 1942 or 1943. ACKNOWLEDGMENTS Those to whom I owe thanks for help and encouragement while preparing and writing this report are legion. Foremost, I wish to thank the Trustees of the Julius Rosenwald Fund for their generous grant without which I would never have initiated the study. Equally, I am deeply appreciative of the aid and encouragement given to me by the Bureau of American Ethnology, Smithsonian Institution, and by the Department of Anthropology, Columbia University. Dr. Matthew W. Stirling, Chief of the Bureau, has not only given freely PREFACE XXi of his help and advice on matters pertaining to Florida archeology, but he has, as mentioned above, turned over to me for my use the notes and collections of his own field work. At Columbia University, Dr. W. D. Strong and Dr. Ralph Linton gave their whole-hearted support to the 1940 survey work and permitted me to continue the research while acting as instructor in anthropology at Columbia during the year 1942-43. Other institutions and individuals who have assisted me in the field or in the study of collections include: the National Park Ser- vice of the United States Department of the Interior and the former director of its Archaeologic Sites Division, Dr. A. R. Kelly; the National Museum of the Smithsonian Institution and its archeologi- cal staff, F. M. Setzler, Neil M. Judd, Dr. Waldo R. Wedel, and J. R. Caldwell; the Peabody Museum, Harvard University, and its present and past directors, Dr. J. O. Brew and Donald Scott; the Peabody Museum, Yale University, and Dr. Cornelius Osgood, Dr. Irving Rouse, and Dr. John M. Goggin; the R. S. Peabody Foundation, Andover, and Frederick Johnson; the Museum of the American Indian, Heye Foundation, and George Heye, E. K. Bur- nett, and Charles Turbyfill; the Florida State Museum, Gainesville, Fla., and its acting director, Nile C. Schaffer; and T. M. N. Lewis. To the field archeologists who conducted the investigations under the Bureau of American Ethnology-State of Florida and Federal Relief sponsorship, I am indebted for the use of their notes and for other information which they have generously offered to me. The section on “Excavations on the West Coast: 1923-1936” is based upon these data; however, I absolve them from any interpreta- tions which I have placed upon their recorded or verbal information. Preston Holder was the archeologist at the Thomas Mound (Hi-1) and Cockroach Key (Hi-2), and J. Clarence Simpson also was in charge of part of the work at the first-named site. Dr. Marshall T. Newman excavated the Englewood (So-1) and Perico Island (Ma-6) sites; and D. Lloyd Reichard was the field supervisor at Parrish Mounds 1-5 (Ma-1-5). Of my archeological colleagues in the Southeast, I am, above all others, in the debt of Dr. John M. Goggin. Since 1944 we have worked together on Florida problems, and his constant counsel and wide knowledge of the State as a whole have been of great benefit to me. The instances of his help are beyond enumeration. Sec- ondly, sincere thanks go to my friend, Dr. R. B. Woodbury, assis- tant during the summer survey of 1940. Only with his energetic XxXil PREFACE and intelligent help could we have accomplished all that we did in that season. In addition, Woodbury prepared a descriptive classifica- tion of all the stonework which was found during the 1940 trip, and this unpublished classification has been of value in the preparation of the present report. For perhaps less specific aid and stimulation I am grateful to a host of other colleagues and friends, particularly Dr. A. R. Kelly, Dr. J. A. Ford, Dr. James B. Griihn, Dr ae Waring, Jr., Dr. Irving Rouse, Dr. Philip Phillips, Paul M. Cooper, Dr. Marshall T. Newman, Charles H. Fairbanks, John W. Griffin, R. P. Bullen, and Hale G. Smith. I am indebted to Drs. Goggin and James B. Griffin, as well as Dr. Irving Rouse, for making available to me their extensive bibli- ographies on Florida archeology. Special assistance in the identification of marine shells has been given by Dr. J. C. Armstrong, American Museum of Natural History, Prof. Robert Webb, University of Tampa, and Dr. Harald Rehder, National Museum. Mammalian bones were identified by Dr. George Goodwin; fish remains by Dr. J. T. Nichols; reptiles by Dr. C. M. Bogert; and birds by Dr. J. T. Zimmer. All these experts are on the staff of the American Museum of Natural His- tory. Dr. T. Dale Stewart, U. 5. National Museum, identified and commented upon a human bone pierced by a shark tcoth. In Florida, I wish to thank all those persons who expedited our work in 1940, especially Alexander Key, of Apalachicola, J. C. Simpson and Dr. M. F. Boyd, of Tallahassee, Julian Yonge, of Pensacola, and Prof. R. F. Bellamy, of Florida State University at Tallahassee. I should also like to mention the name of Montague Tallant, of Bradentown, who provided me with notes and photo- graphs of sites visited and excavated by him in Dixie and Levy Counties. Finally, a note of appreciation is due the various land- owners of northwest Florida and central west Florida who per- mitted excavations to be carried out on their property. Photography of specimens shown in the report was done in part by the Smithsonian Photographic Laboratory and in part by myself with the help of Paul M. Cooper. Most of the stipple drawings of specimens are the work of Jack Anglim and John Adams. Line drawings, maps, and charts were largely made by Edward G. Schumacher, although a few were done by E. G. Cassedy. And finally, for her unfailing good spirits and constant help in the reading and re-reading of this manuscript, I am deeply grateful to my wife, Katharine W. Willey. PREFACE XXili PLAN OF PRESENTATION The preliminary sections of this report are devoted to method- ological considerations, to the natural environment of the area with which we are concerned, and to a review and critique of past archeological research in Gulf Florida. The presentation of the essential data follows this and is contained in three sections, dealing, respectively, with stratigraphic excavations in northwest Florida, with mound excavations in central west Florida, and with a review and analysis of virtually all recorded archeological sites in the total Florida Gulf area. The basic chronology and the characterization of the successive culture periods which follow are derived from this presentation. A summary of the ethnohistory of the area is placed after this exposition. Conclusions consist of a tracing out of cul- tural continuities in the Gulf Coast area, a discussion of extra- areal contacts and affiliations of the area, and a final summary reconstruction. Gorpon R. WILLEy. Bureau of American Ethnology, Smithsonian Institution, September 14, 1948. Since this manuscript was originally submitted I have had the benefit of several new findings and field researches in the Central Gulf Coast region. For these I am largely indebted to Dr. John M. Goggin, John W. Griffin, and R. P. Bullen. A number of their suggestions and data have been incorporated in the text or as foot- notes. Also, my colleague Dr. Marshall T. Newman has written the section “Peoples” in the chapter on “Conclusions.” I am most grateful for his collaboration. G: RW. June 10, 1949. An j tm a iO a iT 1 # « } u nae i god ih § j ’ . x ; “ i ee ot F \ oe eA % 7 *~ be - : ay Waayeti i Lr al cal i ont ay Fy) 4 wally ware ee fea .J uD Rae PP get (ai Oh, COR ua a es Mia Vers aoe Epes mn - ay Ay can ae 7 ; mY : co) OP aren te aoe ] , ae ¢ wi 00) Se re ae ee i] @ oo mr =|, Ae 7 ae ~ as) ¢ a); A “ youre io = 3 ' s 4 By - ; ' ARCHEOLOGY OF THE FLORIDA GULF COAST By GORDON R. WILLEY Bureau of American Ethnology, Smithsonian Institution (Wit 60 PLATES) INTRODUCTION PURPOSE OF STUDY The purpose of this report is to provide a complete and integrated synthesis of the archeology of the Florida Gulf Coast to the present date. It is both a field report of unpublished excavations and a compilation and reinterpretation of previous archeological work in the area. I have attempted to show how recent archeological investi- gations along this coast give the clues which enable us to incorporate the results of almost a century of earlier digging into such an over- all reconstruction. This by no means implies that I pretend to have written the “final word.” As in most other parts of the southeastern United States and the Americas, the archeology of Gulf Florida is in its initial stages of synthesis. If this work serves as a general stock-taking and as a base of departure for future researches it will have fulfilled its purpose. The orientation of the report is historical reconstructive. Spe- cifically, it is a reconstruction of the skeletonized time-space system- atics of prehistory. Within the Gulf Florida area I have attempted to show the succession of cultures and culture growth throughout several centuries of past time. In conclusion, I have also tried to indicate something of the contacts and interplay between these prehistoric Gulf Florida cultures and those of neighboring areas. The spatial and temporal arrangement of cultural forms or types which is the framework for such a reconstruction submits to demon- stration and is, I think, reasonably well proved. Beyond this realm of the demonstrable, several lines of speculation are pursued. One is also historical and concerns origins and diffusions; the others relate to population density, time duration, and cultural trends. The nature of each, as it is followed, marks it as speculative and exploratory as opposed to evidential and established. In offering these speculations I have tried to explain the factors which have guided my choice of interpretations. SMITHSONIAN MISCELLANEOUS COLLECTIONS, Vol. 113 (Whole Volume) 2 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I1I3 DEFINITION OF THE FLORIDA GULF COAST AREA The Florida Gulf Coast archeological area is comprised of a coastal strip which extends from Perdido Bay on the northwest to Charlotte Harbor on the southeast (map 1). The inland extent of the area varies from 20 to about 100 miles. All the counties of northwest Florida are included as is a bordering strip of south Ala- bama and southwest Georgia. Southward along the peninsula the area is made up almost entirely of the coastal counties. Ecologically the area is valid as it is all a part of the Gulf Coastal Plain. There are some physiographic, soil, vegetational, and cli- matic variations; but, relatively speaking, these present no striking differences. The area also has an adequate cultural validity. Thomas, in his mound surveys, included all of Florida in a greater lower southeastern or Gulf area (see Kroeber, 1939, pp. 102 ff., for a simplification of the Thomas groupings). Holmes (1914, pp. 420- 424) placed all of Florida and adjacent south Georgia in a Georgia- Florida area, pointing to over-all similarities in pottery types, stone artifacts, and burial mounds. Wissler (1938, p. 268) includes Florida along with Georgia and the Carolinas in a South Atlantic archeological area. All these areal classifications are similar in that Florida is treated as a whole, or grouped with portions of other States, to create a large area of which the Florida Gulf Coast is but a part. The narrowing of archeological culture areas and the in- creased number of subdivisions in those previously projected has fol- lowed as a natural result of additional research and increased knowl- edge, particularly with reference to culture sequences. M. W. Stirling (1936) was the first to subdivide the greater Florida area and to delineate a Gulf Coast area. Stirling’s Gulf Coast area was defined as: “. .. the area of Florida draining into the Gulf of Mexico, as far south as the Caloosahatchee River.’’ With minor revisions and quali- fications this is the Gulf Coast area as it is accepted today. Goggin’s recent (1947b) definition of Florida culture areas has modified the southern boundary, bringing it north to Charlotte Harbor. Goggin has also drawn more definite eastern boundaries separating the Gulf Coast area from those of the central and eastern parts of the pen- insula. The area as it is defined in this report and presented on map I follows the Goggin revisions. A degree of arbitrariness in the definition of the area is not de- nied. The realities of natural environment and prehistoric cultural distributions have been modified or approximated to suit our pur- pose and convenience. The southern boundary is quite definite and 7 af. : il ; : i ‘ ” » ” . + ts a ‘ - 7 ‘ “ cy -] ad v E\VIAG Ag . u & x H , H H x fern ; Q ' Ls 4 ' & ' AG H t z ' 1 t 3 t ° J N . eat q ' t : / Q ‘ Q 1 9 arnt i * weeee eee -d---- 2 dee rf 1 aout 1 ' Cie | ° S 1 ' \ 1 3 Vege a w g | H ae: ' ' .¢) eh | Ss ! i] w q SOS ms Sn edt ateam ios ' t $s! : ' SY 6 irk 7 | 3 Lay ce ° H | : ; ' ) 0 cc Mane == St ne Fa Wel (oar gata is he oes ' ' x ' . ' ehane = ' 5 reg es SSeS H - ' -s 2, i z Cain, re ° prapeeccan us Ness N De ee ager ° $/ 3 | 3: z vba wet 0 9 y N SCALE IN MILES Mar 1.—Florida counties and the Gulf Coast archeological area. ‘. WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—-WILLEY iS has a substantial basis both in natural environment and culture types. South of Charlotte Harbor the climate is subtropical, and the archeo- logical culture of the Glades area, as it is called, differs markedly from the Gulf Coast. To the east the lines are less sharp both as to environment and prehistoric cultures. There is more shading of one type into another. Nevertheless, differences have been observed (Goggin, 1947b; n.d. 2). The northern boundary is, perhaps, our weakest as the archeological picture in south Georgia and south Alabama is only hazily recognized. I am inclined to think that here the Florida Gulf Coast culture area might eventually be extended inland for a greater distance than we have indicated in this paper. Westward the boundary is also largely arbitrary. Close similarities are known to exist between the Florida Gulf Coast and southern Louisiana and Mississippi. There are also differences. Pre- sumably a division could be made somewhere along the Alabama or Mississippi coast, but, at the present time, we do not know where it could most accurately be placed. The Florida Gulf Coast has unity as an archeological culture area in that it has been the locale in which the same archeological patterns and fusions of these patterns have persisted over a considerable span of time. There are observed, however, a number of secondary differ- ences in these patterns which are reflected in subarea variations of cultures. These subareas are referred to here as “regions.” Three such regions have been formalized by Goggin (1947b; n.d. 2). These are, from north to south, the northwest coast, the central coast, and the Manatee region. The northwest coast includes all of Florida and adjacent Alabama and Georgia as far east as the Aucilla River on the border of Jefferson and Taylor Counties; the central coast extends from that division point down to the southern end of Tampa Bay; and the Manatee region takes in the southernmost counties of the area going as far south as Boca Grande Pass, Charlotte Harbor (see map 1). Although the three regions are closely united during all culture periods, and virtually identical during some, there are period-to-period differences. Of the three the northwest and central coastal regions are most similar to each other, and the Manatee region is the most divergent. This matter of regional differences is con- sidered in detail in the ensuing sections of this report. PROCEDURES OF CLASSIFICATION The formulation of culture periods—Archeological data offer al- most limitless possibilities for the classificationist. A classification is 4 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 113 an arbitrary procedure; the grouping or categorizing of phenomena reflects the attitudes of the classifier toward his data rather than any inherent “truths” in the materials themselves. In view of this it is incumbent upon the classifier to define his particular attitudes or the ends for which his classification is but the means. In establishing the culture “periods” as the major classificatory device of this report we are operating with the basic assumption that culture changes through time. The culture period is the means of measuring and describing cultural forms, both material and non- material, as these have existed in a time continuity. Ideally, this period-to-period change should be uniform throughout the geo- graphical area of study; however, few culture areas of any size display such time-depth regularity. The Florida Gulf Coast ap- proximates it but does not fully attain it; to accommodate these spatial variations in the time continuity of the cultures of the area we have devised the above-mentioned subareal divisions for the regions. Lacking any satisfactory means of establishing an absolute time scale by which to measure and segment the prehistory of the Gulf Coast we have defined these culture periods by their own content. Certain qualitative and quantitative cultural features have thus come to symbolize a cultural period. Both the period and the complex of features are designated by the same name. The particular points in the time continuum which have been selected as the dividing lines between a period and its predecessor or successor have been chosen in an attempt to emphasize certain trends and changes in the pre- historic developments. For example, the appearance or disappearance of a pottery type or the shift from mound to cemetery burial have been capitalized upon as markers on the time scale. The primary method by which the culture continuum has been arranged and given direction from early to late has been stratigraphy. This has been continuous rather than discontinuous stratigraphy and has effected a correlation between arbitrary depth in refuse deposits and cultural succession. This method was successfully employed on the northwest coast, and its results have been extended by com- parison and inference to the central coastal and Manatee regions. In the northwest each of the culture periods in that region was first established upon the presence of certain pottery types and upon the percentage configuration of these types in a particular stratigraphic depth context. By this it is meant that the simple presence of a pottery type is not always enough to denote a specific period; its WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—-WILLEY 5 association with other types and its percentage relationships with these types in context were more often the significant factors. It will be seen from this that the culture periods, as they were first defined, were essentially ceramic periods. The general avail- ability and stratigraphic occurrence of pottery made it the handiest media for our purpose. Furthermore, the assumption that pottery is one of the most sensitive reflectors of culture change seems sub- stantiated in the Gulf Coast area where its observable changes pro- gressed at a much more rapid rate than did those of other prehistoric manufactures or mortuary customs. The period constructs as first outlined by ceramic stratigraphy were tested and expanded by pro- jecting them against the broader archeological background of the area. The ceramic stratigraphies of the northwest coast were com- pared with pottery series from burial mounds and cemeteries in the same region and with burial-mound collections of the central coast and the Manatee regions. The final results of this were much fuller definitions and characterizations of the culture periods. Pottery classification —As the basic time guide pottery has been described and classified in accordance with a definite system. The basis of this system is its smallest classificatory unit, the type. The concept of the type as used in this report follows procedures of defi- nition which are current in the archeology of the southeastern United States. The function of a pottery type is as an historical tool. It is conceived of as an abstraction based upon a specified range of con- structional and artistic variables which are recognized in a group of pottery specimens. To be of value in the solution of historical problems each type must have a definable time and space position. Krieger (1944, pp. 272-273) has made the point that this time-space position of the type must be demonstrable by archeological data and not in terms of “logical” assumptions or unproved hypotheses of culture development. As the type is thought of as representing a number of combina- tions of techniques in manufacture, materials, form, and decorative features it is obvious that not all of these features, or modes (see Rouse, 1939), will appear on any one specimen but will tend to fall in a normal frequency of occurrence curve for the type. The degree of allowable variation within a type follows no set rules. It has, however, been a principle to establish a new type when any material, form, or style variation is found to have temporal, spatial, or associa- tive significance. This has been the general intent in the present re- port although there are some exceptions. 6 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 113 Type nomenclature has followed Southwestern procedures, now also in use in the Southeast. This is to let the names of sites or geographical areas which are associated with a certain type be used as the “designant name.” Techniques of pottery decoration serve as the second name. Where necessary an intermediate modifying term is introduced. Thus, the type name Swift Creek Complicated Stamped is composed of the type site designant, “Swift Creek” ; the decorative term, “Stamped”; and the modifying adjective, “Complicated.” The classificatory term of the next order is the series. I have introduced this in the present report as a means of grouping together a number of types which bear a very obvious relationship to each other. Usually companion types in the same series have about the same tem- poral and spatial distribution although minor differences may exist on this point. The series is named after one or more prominent types which compose it. For example the Weeden Island Series consists of Weeden Island Plain, Weeden Island Incised, Indian Pass In- cised, Keith Incised, and a number of other closely related types. In this report I have consistently excluded types from a series if identity of paste and temper was lacking. Thus, the Papys Bayou Series types, although clearly related to the Weeden Island Series in decorative technique and design, have been set apart as a separate series be- cause of a difference in ware qualities.1 The most inclusive classificatory term in the pottery classification system is the complex. A complex is synonymous with culture period in that it represents the group of pottery types or the various series of types that occur together in the same general area at the same time. Thus, the Weeden Island Complex is composed of several series and each series consists of several pottery types. It is, of course, possible for types of two different complexes to be contempo- raneous. It must be remembered that the cultural development of the Florida Gulf Coast was a continuum and the actual picture was a dynamic one. The dying types of one complex or period may have overlapped in time with the inception of types of a new period and complex. Actually, they were contemporaneous for a time, but they have, for purposes of description and analyses, been made static and classified as components of one complex or the other. Techniques of decoration such as incision, punctation, stamping, rocker-stamping, etc., are well known and common throughout Southeastern and Eastern literature. Similarly, material and con- 1 This difference in ware was demonstrated to have distributional significance. WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—WILLEY 7 structional terms need no glossary qualification as the range found in Gulf Florida is about the same as that encountered elsewhere in the eastern United States. Vessel forms, however, are extremely varied and in many cases unique to the Florida Gulf. For this reason, a vessel form classification is added as an appendix to the section on “The Culture Periods,” following the various pottery type descrip- tions (see pp. 496-506). This classification has been formulated after a thorough review of the C. B. Moore publications, the Moore mu- seum collections, and other Gulf Florida ceramic collections. Descriptive classifications—With the exception of the pottery, all artifacts are classified by an informal descriptive system. For example, projectile points are listed by blade and haft form and size; stone celts are usually referred to as “pointed polled” or “rectangular” ; copper ornaments as “bicymbal ear spools of copper,” etc. Similarly, such features as mounds are designated with self-evident descriptive terms such as “rectangular flat-topped temple mounds” or “circular, conical burial mounds.” In general, the lack of sufficient numbers of artifact specimens, the absence of adequate provenience data, or an extremely limited number of forms or variations of an artifact or feature made formal systematic classification unnecessary. : ets | 7 ee A a4 Vs aa, ae cf Part : iat an ‘yes iinet : ‘ } an iv mapele Pues, | OAV ee 4: a a, we a > oa . er é ih) e hated Les SPUN, Pe ee oer 5 ae? 7 FOS aK > tate t fits Ae - a | pe Zar pave F) CY eee ee © en 4 j nb Me Ma i) i ioe “Tay oo oy: i) j ; iT ; im Vy a ® Wo iA Ba 7 ‘ woe ; ’ Toe | Ai Ly > : ly Cott - dpi } > Gat . f je 2 « = = ~~ 4 J rh} 4 iad. . + “ a . s | a J é ’ o * Cede ' 4 Penk i hi y ‘ ' thy 7c 4 he cae bw yg t i s J * =. 7 - ‘he u ft Role Mi f. a2 a ‘pe t) “5 ‘i ba ul R p : * os vy, La Piit, oh hae OM! iy } Bn bam ca ; A i dis a i ai ir rh nM | Le ee ee AT! I. THE NATURAL AREA GENERAL CHARACTERIZATION The area of this study (see map 1) a strip approximately 50 miles wide which extends around the Gulf Coast of Florida from Perdido Bay on the Alabama-Florida line to Boca Grande Pass, Charlotte Harbor, in the southwestern part of the peninsula, is a part of the Gulf Coastal Plain of the eastern United States. In general, there are two physiographic types in the area, the flat marine terraces of the coast and the rolling hills which lie several miles inland. The dominant vegetation pattern is that of the southeastern meso- phytic evergreen forest; however, this is interspersed with swamps, marshes, and hammock lands of oak and other hardwoods. Year- around climate is mild to warm, with hot, wet summers and rela- tively dry winters. There is a gradual increase in summer heat and rainfall, proceeding from the northwest to the southeast. Drainage is rather poor, owing to the low-lying terrain and the height of the water table. Most of the rivers that flow into the Gulf are short, although there are some exceptions to this in the cases of the major rivers whose headwaters are far to the north in the uplands of Alabama and Georgia. Along the northwest coast there are several large, well-protected bays into which many of the streams drain. There are fewer of these bays on the west coast, with Tampa and Charlotte Harbors being the only important exceptions. GEOLOGY, SOILS, VEGETATION, AND CLIMATE The Florida Plateau is an ancient geologic feature which today is only partially above the waters of the south Atlantic and the Gulf. It dates back at least to the Paleozoic when it was a part of the old land mass of Appalachia. During the Upper Cretaceous the entire plateau was covered by the sea. Afterward, in the Cenozoic, the Florida shoreline underwent many shiftings, but it was never sub- merged to any great depth nor, correspondingly, was it ever raised very high above the water level. Sedimentation during the Cenozoic was from the Piedmont Plateau and was predominantly limestone, especially during the Eocene and Oligocene. Miocene deposition combined sand and limestone and the later epochs sand and clay. The topography, soils, and vegetation of Gulf Florida can be di- vided into a number of small zones which differ from each other in 3 9 se) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. II3 varying degrees. In west Florida, large portions of Escambia, Santa Rosa, Okaloosa, Walton, Calhoun, and Gadsden Counties make up the west Florida pine hills. This zone lies back away from the coastal flats and extends on into Georgia and Alabama. It is composed of red-orange clay hills. Rivers are numerous and there are occa- sional swamps. Open forests of long-leaf pine are the characteristic vegetation cover. A similar upland is the Knox Hill country in Walton and Washington Counties. A little to the east of this in Holmes and Washington Counties, is the west Florida lime-sink region, a low-lying area of thin soils with open forests of pine. Solution of underlying limestone has produced ponds and swamps around which hardwoods and cypress cluster. To the northeast, in Jackson County, is a region known as the Marianna red lands. This is red-clay hill country but with numerous limestone outcrops and natural caverns. The soil here is rather rich. Another particu- larly favorable area for agriculture is the Tallahassee red hills in Leon County. This is upland covered with short-leaf pine and de- ciduous trees. Around Tallahassee there are several large, shallow lakes, the result of river channels dissolving out the underlying limestone and causing the valley floors to sink below the level of the outlets. The valleys are large and broad-bottomed, the lakes extremely shallow, and the land the richest in Gulf Florida. The Tallahassee red hills are surrounded by another topographic-vegeta- tion zone, the middle Florida hammock belt, an alternately hilly and flat region with moderately good soils. South of the hills, but not a part of the immediate beach country, are the Apalachicola flatwoods, embracing parts of Gulf, Calhoun, Liberty, Franklin, and Wakulla Counties. This is a rather poor lowland sustaining pine, palmetto, and wiregrass, and dotted with cypress ponds. An exception to this topography is a series of high bluffs forming a narrow strip along the east bank of the Apalachi- cola River from Bristol to the Georgia border. Other zones occupy- ing a similar transition place between the clay-hill uplands and the coast proper are the Bellair sand region of Leon and Wakulla Coun- ties and the Wakulla hammock region of that county. Here the soils are deep sands and not well suited to agriculture. The middle Florida flatwoods, farther to the east and south in Taylor, Lafay- ette, Alachua, Marion, and Levy Counties, are comparable to the Apalachicola flatwoods in soils, lack of topography, shallow lime- sink depressions, and sluggish streams. These middle Florida flatwoods lie back about 20 miles from the coast. WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—WILLEY II North and east of the middle Florida flatwoods, stretching from Georgia southward through Madison, Hamilton, Suwannee, Gil- christ, Alachua, Marion, Sumter, and Pasco Counties, is the penin- sular lime-sink region. This is a profusion of low hills and basinlike depressions resulting from limestone solution. Rivers and swamps are scarce but springs or underground streams common. The country is covered with pine except along occasional streams and in the deeper limestone sinks where hardwoods grow. The soil is sandy but can produce good crops. The actual coastal strip, varying from 5 to 20 miles or so in width, can be divided into three zones: (1) In west Florida, from Fs- cambia Bay to the Ocklockonee River there is a topography of simple beach dunes and marshes. Beaches are wide and composed of fine sand. (2) South and east of this, from the Ocklockonee to Tarpon Springs, the shore is marshy. Along this coast the water for a long way out into the Gulf is quite shallow and large waves do not reach the shore, preventing the formation of smooth sandy beaches. In many places along this stretch there are limestone outcrops at the water’s edge. (3) South of Tarpon Springs, down to Charlotte Harbor and beyond, the outer bars are sandy beaches, but the pro- tected inner shores are lined with mangroves. The beaches or marshy shore, as the case may be, are backed up by several miles of flat pine lands, marshes, and hardwood hammocks. Neither in the north- west nor the west is this type of terrain suitable for agricultural production. From Gordon’s Pass southward, a region lying out of the Gulf Coast area as we have defined it, mangrove swamps line the outer as well as the inner shore. Climatic variation in the Gulf Coast area can best be schematized from the following figures based upon three weather stations: one, in the extreme west at Pensacola; another, near the center of the coastal area at Apalachicola, and a third in the extreme south at Fort Myers. The Pensacola station gives average temperatures for January of 53° Fahrenheit and for July of 80°; there are 300 days to the annual growing season; and rainfall is 57 inches annual with 4.6 inches in a maximum spring month and 7.5 inches in a maximum summer month. The Apalachicola station shows temperatures of 55°-81°; 305 growing days; and 57 annual inches of rain with 3-5 inches for a maximum spring month and 7.7 inches for a similar summer month. Fort Myers has a temperature variation between January and July of 64° and 81°; growing days per year exceed 335; I2 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. II3 and with 52 annual inches of rainfall the maximum spring month is 2.2 inches while the maximum summer month is 8 inches. These figures show that while there is more annual rain in the northwest, there is a greater summer precipitation in the south. Semitropical to tropical conditions prevail in the southern part of the area for several months each year. On the whole, the climate and precipita- tion are well suited for corn and other crops that are best adapted to a mild, relatively dry winter and a hot, wet, and humid summer. LIFE POTENTIALITIES The potentialities of the Gulf Coast area for human exploitation on a preindustrial level are definitely restricted. Although climate and soils are generally favorable to heavy native vegetation, the soils and drainage near the coast are, for the most part, unsuited for inten- sive agriculture. Mixed sandy soil which is best for crops is found only in small scattered patches along the coast; the more plentiful pure sandy soils and organic soils require modern cultivation tech- niques for successful production. Inland, in the hilly country, es- pecially in the north, as has been noted, there are many sections where agriculture now flourishes and undoubtedly did so in the past. Today, corn, beans, and sweetpotatoes, as well as many other food crops, are grown in abundance in these uplands but less success- fully in the flat lands. The native resources of the Gulf Coast were, however, quite ample to offer subsistence for a small and well-distributed population. Previous to European settlement, it was a region of wild game, such as deer, bear, panther, wildcat, fox, opossum, raccoon, skunk, musk- rat, and rabbit. In early historic times buffalo were found as far east and south as western Florida. Turkey, duck, various sea birds, fish, turtle, alligator, and mollusks were extremely plentiful. The natural conditions of the Gulf Coast waters were, in fact, ideally suited for the breeding of shellfish. Oysters breed only where a combination of shallow, salt-water shoals and river deltas obtain. The few large and many small rivers emptying into the Florida Gulf offer these conditions to a degree that is unmatched almost any- where else in Atlantic North America. It is believed that a rising sea level in the later geologic history of the Florida peninsula cre- ated more of these coastal shoals and marshes resulting in more mollusks. Today, marine foods remain one of the principal economic resources of the coast although there are sections where oysters and WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—-WILLEY rs clams are no longer extant. The prehistoric refuse middens contain numerous fish and animal bones as well as shell.” In addition to the fauna, wild plum, crabapple, blackberry, cherry, persimmon, huckleberry, and swamp chestnut grew in profusion and served as supplementary diet. Native materials for building and for artifactual use were lim- ited to wood, shell, bone, and flint. Stone, other than flint, limestone, or soft coquina and occasional sandstones, was lacking. Metals, such as copper and galena, were not present. Communication on the numerous streams or along the coast in the places protected by offshore bars was reasonably easy and rapid by canoe.° 2 Identification of such remains from the old village middens excavated in 1940, at Carrabelle, Mound Field, Sowell, Fort Walton, Gulf Breeze, and Lake Jackson, revealed the following species: Mammals: Opossum (Didelphis virginiana) ; racoon (Procyon lotor). Fish: Sea drum (Pogonias cormis) ; jack fish (Caranx hippos). Turtles: Box turtle (Terrapene major); mud turtle (Kinosternon ?); gopher tortoise (Gopherus ?). Birds: Florida cormorant (Phalacrocorax auritus). Shells: Polynices duplicate; Rangia cuneata; Chione cancellata; Pecten irradians; Englandia rosea; Fasciolaria tulipa; Oliva sayana; Ostrea virginica; Melongena corona; Arca sectiocostata; Strombus pugilis; Murex fulvescens; Busycon contrarium. This is only a partial list of prehistoric fauna from the area. 3 This section on “The Natural Area” is based upon first-hand observation and upon the following references: Cooke, C. W., and Mossom, S., 1929; Fenne- man, N. M., 1938; Goggin, J. M., 1948b; Harper, Roland M., 1914; Kroeber, A. L., 1939; Martens, J. H. C., 1931; U. S. Department of Agriculture Year- book, 1941. II. HISTORY OF ARCHEOLOGICAL RESEARCH: 1846-1946 EARLY INVESTIGATIONS The period of contact and conquest in Florida begun in 1513 by Ponce de Leon and continued through Spanish, French, and English colonization, came to an end with the cession of Florida to the United States in 1819. The country remained wild and semipopu- lated for several decades after this, but by 1850 the Indian was no longer looked upon as a serious problem for the settler but rather an object of curiosity. Archeology, as distinct from general ex- ploration, began in this second half of the nineteenth century. With the widespread stimulation of interest in the ancient past of the New World, following the rediscovery of the Mayan ruins of Yuca- tan and the explorations of the mounds of the Ohio Valley, notes and comments on the antiquities of Florida began appearing in the scientific journals of that time. Aside from William Bartram who in the 1770’s made his cele- brated journey through north Florida, traveling as far west as Mo- bile (Bartram, 1940), the first published accounts of Gulf Florida archeology appeared just before the Civil War. In the 1830’s and 1840’s the small towns along the west coast of Florida were established and the region was more thoroughly explored. Two naturalists, J. H. Allen (1846) and T. A. Conrad (1846), primarily interested in making conchological collections, visited and described the large shell middens near the mouth of the Manatee River in the middle 1840’s. Both these men were of the opinion that the huge shell de- posits were the result of natural wave action. As Allen states: “.... their immense quantity precludes the idea of their having been accumulated by the aborigines of the country.” In 1854, in Schoolcraft’s monumental though somewhat uneven “Indian Tribes of the United States, Part III,” there is a discussion of aboriginal pottery from the “low mounds of the Gulf Coast’ (Schoolcraft, 1854, pt. III, pp. 77 ff.). The collection under consideration was submitted to Schoolcraft by a Mr. Hitchcock who had obtained it in the field, in the vicinity of Apalachicola Bay, in 1841. The data are not sufficient to locate the site. Typical burial mounds are de- scribed as being 12 to 18 feet high and 30 to 50 feet in diameter ; they were constructed of black soil and sand, and were encircled with trenches (the borrow trenches). Shell middens are also mentioned 15 16 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 113 as a distinct type of site. Such features as artificially perforated or “killed” pottery, clay pipes, lump galena, and metal are listed as coming from the mounds. The art style of the pottery was classified, in conception and skill, as intermediate between Mexico and the northeastern United States. Apparently Schoolcraft’s plate 45 is the first illustration of Gulf Coast archeological pottery. The sherds shown are of the Fort Walton culture period. The distinguished anthropological scholar Daniel G. Brinton vis- ited Florida in 1856-57, and, although he spent most of his time on the St. Johns River, he made a quick trek to the Gulf Coast from where he reported large mounds in Marion and Alachua Counties, along the Suwannee River, and as far south as Charlotte Harbor (Brinton, 1859). Brinton, too, marked the distinction between shell middens and sand mounds, and in a later paper (1867) he called attention to the great shell refuse heap on the Crysta! River. This Crystal River report was based upon the field observations of F. L. Dancy, at that time State Geologist of Florida. Dancy located the shell mound at 4 miles upstream on the Crystal. Another distinguished scientist of about the same time, Jeffries Wyman, described a shell mound at Cedar Keys, apparently the great midden pile reported on by various parties since that time (Wyman, 1870). In 1869 R. E. C. Stearns studied the shell deposits around Old Tampa Bay, on Point Pinellas, and at Cedar Keys. He commented upon the distinction between the shell heaps, which were refuse middens, and the purposefully built sand burial mounds (Stearns, 1870, 1872). The first description of excavations on the Gulf Coast was offered by an army surgeon, G. M. Sternberg, in 1876. He lists two shell mounds, one at Bear Point, Ala., and the other at Anerierty’s Point, Fla., on Perdido Bay, both of which he recognized as middens. His excavations were principally directed toward a sand burial mound at the first location, a tumulus some 12 to 15 feet high and 100 feet in diameter. Sternberg’s listing of burials and artifactual remains is the most detailed up to that date.* Sternberg also excavated in both the shell midden and the big platform mound at Fort Walton at the eastern end of Santa Rosa Sound. He does not, however, describe the site under this name (Sternberg, 1876). Both Bear Point and Fort Walton were the scene of later archeological activities by Clarence B. Moore and others. 4A Francis H. Parsons, of the U. S. Coast and Geodetic Survey, obtained a large collection from this mound in about the year 1889. Holmes (1903, p. 105) refers to this, but, apparently, Parsons never published on his findings. WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—-WILLEY 17 There were a number of other rather desultory attempts at investi- gation in the 1870's, none of which were very fully reported. Calkins (1877-1880) described some excavations by a Lt. A. W. Vogdes in Tampa Bay shell mounds. This work was done in 1876 (see also Vogdes, 1879), and only the finding of human skeletons is reported. Vogdes also dug into the big shell refuse at Cedar Keys. Charles Rau reported upon a gold effigy woodpecker ornament found in a mound, along with pottery and human bones, in Manatee County (Rau, 1878). C. J. Kenworthy, making a boat trip from Key West to Cedar Keys in 1877, listed shell mounds between the Crystal and Homosassa Rivers and others around Orange Lake in Alachua and Marion Counties. Kenworthy also evinced considerable interest in the artificial canals on Pine Island, in Charlotte Harbor, and the large shell mounds in this same vicinity (Kenworthy, 1883). Henry Gill- man (1879) excavated a burial mound northeast of Santa Fe Lake, Alachua or Putnam Counties, and found painted and punctated pottery, a stone ax, and two human skulls which he believed to have been filled with calcined human bones from other parts of the cremated bodies. A somewhat different note was struck in Florida studies by Ecker’s paper published in 1878. His “Zur Kenntniss des Korper- baues friiherer Einwohner der Halbinsel Florida” was the first serious attempt at physical anthropology for the area. It is in the central European tradition of the period. Measurements and descrip- tions of crania from a sand mound in Cedar Keys region are pre- sented in great detail and with exemplary orderliness. All fall within the brachycephalic and mesocephalic ranges. Like so much scientific industry of its kind it is rendered sterile by a virtually complete lack of cultural data concerning the skeletal material in question. Ecker was not to blame for this, however, as the material was gathered by others in 1871. In the Smithsonian Annual Reports for 1879, 1881, and 1883, S. T. Walker (1880a, b, 1883, 1885) gave reasonably thorough accounts of his observations and excavations on the Pithlochascootie River, on the Anclote River, at John’s Pass, at Maximo Point, in Old Tampa Bay, at Shaw’s Point at the mouth of the Manatee River, at Cedar Keys, and on the northwest coast. At the Cedar Keys site he con- ducted some crude stratigraphy and noted changes in pottery types by depth, but, unfortunately, included no good descriptions or illus- trations of his ceramic material. In the northwest, Walker visited and excavated some of the sites later investigated by Moore. These 18 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 113 include Escribano Point, Fort Walton, and Hogtown Bayou, all between Pensacola and the eastern end of Choctawhatchee Bay. Walker was the best Florida archeologist of his time. Obviously influenced by the cultural evolutionism of his day, he had some awareness of the principles of culture growth as interpreted through archeological remains. In this way, he had a generalized interest in chronology although his attempts to demonstrate it from his field work were not systematized. His descriptions of materials recovered were fairly full although he did not completely itemize his findings. In this he is the equal of Moore who was to follow him in the area some 20 years later. Walker’s total contribution was, however, much more limited than Moore’s. After Walker, little of consequence was achieved in Gulf Coast archeology until almost the turn of the century. IF’. LeBaron, in making a trip from Orlando to Charlotte Harbor in connection with explorations for a steamboat route through Florida, saw a small mound near Fort Myers on the Caloosahatchee and described the artificial canal through Pine Island in Charlotte Harbor (LeBaron 1884). James Bell described six sand mounds in the vicinity of Gainesville, Alachua County. Most of these mounds contained nothing in the way of artifacts or burials, but in one he found sec- ondary burials and pottery which, from his description, may have been of the Weeden Island style (Bell, 1883). M. H. Simons, of the U. S. Navy, published a brief account of shell-midden keys, mounds, and artificial canals from the southern end of Charlotte Harbor. Simons also explored the Caloosahatchee River for a dis- tance of 25 miles inland but reported no archeological sites (Simons, 1884). James Shepard, in making a few investigations around Tampa, gathered a collection of stone and shell tools as well as potsherds. Among the sites he mentions are Rocky Point, 5 or 6 miles west of Tampa, the Culbreath site on the shore of old Tampa Bay, and the Bull Frog mound at the mouth of the Alafia River sev- eral miles below Tampa. Shepard was a good observer and interested in the natural resources that were available to the primitive inhabitants of the area. He pointed out that all the marine shells which he saw corresponded with live species now inhabiting the nearby waters and that chalcedony and quartz were available at a location about 5 miles below Tampa in the form of silicified geodes (Shepard, 1886). G. F. Kunz described in considerable detail a flat rectangular gold pendant and a silver ornament, one from a mound on the west side of Lake Apopka, Sumter County, and the other from a mound near Tampa. Kunz was careful to point out in the case of the gold WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—-WILLEY 19 object that it was found deep in the mound in association with the bones of “hundreds of Indians” and was not, therefore, a later intrusion (Kunz, 1887). Much of this survey work of Walker, LeBaron, Bell, and others was assembled by Cyrus Thomas in his exhaustive mound survey of the eastern United States (1891). Although little more than a catalog of county, location, and type of Indian site, this work served to bring together for the first time a skeleton outline of just what had been accomplished in archeological research in Gulf Florida as elsewhere. Thomas’ field agent, Rogan, excavated two mounds in Alachua County and these operations are described by Thomas in the Twelfth Annual Report of the Bureau of American Ethnology (1894). Work on a much larger scale than heretofore reported was under- taken on the Gulf Coast by F. H. Cushing and party at Tarpon Springs in 1896. This was a prelude to the later and more famous excavations at Key Marco in the Glades area keys. A small sand mound, known as the “Safford mound,” was thoroughly excavated and was found to contain over 600 human skeletons as well as a large amount of pottery. This was a Weeden Island Period mound, and Cushing’s work, for the first time, drew considerable attention to this very distinctive style of pottery. Cushing and party also excavated in a mound at Finley Hammock, 9 miles northwestward of Tarpon Springs. Unfortunately, the Tarpon Springs material was never reported upon in detail, although a few years later some pottery specimens were illustrated by Holmes (1903). In the brief preliminary report (Cushing, 1897) Cushing offered some very colorful theories with regard to the nature of the mound construction, the disposition of burials, and the general cultural affiliations of the builders of the Safford mound. Of all the early Florida Gulf investi- gators he was the most given to extreme speculation. In connection with Cushing’s cruises to Key Marco he made a number of interesting observations among the keys in Charlotte Harbor, Pine Island Sound, and Caloosa Bay. This region of the keys is the borderland between the Gulf Coast and Glades archeo- logical areas as they have been defined in more recent times. The phenomena which Cushing saw and described here are somewhat more typical of the south than the north. He was considerably im- pressed by the artificial or semiartificial terracing, the platforms, and the mounds on the shell-refuse-covered keys at Demorey, Josselyn’s and Battey’s Landing, and he described these in some detail, includ- ing photographs and drawings (Cushing, 1897). 20 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I13 A contemporary of Cushing, Thomas Featherstonhaugh, reported on some mounds and mound excavations in central Florida at about the same time (Featherstonhaugh, 1899). These were near Lake Apopka and Lake Butler in Lake and Orange Counties. Although this region is as close to the east coast as it is to the Gulf, the pottery which Featherstonhaugh found in the mounds pertains to the Weeden Island Period. Featherstonhaugh’s account is fairly complete but lacks illustrations. Fortunately, his material is available in the United States National Museum. From this brief summary, it is obvious that the period between 1850 and 1900 was not a brilliant one in Gulf Florida archeological research. Archeology as a profession in the United States was in its infancy, and techniques in the field as well as manner of presenta- tion of data were undeveloped. Of all of the men named, only Brinton, Wyman, and Cushing could be considered professionals, and of these three only Cushing worked to any extent in the Gulf area. Of the entire group, S. T. Walker shows up the best, as far as Gulf Florida studies are concerned, and he was an interested layman. Most of the accounts are merely the casual relations of educated men, naturalists, doctors, military and naval officers, lawyers, and engineers who had an appreciation for both natural and human history. The fact that much of this material was published at all was due to the stimulation by Powell and Thomas, of the Bureau of American Ethnology, who were encouraging responsible persons in all sections of the country to make investigations and to communicate some record of these to the Smithsonian Institution. Similarly, the editors of the American Naturalist and the American Antiquarian encouraged their readers to report their archeological findings. Of various unpublished investigations of the period there is almost nothing known except where museum collections exist. With the passage of time old correspondence is lost or forgotten and artifact collections often scattered or separated from any provenience data that may once have accompanied them. As the Gulf Coast was less thickly settled then than now, it is not likely that there was as much unrecorded digging and “pothunting” previous to 1900 as afterward. I would imagine that a good proportion of the gentlemen of “scientific bent” of the last century made, at least, some brief published entry of their activities. The solid results of the “early investigations” are these: The great number of aboriginal remains of the Gulf Coast were made known to the scientific world. These were accurately grouped, as far as we can yet tell, into functional categories of burial mounds, house WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—WILLEY 21 platform mounds, refuse piles, and the canals, terraces, basins, etc., of the keys below Tampa. Some knowledge was gained as to the nature of the artifacts found in the Gulf Coast sites, although this knowledge was not systematized and very few illustrations of artifacts were published. From the discovery of European trade materials some of the mounds were identified as post-Conquest while others, with- out such objects inclusive, were considered wholly aboriginal. It was generally conceded that the sites and remains were those of the Indians or the ancestors of the Indians who occupied Florida at the coming of the Europeans. The controversy as to “Indians versus a mysterious race of Mound Builders” was not fought out on the Florida scene, but the outcome of the dispute which raged farther to the north seems to be reflected in most of the statements of the men writing on the Florida Gulf. THE SURVEYS OF CLARENCE B. MOORE In volume of work accomplished and in amount of published data Clarence B. Moore ranks as an outstanding investigator in the Florida field. Earlier workers carried on their researches rather sporadically and published only brief notices in the scientific journals. Moore, through sheer dint of effort, persistence, and thoroughness in cover- ing the ground, brought a certain system into Florida studies. By faithfully publishing his findings he has left an invaluable record that forms the groundwork for any synthesis of Florida prehistory. In method and sense of problem, it is true, he differed but little from most of his predecessors or colleagues. Moore introduced no new point of view to Southeastern studies ; his contribution was of a quantitative rather than qualitative value. His interests were in phenomena themselves, in their appearance and diversification, rather than in any scheme into which they could be placed. He had little appreciation of chronology and took only casual note of geographic distributions. Nor was he scrupulously careful to itemize all features, burials, and artifacts in the sites that he excavated. He was aware of the ethnohistoric contacts of the early Spaniards on the Gulf Coast, and he occasionally mentioned the European explorers in connection with a particular vicinity in which he was carrying out explorations. There was no attempt, however, definitely to “document” a site in terms of historical incidents. Similarly, al- lusions are made to the early historic tribes of the area without any systematic effort to identify tribes with archeological sites. Like many of his contemporaries, Moore was impressed with the value of 22 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 113 physical anthropology and made and recorded various observations on the skeletal material found in the mounds. These observations in- cluded pathology, head form, size and shape of long bones, and cranial deformation. They are not metrically or systematically pre- sented, nor are they, except for a general statement on cranial flat- tening, synthesized for the area or any part of the area. On most of his trips he was accompanied by a Dr. M. G. Miller, a medical man, who aided in the task of identifying and commenting upon human skeletal material. In other lines Moore also sought specialists’ advice. Rocks and minerals were turned over to geologists for classi- fication, metals were analyzed, and even soil samples and potsherds were microscopically examined. Unlike many of his colleagues, Moore was not given to sweeping speculations, at least in these Florida reports. Occasionally he offered a theory, often concerning the use or function of some archeological object or feature. Some- times he pointed out what seemed to him similarities between one of the specimens uncovered during his survey and artifacts from other parts of the Southeast or, rarely, the Americas at large. But, gen- erally, his work was factual and not conjectural. Moore made his first excursion into Gulf Florida in the winter of 1900. Previous to this, he had been surveying and excavating sites in the St. Johns drainage of east Florida, but the remarkable finds which Cushing (1897) uncovered in the muck at Key Marco at the head of the Ten Thousand Islands attracted his attention to the west. Much of Florida was still a wilderness at that time, and land trans- portation was difficult in the thinly settled regions. Because of this, Moore’s surveys were conducted by boat. Beginning at Clearwater Harbor, north of Tampa Bay, Moore’s party, consisting of associates, ship’s crew, and laborers, proceeded southward. Excavations were made at mounds and shell piles within Tampa Bay, on the Little Manatee and Manatee Rivers, in Sarasota Bay, in Pine Island Sound, on the Caloosahatchee River, and, finally, south into the islands and keys. The southern extent of this survey of 1900 was the Chatham River of Cape Sable. Within the Gulf Coast archeological area the sites he investigated range from Four Mile Bayou to Indian Hill in Hillsborough County. He critically examined Cushing’s (1897, pp. 338-339) claim for truncated pyramids faced with conch-shell masonry and successfully disproved it. In general, Moore considered it a very unsuccessful season as he was unable to locate anything comparable to Cushing’s “Court of the Pile-Dwellers.” The 1900 survey was published as “Certain Antiquities of the Florida West Coast” (1900). WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—WILLEY 23 In the following year, survey was resumed but in the extreme northwest, on Perdido Bay. The first site explored was the “Mound at Bear Point,” Baldwin County, Ala. Moving eastward Moore in- vestigated and recorded 16 mound or cemetery sites. The season of IQ9OI was terminated on the southern side of Choctawhatchee Bay at Hogtown Bayou, Washington County, Fla. These excavations are described and the materials from the sites are illustrated in Moore’s “Certain Aboriginal Remains of the Northwest Florida Coast,’ Part I (1901). In 1902 Moore continued his survey, beginning at St. Andrews Bay which lies next in order to the east of Choctawhatchee Bay. The first site was on West Bay Creek, a small stream flowing into St. Andrews Bay from the west. Proceeding eastward along the coast, 52 sites, mostly sand burial mounds, were excavated and reported upon during this second season in northwest Florida. The last and most southeasterly site was the “Mound near the Shell- Heap,” Levy County, Fla., just south of the mouth of the Suwannee River. The area covered during 1902 included St. Andrew’s Bay, the Bay of St. Joseph’s, Apalachicola Bay, St. George’s Sound, Ock- lockonee and Apalachee Bays, and Deadman’s Bay. A detailed account of this work is given in “Certain Aboriginal Remains of the Northwest Florida Coast,” Part 11 (1902). In 1903 Moore continued by moving south from the region of the mouth of the Suwannee as far as Tampa Bay, thus completing his first inspection of the entire Gulf Coast area. He began in the north at “Fowler’s Landing” on the Suwannee, and, after a brief 50-mile trip up that river, went southward, concluding at Long Key, an island fronting Tampa Bay. From north to south he surveyed the lower reaches of the following rivers, as well as the intervening coastal stretches: the Waccasassa, Withlacoochee, Crystal, Homosassa, Chas- sahowitzka, Wekiwachee, Pithlochascootie, and Anclote. This sea- son’s account appeared as “Certain Aboriginal Mounds of the Florida Central West Coast” (1903). Supplementary investigations were made in 1903 on the Apalachi- cola River which drains into Apalachicola Bay. On this waterway, from the Gulf to the point of confluence of the Flint and Chat- tahoochee Rivers, 14 mound sites were excavated, ending with “Mounds at Chattahoochee Landing” in Gadsden County, Fla. These 14 mound sites are discussed in Moore’s “Certain Aboriginal Mounds of the Apalachicola River” (1903). In 1904, after an unfortunate beginning on the Kissimmee River in interior Florida, north of Lake Okeechobee, Moore returned to the west coast to make excavations in the vicinity of Charlotte Har- 24 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I13 bor and to the south. He eventually rounded Cape Sable, visited various southern keys, and put in at Miami on the east coast. This account is given in his “Miscellaneous Investigations in Florida” (1905). Still later, in 1906, Moore went farther up the greater Apalachi- cola drainage, on the Chattahoochee and Flint Rivers, respectively. Twenty-one sites, mostly burial mounds, were opened on the Chatta- hoochee in Florida, Alabama, and Georgia. The northernmost site was in Muscogee County, Ga., just below the city of Columbus. How- ever, the presence of burial mounds and an essentially Floridian type of culture were not met with beyond 50 miles from the Gulf Coast. On the Flint River, in Decatur County, Ga., Moore dug four sites. The Chattahoochee and Flint excavations are the subject of a paper appearing in 1907 with the title “Mounds of the Lower Chatta- hoochee and Lower Flint Rivers.” At about this same time, a brief period of excavation at Crystal River, in Citrus County on the west coast, terminated the explora- tions of the Florida Gulf for a good many years. These excavations are reported on as “Crystal River Revisited” (1907). Moore had, previously, examined the burial mound at Crystal River in his west coast survey of 1903. From 1906 until 1918, Moore was occupied elsewhere in Florida and the Southeast, but early in 1918 he made a final visit to north- west Florida which he summed up in “The Northwestern Florida Coast Revisited,” published in the same year. The 1918 investiga- tions included 9 mounds on the Choctawhatchee River, 13 sites on Choctawhatchee and St. Andrews Bays; 4 sites on the Apalachicola River in Florida and Georgia, and Io sites along the coast between St. George’s Sound and the “Greenleaf Place,” Citrus County. Most of Moore’s excavations were in small sand burial mounds and although he often mentions shell middens and other refuse areas near or in association with the mounds he seldom excavated these. An occasional cemetery drew his attention, and in several instances he sampled pyramidal platform mounds of the sort considered as temple or house substructures. Moore also mentions the excavation of small, low, circular or oval sand mounds in which no burials or purposefully placed artifacts were found. This type of mound was usually located near a burial mound, and the author considered them as domiciliary mounds, presumably erected by the builders of the burial mounds. Moore attempted no stratigraphic excavations on the Gulf Coast, and his conclusions are not primarily concerned with relative chro- WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—WILLEY 25 nology. In his earlier work on the St. Johns River he had discussed the temporal relationships of the ceramics and other artifacts from the various layers of the great shell deposits in that area. Following Wyman (1875), he pointed out evidences of nonpottery levels under- lying pottery-bearing strata, of intermediate fiber-tempered pottery levels, and of top levels showing decorated sand-tempered pottery. In his west Florida work he failed either to find or to look for indi- cations of sequence. Perhaps this was because nearly all the pottery of the Gulf area which Moore encountered has a general resem- blance to the upper-layer pottery of the St. Johns and he neither ex- pected nor tried to distinguish periods in what he considered to be a single horizon. Nor did Moore make any serious effort at classifica- tion, either of sites or of pottery. He recognized close similarities among all the burial-mound sites from Pensacola to the region of Tampa Bay, and it is now clear that this was a sound observation. Yet, in spite of many important features held in common throughout the area, there are many easily recognizable differences, especially in the pottery of the various sites. Some of these distinctions in ware and pottery types were known to him, for, along with Holmes, he saw a divergence in style between pottery of the Mobile-Pensacola district and that of the Apalachee Bay region. He was also well aware of the importance of European trade goods inclusive in a burial mound or grave. This makes it seem strange that he should have missed the suggestive correlation of European objects being found only in sites which contained pottery of Mobile-Pensacola or Mississippian affiliations. Moore’s conclusions deal with minor problems of distribution of selected traits. In his 1901 report he emphasizes the “mixed” nature of the cultures of the Choctawhatchee Bay district, pointing to the Lower Mississippi Valley ware, Georgia stamped ware, and character- istic Florida pottery all within this one region. The trait of basal perforation of funerary vessels, or the “killing” of pottery, was briefly examined upon an areal basis, and Moore advanced the hypothesis that this idea was Floridian, or at least had its first focus there after being transplanted from an unknown source. He concluded that the pottery in northern peninsular Florida with ready-made mortuary perforations was a refinement of the original idea of breaking or “killing” a mortuary vessel and was more limited in dis- tribution and later in time. Comparing his 1900 and 1901-02 field seasons, Moore (1902) calls attention to the fact that pottery in the burial mounds of peninsular Florida is neither equal in quality to that of the northwest coast nor 4 26 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. II3 does it occur as abundantly. He notes, in addition, that the custom of placing pottery with the dead does not obtain on the lower south- west coast of the peninsula or on the east coast. The custom of urn burial is observed in the northwest, and cremation, or partial crema- tion, is mentioned as being exceedingly rare in northwest Florida as compared to the Georgia coast. Of the mounds inland, on the Apa- lachicola River system, Moore concludes his 1903 report with the fact that these sites closely resemble those of the Gulf coast. After Moore’s work between the Suwannee River and Tampa Bay (1903), he drew some comparisons between the west coast and the northwest. He pointed out that no urn burials, no general deposits of earthenware found in blackened sand in mounds, no cranial flatten- ing, no prefired “killed” pottery, and almost no pottery vessels modeled after life forms were found below the “bend” of the State. But west coast burial mounds were numerous and similar to those of the northwest. Pockets of calcined human bone in the mound were met with in both regions. The general ceramic pattern was much the same, with graceful incised and punctated pottery, some complicated stamped, and abundant check stamped. Finally, he notes, with sur- prise, the presence of solid copper artifacts well to the south. THE CLASSIFICATORY STUDIES OF W. H. HOLMES W. H. Holmes was the first man of considerable intellectual stature to bring his talents seriously to bear upon the problems of Florida Gulf Coast archeology. Not primarily a field investigator, nor a Florida area specialist, his impress has, nevertheless, been the greatest of any of those of his time upon the later generation of archeologists. Holmes worked with the artifact collections gathered by others, and his energies were chiefly directed toward injecting order into the great mass of ceramic materials that had been brought to the museums of Washington and Philadelphia. He was one of the first to visualize Florida prehistory as but a part of the larger fabric of the Indian past of the eastern United States. With him, Florida archeology, for the first time, moved away from a consideration of discrete phenomena toward a comprehension of broad categories of related phenomena. In his introduction to the “Earthenware of Florida: Collections of Clarence B. Moore,” an analysis of the Moore and other collec- tions as of 1894, Holmes (1894a, p. 105) states: Exploration has not yet gone far enough on the peninsula of Florida to give archaeologists a firm grasp on the problems of its prehistoric art. The general nature and range of the remains are pretty well understood, as they form no WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—-WILLEY 27 marked exception to the rule in this latitude, but little has been done in the study of those details that must be relied upon to assist in assigning the art remains to particular tribes and stocks of people, in correlating them with cul- ture features of neighboring regions and determining questions of chronology. By this, Holmes demonstrated a grasp of the essentials of: (1) archeological-tribal correlations; (2) correlations with other areas or extra-areal cross ties; and (3) culture chronology. To that time none of his predecessors or colleagues had equaled him in understanding, or in any event clearly formulating, these principles. And again (Holmes, 1894a, p. 105) he pointed the way toward modern methods with: It is on ceramic evidence perhaps more than any other that we must depend for the solution of problems of time, people and culture, and to this branch of investigation the most careful and painstaking attention must be given. In both his Florida study and in his monumental “Aboriginal Pot- tery of the Eastern United States” (Holmes, 1903), Holmes stressed geographical groupings of ceramics. In defining these ceramic areas he considered it necessary to evaluate a number of characteristics or a complex of features which distinguished the ware of one such area from another. Form, temper, method of manufacture, and decoration were all taken into account. Nor was Holmes unsophisticated in his conception of the validity or usefulness of his major pottery prov- inces. He makes it quite clear that within a given area there might be pronounced ceramic diversity due to a number of complicating causes such as differences in culture period or the intermingling of different ethnic groups upon the same time horizon. Nor were the areas absolutely defined, for, as he says, “limitations of these varieties (in pottery), geographically or otherwise, are not well-marked, one grading imperceptibly into the others, features combining in such ways that many specimens occur that cannot be definitely assigned to any one of the groups” (Holmes, 1894a, p. 111). But the im- portance of this first step, difficult as it was, was indisputable, and Holmes did not hesitate to make it. Along the Gulf in northwest Florida Holmes recognized three major ware groups: the “Mobile-Pensacola,” the “Apalachicola,” and the “Appalachian” (Holmes, 1903, pp. 104-114). The Mobile-Pensacola group was based upon the Parsons collections from the sand mound at Bear Point, Ala., on Perdido Bay, on the Moore collections from the same site, and on Moore’s other collections from Choctawhatchee Bay. The vessels are well described and well illustrated, and the classificatory group is a good one. It is, essentially, the same material 28 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I13 that makes up the Fort Walton Period ceramic complex. Holmes, however, had no chronological evidence to support his classification. He did, though, draw comparisons between the Mobile-Pensacola Ware group and the pottery of southern Alabama, of Tennessee, and of Mississippi. Similarities were demonstrated in vessel forms and decorations, particularly the skull and hand designs, eagle and rattle- snake designs, and the frog-effigy bowl. From the vantage of a more recent point of view, it is clear that Holmes visualized this ware group as the strongly Middle Mississippian and late Lower Mississippi Valley influenced styles which are outstanding along the western Gulf Coast. The Fort Walton, or Camp Walton, collections were also lumped by Holmes into the Mobile-Pensacola group. At Point Wash- ington, a little farther east, he felt that there were significant differ- ences between the predominantly three-lined incised styles of that site and the Mobile-Pensacola group as a whole; at the same time he recognized their very great similarities. Between the Choctawhatchee Bay and the Apalachicola River he noted a lessening of Mobile-Pensacola Ware and an increase in what he called Apalachicola Ware.> This latter group is what is now considered as the Weeden Island ceramic complex. Holmes thought it peculiarly Floridian but showing interesting analogies to both the Caribbean Islands and Yucatan. He also observed the more im- mediate connection with the pottery from the Florida west coast site of Tarpon Springs or Safford which Cushing recovered. The presence of complicated stamped pottery in this section also helped to distinguish it from the Mobile-Pensacola-Choctawhatchee region. The inland Georgian affiliations of the complicated stamped pottery were appreciated by Holmes, but at the same time he was also puzzled by the design elements which were, to him, similar to those of the West Indies. In a separate paper (Holmes, 1894b), he advanced a hy- pothesis for West Indian-Southeast contact as expressed in the stamped pottery designs. Concerning the Moore collections between St. Andrews Bay and Cedar Keys, Holmes felt that there was a gradual shading off from the pottery groups of west Florida into those of the peninsula. He is not very clear about this, probably because the relatively minor stylistic differences within the Weeden Island Complex are more diffi- cult to synthesize and clearly point up than the sharp contrasts between the Mobile-Pensacola (Fort Walton) and the Apalachicola (Weeden 5 This study was based in part on a collection from Gulf County given to the U. S. National Museum in 1893 by C. H. B. Lloyd. (U.S.N.M. Nos. 155318- 155329.) WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—-WILLEY 29 Island) groups. Holmes also was aware that pottery of the Mobile- Pensacola group was found all along the northwest coast and as far south as Tampa. As indicated, the Holmes area classification of pottery still has sig- nificance for archeological research on the Florida Gulf. Some of the problems that bothered Holmes have been resolved in terms of period differences. Finer and more exact regional and subregional break-downs have been made in the Florida ceramic data. Extra-areal affiliations have come to be more clearly understood as a result of addi- tional work both in and out of Florida. On the other hand, many of the questions raised by Holmes, such as the hints of Middle Ameri- can or Caribbean resemblances, have not yet been explained to the satisfaction of the archeologist. THE PROBLEM OF EARLY MAN IN GULF FLORIDA As in most parts of the United States some claims for “Early” or Pleistocene man have been made for the Florida Gulf Coast. The status of these finds has never been completely settled, but the original materials and data were so sparse that proof for the occurrence of Early man on the west side of the Florida peninsula must await new finds. In 1871 J. G. Webb, of Osprey, Fla., sent a number of human bones to the Smithsonian Institution accompanied by the data of their discovery. They had been found on Sarasota Bay and, presumably, were from intentional burials. They were uncovered at 3 to 4 feet below the surface and were imbedded in soft ferruginous rock. Sub- sequently, a number of other human bones were discovered and de- scribed from the same general region. Some of the outstanding scholars of the day visited the area, and in 1887 Angelo Heilprin, the geologist, published an account of human finds made at Hanson’s Landing, Sarasota Bay. The remains of a human skeleton were ob- served exposed in a low bank of ferruginous sandstone along the shore. The human bone was fossilized, being completely replaced by limonite. Heilprin’s attempts to correlate this sandstone bed in the geological scale were not successful, but he believed the bones to be inclusive in the bed and the whole to be of a great age. Joseph Leidy (1889), who was also interested in the Hanson’s Landing discovery, pointed out that the bones did not differ from modern human bones but withheld comment on the geological evidence. A critical note on the geology was injected into the discussions, however, by W. H. Dall, writing in 1887. It was his contention that rock formation pro- 30 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 113 ceeded at a much more rapid rate in this part of Florida than was generally believed. The fact that the natural springs of the Sarasota region contained considerable iron in solution was held to be the cause of the consolidation of gravel, sand, and shells into a sort of rapid-forming pseudocoquina. Dall noted that a potsherd had been found imbedded in such a conglomeritic mass of material in a find made near an Indian shell mound on the Webb estate. It was his further observation that the human bone remains reported by Heilprin from nearby Hanson’s Landing were encased in a similar formation and that they were of recent age. Hrdlitka tackled the problem of Early man in west Florida several years later (1907). After a careful description and analysis of finds from Osprey and Hanson’s Landing he ruled out the possibilities of a Pleistocene or early dating. Somatologically, he pronounced all the remains to be American Indian. His critique of the geology, supported by a statement from the geologist T. Wayland Vaughan, follows that of Dall. He held that the skeletons were all recent Indian burials that had subsequently, and within a relatively short time, been covered and impregnated with ferruginous sediments. That the bones are well within the range of those of American In- dians does not surprise us in the light of more recent developments in Pleistocene and early postglacial archeology in North America. This is to be expected. The case of the Osprey bones appears to devolve upon the geological interpretation of the find situations. The fact that they were imbedded in stone was the single factor in bringing them to attention. No distinctive artifacts of any kind were found in association with any of the skeletal remains nor were there correla- tions with extinct fauna. The evidence and arguments adduced to date indicate that the rock formation surrounding the bones could have developed in a very short time. For the present, the claims for Early man in west Florida are unproved. SUBSEQUENT WORK Following the studies of Moore and Holmes there was no concerted field program on the Florida Gulf Coast until the beginning of the Federal Relief Archeological Projects in 1933, but a number of minor investigations were reported upon in the interim period. R. D. Wainwright described some miscellaneous diggings at Palma Sola, Sarasota, and on Bokeelia and Pine Islands at about the same time that Moore was concluding his work (Wainwright, 1916, 1918). In 1923-24 Dr. J. W. Fewkes, assisted by M. W. Stirling, excavated WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—-WILLEY 31 a burial mound at Weeden Island, on Tampa Bay, for the Bureau of American Ethnology. This Weeden Island investigation proved to be of considerable significance as it became the basic datum or the type site of the Weeden Island culture. A collection quite com- parable to that which Cushing recovered from the Safford mound was acquired at Weeden Island (Fewkes, 1924). Shortly after the termination of the Weeden Island work, D. I. Bushnell made a brief survey of the Pinellas Peninsula and gathered a few surface speci- mens from several other sites in the same region (Anon., 1926). At the other end of the Gulf Coast, on Santa Rosa Sound, T. M. N. Lewis made excavations in a small burial mound located about 18 miles east of Pensacola (Lewis, 1931). Midway down the coast, near Crystal River, F. G. Rainey excavated what appeared to be a mound or cemetery on Buzzard’s Island (Rainey, 1935). In 1929 and 1930, several years before the Federal Relief Projects began, Stirling, as Chief of the Bureau of American Ethnology, re- turned to Florida and conducted a survey and series of small excava- tions at Palma Sola, Shaws Point, Horr’s Island, on the Withlacoochee River, and at Safety Harbor (Stirling, 1930, 1931). The latter site on Old Tampa Bay has become, like Weeden Island, a type station for a culture complex and period. As a follow-up to these personally directed field surveys, Stirling then maintained over-all supervision of the Smithsonian Federal Relief excavations at Perico Island and on the Little Manatee River, both in Manatee County, at the Thomas mound in Hillsborough County, and at the Englewood mound, Sarasota County. The immediate field supervisors in this work were M. T. Newman, Preston Holder, and D. L. Reichard (Stirling, 1935). These Relief Program undertakings of the Smith- sonian began in 1933 and were terminated in 1936. They are treated, along with the Weeden Island and Safety Harbor excavations, in the present paper under the section, “Excavations on the West Coast: 1923-1936.” After the Smithsonian withdrew from the field, the Florida State Board of Conservation continued Relief-financed archeology under the supervision of J. Clarence Simpson, of the Florida State Archaeo- logical Survey. Simpson made additional excavations at the Thomas mound and also conducted work at the Spender, Cagnini, Branch, Lykes, Snavely, Jones, and Picknick mounds. Very brief notices of this work have been published in the Biennial Reports of the Florida State Board of Conservation (see Anon., 1937 and 1939a). As a byproduct of his archeological interests Simpson has published an interesting and useful little summary of sources of stone in the 32 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 113 Florida Gulf Coast area which were used by the Indians in the manufacture of both chipped and ground stone artifacts (see Anon., 1939b, and Simpson, 1941). The present author and R. B. Woodbury made a surface survey and a number of stratigraphic tests in northwest Florida (Willey and Woodbury, 1942). This field work was done in 1940 and is reported on in full in this paper under “Excavations on the North- west Coast: 1940.” Following Stirling’s lead, the emphasis in this survey was upon sequence determination of pottery styles, the relating of these sequences to neighboring areas and sequences, and the anal- ysis and interpretation of previously excavated and published or un- published data from the Gulf Coast. A number of very recent surveys and test excavations have been made in the area. In 1946 J. G. Griffin and Hale G. Smith, working under the auspices of the Florida Park Service, conducted a quick survey along the northwest coast. This was done chiefly by way of examination of the area although surface collections were made at a number of sites. Griffin and Smith also made test diggings at the Lake Jackson site in Leon County and the Scott Miller place in Jefferson County (Smith, Hale G., 1948). Their findings at these sites should add much to the sketchy data presented in this paper on these two stations. As this report is being written Griffin is resurvey- ing the important Safety Harbor site in Pinellas County and open- ing new test trenches in the mounds and middens of that group. In connection with Griffin’s Park Service work it is hoped that R. P. Bullen will survey and study mound and village sites along the lower Chattahoochee River this coming winter (1948-49). Coextensive with Griffin’s program, Dr. John M. Goggin made surface surveys at Cedar Keys, Shaws Point, on the Withlacoochee River, and in Alachua County during the summer of 1947. Goggin had made previous trips to some of these regions in 1944. During the present summer (1948), Dr. A. J. Waring, Jr., explored sites in Citrus County and tested a midden heap near the mouth of the Chassa- howitzka River. This last site, from preliminary reports, may be one of the most significant in the archeology of the preceramic periods for Florida. The need for continued archeological research in Gulf Florida is recognized by all these men and others interested in the area (see Stubbs, 1940; Sleight, 1943). The establishment of a division of archeology within the Florida State Park Service (1946) and of a chair of anthropology at the State University at Gainesville (1948) indicate that present interests and research trends will be continued in good hands. WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—-WILLEY 33 Mention should also be made of reference material in the way of numerous collections and a considerable correspondence from various interested persons, many of whom live along the Florida Gulf Coast. These individuals are referred to in various places throughout this report. Some of this correspondence is now in custody of the Smithsonian Institution and many of the collections are accessioned in the division of archeology, United States National Museum. Other significant collections of Gulf Coast materials are in the Heye Mu- seum of the American Indian, Peabody Museum of Harvard Uni- versity, Peabody Museum of Yale University, the R. S. Peabody Foundation of Andover, Mass., the University Museum of Ann Arbor, Mich., the University Museum of Philadelphia, and the Florida State Museum at Gainesville. In addition to the actual field work, the period of 1918 to the present produced a number of summary and general papers on Florida archeology, all of which, to a greater or less degree, touch upon the Gulf Coast. Taken in chronological order, the first of these is Hrdlitka’s “Anthropology of Florida,” published in 1922. This study was, in part, based upon a field trip which the author made in 1918 around the southern end of the Florida peninsula ; but, in greater part, Hrdli¢éka drew upon Moore’s data and Moore’s skeletal collections from the various Gulf Coast mounds. As a synthesis, the work is principally concerned with physical anthropology rather than archeology. Another summary paper, but one more archeologically and ethno- historically oriented, was published by Rhea M. Smith in 1933. This is a competent statement of what was known of Florida prehistory as of that date, although there is no attempt at synthesis or integra- tion of the data into any scheme. In 1936, as a result of the several seasons of field work along the Gulf, Stirling published his “Florida Cultural Affiliations in Rela- tion to Adjacent Areas.” This short paper defined what the author considered to be the four major cultural regions of the State in terms of pottery wares. The method is comparable to that used by Holmes ; however, in addition to the trial definition of areas, Stirling also pointed to chronological relationships between two of the dominant pottery styles of the Gulf Coast, the earlier Weeden Island and the later Safety Harbor. This sequence relationship, the first for Florida in terms of described styles, was based upon associations rather than stratigraphy. Cross similarities between Safety Harbor and other late periods in nearby areas, such as Georgia, were established ; and, in addition to this, the associations of European trade artifacts with 34 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I13 Safety Harbor sites, but never with Weeden Island sites, gave added proof of the chronological relationship between the two. Other brief papers, following along similar lines, are those of the present author (Willey, 1945, 1948a), Goggin (1947b),® and J. W. Griffin (n. d.). Going somewhat farther afield are the papers by Greenman (1938) and Willey and Phillips (1944) which treat of Gulf Coast prehistory in relation to other cultural manifestations outside of Florida, and the syntheses by Ford and Willey (1941), J. B. Griffin (1946), and Martin, Quimby, and Collier (1947) which attempt to place the Florida Gulf Coast in terms of the total chronological and develop- mental framework of the eastern United States. The same approach, but from another geographical quarter, is seen in the papers of Gower (1927), Rouse (1940), Stone (1939), and J. W. Griffin (1943), where the authors are interested in the problem of West In- dian and Middle American influences into the Gulf Coast and other parts of Florida. The trends in recent archeological research along the Florida Gulf have been in keeping with those of American archeology in general. In short, the definition of the details of culture complexes and the tracing of these through time and space have been the primary ob- jectives. We are currently engrossed in this phase of prehistoric study in Gulf Florida, and as investigations progress it is possible to comprehend the local scene more satisfactorily against the larger background of eastern North America. At the same time, a new level of research effort is anticipated in that the archeologist will soon be in a position to go forward from the attained platform of time- space formulations toward an understanding of culture growth and process. THE ETHNOHISTORICAL STUDIES OF JOHN R. SWANTON As archeology in its most restricted sense gives us only a partial picture of culture history, a summary of the development of Florida prehistory would be incomplete without taking into account the im- portant strides made in the ethnography and history of the region. This ethnohistoric material has been compiled and interpreted in able fashion by John R. Swanton. In his long and distinguished career 6 Even more recently (spring 1948) Goggin (n. d. 2) has completed a Florida- wide summary treating with the relations of prehistoric cultures to their natural environment (Ph.D. thesis, Yale University). WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—WILLEY 35 in the field of Southeastern studies Swanton has consistently worked to bring clarity and integration out of a complex assortment of ethno- historic and linguistic sources. His two principal researches that deal with the Gulf Coast area of Florida (Swanton, 1922, 1946) are works of critical scholarship. In Gulf Florida as in other parts of the Southeast Swanton’s tracings of tribal movements during the sixteenth, seventeenth, and eighteenth centuries, his ethnographic descriptions, and his linguistic analyses have provided the starting points for the archeologists who are following him. eee Pace on - < ae f, | | . wigs je Taste, dalla og etl RU BE ees MT i ty Heese | Ww Ca a) ein Hi) oy inl ab eh OWE? ne Sat) 1 eae od % PLM pa edie deer tei Thaw teil nen a ML prs th ae hieeey MR ss eye vm me) Fide 4008 AR iol ge nh ea Ade Li An hg ithe A OL ith }a lh erga Ve Wl MPV inet S ray III. EXCAVATIONS ON THE NORTHWEST COAST: 1940 INTRODUCTION In the summer of 1940 an archeological survey of the northwest coast of Florida was conducted by the writer assisted by R. B. Wood- bury. These investigations were under the auspices of the Depart- ment of Anthropology, Columbia University, and were financed jointly by Columbia and by the National Park Service of the United Sa aee GULF BREEZE SOWELL N -- oe. ecw eeersy SCALE IN MILES Map 2.—The northwest Florida Coast, showing sites excavated in 1940. States Department of the Interior. At that time both the writer and Woodbury were graduate students of anthropology at Columbia. Taking the field on the first of June, surface survey and test excava- tions were pursued until mid-August. In this 2}-month period 87 sites were visited and studied briefly in the region between Pensacola and St. Marks; 6 of these were excavated stratigraphically (see map 2). As it is these excavations on which the ceramic and cultural chronology of Gulf Florida is essentially established, the present 37 38 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I13 section of this report demonstrates these several stratigraphies in some detail.’ Because of the crucial nature of this sequence data for the analyses and interpretations that are to follow, it is placed here, near the beginning of the report. The pottery types that are the principal media for measuring culture change through time, and that are listed only as names in the tabulations of sherd counts and percentages, are described in detail in connection with the cultural periods of which they are characteristic in a subsequent section of this report. Methodologically, this presentation is the reverse of the actual pro- cedure of investigations. In making the survey, the analyses of Moore and Holmes, the ceramic chronologies outlined in neighbor- ing areas, and the inferences to be drawn from a consideration of these were the first points of departure from which trial classifications and tentative groupings of the data were projected. These hypotheses were then checked by a seriation of the surface pottery collections from all the surveyed sites. The final substantiation and correction of the pottery periods and their chronological order were the test excavations which are presented here. The excavations were primarily concerned with ceramic stratigraphy and test pits were laid out at each site with this in mind. Burial mounds and cemeteries were not excavated. No burials were dis- covered in any of the refuse digging, nor were any structural features, such as house floors, post-mold patterns, or house mounds uncovered or explored. A very few nonpottery artifacts, mostly chipped-stone projectile points or tools, were recovered from the stratigraphic tests. These were too scarce in any one excavation or in any one site to compute stratigraphic percentage trends. Together with similar artt- facts gathered in the surface survey, they are identified as to period and incorporated in the several descriptions of the section on “The Culture Periods.” CARRABELLE, FRANKLIN COUNTY (FR-2) Description of the site—In his 1918 report Moore (pp. 557 ff.) describes what he calls an “Aboriginal Cemetery near Carrabelle”’ : About 1.5 miles north-northeast from the town of Carrabelle, on the Gulf Coast, is a low ridge . . . covered with scrub and scattered pine, having its 7 The principal type sherd collection, together with the nonceramic artifacts from the 1940 survey, are stored in the archeological laboratory, Department of Anthropology, Columbia University. The remainder of the materials are at the National Park Service Museum at Ocmulgee, Macon, Ga. WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—-WILLEY 39 eastern extremity almost enclosed by a small, shallow, fresh-water pond some- what in the form of a horseshoe. This ridge is of white sand on the surface, darkened by vegetal deposit and the charcoal of fires that have spread over it. Below the white sand, which is from four inches to one foot in depth, is yellow sand of uniform shade. Moore cut a trench through the east end of the ridge and discovered several deposits of whole and broken pottery on the yellow sand or in it. No material was found at a greater depth than 20 inches. The remains of two cremated human burials were also found in the yellow sand. No other human skeletal material was uncovered. From this description, which is rather unlike any of the other sites that Moore explored along the Gulf Coast, it appears as though cremated burials and mortuary deposits of pottery, both broken and “killed” by perforation, were placed in the natural sand of the ridge at superficial depths. Possibly a low artificial sand mound was built over the burials, but this cannot be satisfactorily determined from Moore’s data. That Moore was correct in his assumption that this site was a cemetery or burial place seems to be borne out by the absence of any amount of black detritus or shell-midden refuse.® In the 1940 survey some test excavations were made at a midden site on the edge of the city of Carrabelle (map 3). At that time it was our opinion that the site which we excavated was not a part of, or near to, the site described by Moore (Willey and Woodbury, 1942, p. 238, footnote 45), but in again reviewing the notes it seems at least probable that our site and Moore’s are the same. The site is about one-half mile from the present center of town in an easterly rather than a north-northeast direction. It is located on a low ridge about 75 meters back from the Gulf beach. The northeastern end of this ridge is surrounded by a swampy lowland, tallying in some respects with Moore’s description of a “shallow, fresh-water pond somewhat in the form of a horseshoe.” Because of modern develop- ment we were unable to excavate in the eastern or northeastern por- tions of the ridge; these are now house sites of a pleasant suburb. Fortunately, on the west side of the ridge, on the Keith property, there was a vacant area, and the northwest corner of the ridge was similarly unoccupied. Permission was obtained for excavations in these portions. Test digging here, as well as an examination of the banks of a railroad cut (see map 3) which has bisected the ridge, revealed abundant black earth and shell midden. 8 See section on “Review and Analysis of Gulf Coast Sites,’ Carrabelle Site, pp. 267-268. 40 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 113 SCALE W METERS Map 3.—The Carrabelle site, Franklin County (Fr-2). WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—WILLEY 4I The general conformation of the site accords with Moore’s descrip- tion, and it seems likely that Moore’s direction of “north-northeast from the town” was in error. If the site was located on the Gulf beach, within 1 or 2 miles of the city, it could not possibly have been in a northerly direction, and the location which we visited is the most likely spot. The difference in the nature of the deposits which Moore encountered and those which we noted might well be due to the fact that his excavations were confined to the eastern or northeastern section of the ridge which had been reserved by the Indians as a burial place. Our observations and excavations, on the other hand, were contained within the old habitation ground. The Carrabelle site, as it appears today, is situated on a ridge paralleling the beach for about 200 meters. The ridge is approximately 75 to 100 meters wide. It lies just above a paved subsidiary high- way which connects as a small side loop with U. S. Highway 319. This highway cuts along the southern edge of the site, between the ridge and the beach, exposing shells, midden, and sherds. The modern houses referred to are built on the crest of the ridge, and back of the houses the railroad cut slices through the ridge on a northeast-southwest axis. In most places the refuse averages about I meter in thickness. On the western edge of the ridge the ground drops away rapidly to a small drainage which empties into the Gulf, passing first under a railroad bridge and then through a highway culvert. A little side road comes into the highway from the north, following along the west side of the ridge and coming up over one corner of it just before it joins the pavement. The midden, shell, and sherds did not extend the full length of the ridge but occupied the western or southwestern half of it. Cul- tural debris was also found in the flat between the ridge and the beach, and some sherds were picked up as far away as the high-tide line (see map 3). A surface collection of pottery was gathered from the railroad banks, the highway cut, the beach flats, and the top of the ridge on both sides of the railroad excavation. The low, marshy area to the north and east of the ridge was covered with deep grass and other vegetation so that surface pottery would not have been visible; however, it is extremely unlikely that the site occupation ever extended down into this terrain. The sand ridge on which the site is located is an ancient dune formation rising above the surrounding lowlands. Several hundred years ago, at the time of Indian occupation, the surrounding marshes may have been flooded regularly by the tides. Even in more recent times, as Moore has remarked upon, water stood in the basin at one end of the ridge. 5 42 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I1I3 Excavations—Tests pits I and II and the two supplementary excavations were put down in the vacant lot near the west end of the ridge. There is no record of modern building in this space, and the spoil dirt from the railroad excavations has not been thrown back over the banks but taken a few hundred meters farther to the west to serve as fill for a low gulley. All other indications were that the small western section where excavation was possible was an undis- turbed deposit. The four tests were dug contingently so that when the work was completed they formed one large excavation (see map 3). Pit I (a 3- x 3-meter pit) was oriented to the cardinal directions, and pit II was later excavated immediately to the west. The supplementary excava- tions I and 2 were made prior to the excavation of pit II, and they were for the purpose of exposing pit II as a stratigraphic block on the north and south as well as the east side. Both pits and both supple- mentary excavations were excavated by arbitrary 10-centimeter levels. The first level of pit I (0-.10 meter) was composed of packed shell and very black midden with abundant sherds. In levels 2 and 3, down to .30 meter below the surface, the midden concentration less- ened, but a gray, charcoal-stained sand continued to yield sherds. In level 4 (.30-.40 meter deep) there was a continued decrease in midden richness and color. Shells were scarcer although all types, including oysters, small clams, conchs, and scallops, were still found. At the bottom of this level small, dark organic concentrations re- mained in the southwest and northeast corners of the pit. The final level, 5, was excavated from .40 to .60 meter below the surface, and exhausted this refuse. The upper portion of the level was of gray sand, slightly stained with organic and charcoal color ; the lower part of the level was clean red-brown loam, completely sterile except for a few shells. The southwest quadrant of the excavation was taken on down to .go meter below the surface before closing the pit. Supplementary excavation I, on the south side of the isolated stratitest block that was to be pit II, was excavated in .10-meter levels down to .80 meter below the surface, a total of 8 levels. It differed from pit I only in that the midden and shell concentration at the top were a little deeper. Its surface dimensions were I meter north- south by 4 meters east-west. Supplementary excavation 2, on the north side of pit II, had the same surface dimensions as supple- mentary 1. Supplementary 2 was not as rich, and the last level, 6, yielded only one or two sherds. The excavation was continued to .80 meter, terminating in sterile red-brown loam. WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—WILLEY 43 Pit II (3 x 3 meters) was excavated to a depth of .80 meter, being removed in 8 levels. Carefully working in from the three sides which were exposed by pit I and the two supplementary excavations, the removal of the levels was well controlled and an accurate picture of the physical strata obtained (pl. 1, top). Soil conditions were very similar to those of pit I. Shell and midden concentration at the top gave way to midden-stained gray sand and finally to sterile loam. Profiles of the faces of pit I and pit II are shown in figures 1 and 2. They may all be generalized as follows: a top stratum of hard- packed shell and black midden, varying from .10 to .30 meter in thickness; a secondary stratum of gray sand and midden; in some places, a third stratum of lighter gray sand with only a little organic coloring ; and basal reddish sand or loam. There are a few intrusive pockets reaching from either the midden concentration or the gray, sandy midden into lower strata, but these are not numerous enough or do not appear to be large enough to have greatly churned the deposits. The dip and strike of the strata are slight although the various beds have small irregularities. The soil profiles are those of the average thin occupation site for the Florida coastal area. Packed debris has mixed with the natural gray top sand of the region and has stained the sand for several centimeters in depth. The pure rub- bish concentration is a small band at the top. Test pits III, IV, and V were all located on the north side of the railroad cut in a small grove of trees. On this side of the tracks there is no present-day occupation. All pits were put down in the crown of the slope in the deepest part of the refuse. Pit III (a 3- x 3-meter pit) (see map 3) was excavated down to .80 meter below the surface, each level .10 meter in depth. The top level (o-.10 meter) had very little shell, a number of sherds, and consisted of brownish midden mixed with humus. Levels 2, 3, and 4, down to .40 meter deep, passed through black midden rich in sherds but with only scat- tered shells. At the bottom of level 5 (.40-.50 meter deep) thick shell deposits were mixed with the black midden, and a fire area was noted. Shell continued through level 6 (.50-.60 meter deep) but diminished in level 7 (.70-.80 meter deep). In the bottom level (.70- .80 meter deep) there was little shell and few sherds. The bottom of the pit showed clean gray sand dotted with a few organic spots which might possibly be post molds. A small additional test revealed that red-brown sand underlay the gray sand at a depth of .go to I meter (see fig. 3). 44 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. II3 DEPTH ET ae / S oM Oo BLACK AMIDDEN ANDO SHELL GRAY SAND AND MIDDOEN .20 ‘ AND (MIDDEN. | STAIN \ ' Veet) ee ~ = RE0D0/SH SAND .40 -6O REDOISH SAND BLACK AWDOEN AND SHELL ta ta Lt Ly / eS 2.70 M Oo BLACK MIODEN AND SHELL 2O 40 GRAY SAND AND AG GRAY SAND ANDO MIDDEN STAIN .80 REDOISH SAND. WEST FACE Fic. 1.—Pit I soil profiles, Carrabelle (Fr-2). WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—-WILLEY 45 Pit IV was a small (1.50 x 1.50 meters) test made just a few meters to the northeast of pit III. Levels 1 to 6 (0-.60 meter deep) penetrated mixed rich black midden and shell with the shell being slightly more concentrated toward the bottom of this zone. Level 7 was made 20 centimeters deep, going from .60 to .80 meter, and in this level gray DEPTH 1N METERS / 2 JIM. o a6 BLACK MIDDEN AND SHELL 40 GRAY SAND AND MIDDEN, GRAY SAND AND MIODEN STAIN REDDISH SANO NORTH FACE 60 60 BLACK MIDODEN AND SHELL GRAY _SANO AND ANNDDEN GRAY SAND AND_NM/IDDEN STAIN REDOISH _SANO SOUTH FACE / 2 3M 1) Bo BLACK MIDDEN ANO SHELL GRAY SAND AND MIDDEN 4O GRAY SAND ANO MIDDEN STAIN 6O CLEAN GRAY SANO 80 ees WEST FACE Fic. 2.—Pit II soil profiles, Carrabelle (Fr-2). sand began to replace midden and shell. The pit ran out into sterile sand in level 8 (.80-1.00 meter deep). Pit V, another 1.50- x 1.50-meter test pit, was located about 25 meters south and west of pit III. The refuse at this point was ex- hausted at the bottom of the 6th level (.50-.60 meter deep). Sherds were plentiful except in the bottom level. There were almost no shells in the pit. A brown midden and humus was slightly less than .30 46 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 113 OLPTH IN METERS / 2 3M Oo BROWN AIIDOEN -20 40 BLACH ANDDEN SHELL AND MIDDEN, -60 GRAY SAND AND MUIDDEN STAIN -6O EAST FACE -2O0 / Pe IM ae BLACK AIWNDDEN .40 60 BLACK AMIDDEN ANDO SHELL GRAY SAND ANDO Yo) AUMODEN STAIN NORTH FACE / 2 280M o -20 BLACK AMDDEN 40 BLACK ANMDDEN AND SHELL 1 ' DARA GRAY 1 STAIN | GRAY SAND AND MIDDEN~STAIN WEST FACE / 2 v BROWN AMUDDEN BLACK M/DDOEN GRAY SAND AND AIODEN STAIN i ‘ ‘ ‘ M -60 -d0 SOUTH FACE Fic. 3.—Pit III soil profiles, Carrabelle (Fr-2). WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—WILLEY 47 meter deep; black midden extended from here to the .50-meter point and gradually shaded off into gray and then reddish sand. The soil profiles on the north side of the railroad track differ from those on the south side in that the midden concentration is deeper and can be divided into an upper brownish zone and a lower black zone. Shells, though not as abundant as they are in the top two levels of the pits south of the railroad are fairly plentiful in the lower portion of the black midden zone. Below the midden and shell there is a zone of gradual transition from the dark occupation layer to the natural sand of the ridge. (See fig. 3 for profiles of pit III.) As in pits I and II, there is nothing in the profiles of pits III, IV, and V to indicate an interval of nonoccupation for this part of the site. Stratigraphy—tn pits I, II, and the supplementary excavations there is a marked complementary relationship in the vertical distribu- tion of the pottery types. This is keynoted by the two types, Swift Creek Complicated Stamped and Deptford Linear Check Stamped. (See tables 1 and 2 and the graphs on fig. 4.) In pits I and II, Swift Creek Complicated Stamped is 40 percent of the total number of sherds for the top levels. There is a gradual diminution of the type toward the bottom of both pits. In pit I Swift Creek does not occur in the two bottom levels, and in pit II it occurs in those levels only in very small quantity. Deptford Linear Check Stamped, on the other hand, shows a numerical and percentage increase in the bottom levels of both pits. Compared to Swift Creek, there are not, proportionately, many Deptford sherds in either pit. It is extremely significant that those few which are present should be, with few exceptions, in the lower levels. The supplementary excavations show approximately the same percentage changes in this regard as the test pits. Occurring in small percentages in the same levels as the Swift Creek Complicated Stamped are three other complicated stamped types: Crooked River Complicated Stamped, New River Compli- cated Stamped, and St. Andrews Complicated Stamped. On the whole, their sporadic distribution in the various levels seems to indicate their temporal affinity to Swift Creek. Typologically, they are also more closely related to the family of complicated stamped wares rather than to the Deptford pottery. The supplementary ex- cavation I, vaguely suggests that New River Complicated Stamped is a little earlier than the other two. Paralleling the vertical distribution of Deptford Linear Check Stamped is Deptford Bold Check Stamped. This is best shown in pit II. However, the other pits show Deptford Bold as being well SISAL CILYYOIIS ONYTIS/ NICGFIIM CISW ELS SGWOD SMIVYONY LS OFALVOLS SWOD HIFYD LIIMS PP Sy aT ST SURES NU te PDISV OLS pr —————eeeEEEEy HIFTHD G70F GYOSLASII CSIASALVELS YIFHD YPINIT GYOSLAIC LEVELS SCALE /00% Fic. 4.—Graph on pottery stratigraphy, pits II and III, Carrabelle (Fr-2). 48 ARCHEOLOGY OF FLORIDA GULF COAST—-WILLEY 49 WHOLE VOL. teeers Te10} Wg iste ABA 300 aatels 5000 oz* 09" ap00 Sens ASOD ad08 gos atciste oe" cietole Briere ee eeee eee seer eee seer I ¢ eoee eeee eeee seer eeee ener I eons eeee eee 9 6z0" 2Zo° rIo° eee eore SATs ges eee fro eeee FIO" £to° 6z0" £to* eeee weer eee z ¢ 11 eeee eee ZI Ze eee i eere I € tA £ eeee ener eee S$ 9£0° 6£2° eee eeee Io* gor* 902" IZzo" Zo" ener Stor ener 2fo° §2e" eeee eee eee Z €z eee seer I OL 61 tA £ eree v eree £ 1z ener enee eee v 6II* £62" bSo° eeee eeee oro’ eeee eeee seee Izo*° 250° oro" S90" 16£° eree eeee eee II lz Ss eeee eevee I eeee ener eree g +, I 9 gt ener eeee weer £ $zo" gos 650° voo" essvele 910° 210° seas g00° 9r0° 910° asters 1Z0° Lor seueke ¥oo" goo" 9 tZ PI 1 eere v ¢ weer Zz v v ener S QIl eeee I z z tzo" Z9¢° 220° OHSY oe S10" £00" I00° £00" $00" I00* £00* 0zo* bOE* 6£0" 100° 200° £1 £sz ZI stores seis 8 z I z £ I z II goz z I v I aqfeqesse) “TI Wd serees 1e}0} Hg O6r’ eee eeee Ltro’ gro" eri’ f2S° eeee Lto’ ener eeee eee eee eeee eee weer wee v eeee eoee I I £ II eee I eoee seer onee see eee wees wane sees S Tir’ eee eee weer eeee eoee cos: eeee Zee’ enee ener see Tite eeee eeee eee eee I eeee eoee eee eeee eeee S$ eeee Z eeee eens sere I ewer eeee eoee eeee v tro’ 02* 1QVae ener ener tro’ 892° 220" 2zo° eeee 990° 990° tto’ 990° ener eee eee CA 6 S$ eee eree A £1 Vt I eee v ¢ A £ ener eeee see £ 6z0" Zos° 490° ener 600° 6z0° 600° oeee 600° 600° 6zo* 600° gf" ore’ ener eee ener £ ze /4 eee I £ I eree I I £ I v 0) 4 enee eee eter z GNC bre’ rSo° BORG too" too" Foo" goo" goo" $zo" goo" sieisie 1Z0" Zgt° OP an sforeie too" eee zg 1 4 eeee I I I z Z 9 Z eeee S gir eeee eee I tT ayfoqesiey ‘I Hd Gi - g chad ny AY els az ne? ay noe nas na? noe wD re we s[PAo'] 3 ae 8 p Sy ihe ro one ies a 8 Bg BSa BSS Roe EF 2.0 Do bo. a 2 Bo BG Get. Gem tach «8 Bee Bee ope BES gEP BS rp et oh eS A Boge ro 9 be Qa Pi 28 UH 50.0 2 ot o Bo B Fy B A 5 ce =a Sse zo elon hebieiges Se heyon SEN@l ais r=¥i=} 5B es a og © €& wh we go 88 B O85 864 28 888 #2 — ® tg re a n B by Bt no. fee & aR steel Lo oe 7) a Qs Xp Be me 36 Se. Ley x as o oa ow ap a ae 5 ) BS BS ep a? A. Ad ro 5 S 5 3 4g a ve a A. oO Qu arn (Sole} UL UdDATS JOAd] Aq 9dUaIINIO 98ej199.9 J ) (2-4) ayjaqnaavy ‘TI pun 7 sng fo uoyvoifissyja paayS—t1 Av VOL. 113 SMITHSONIAN MISCELLANEOUS COLLECTIONS 50 a fz teres e103 Hq wees oF: wees z oeee eee bSo° £ eoee /4 wees eeee SI gfo° && v ++ 1230} Hq eee C2: eens z peyissepuy uel [ENpIsoy ure[g y}OoUrS (2-47) ured SUOLUIS 4S ayjaqnadDy) ‘2 puv I Suoyvavaxa Cavjuamasgdns fo uoypoyisspj2 paayS—z% AIav]L pedureys ajduig ps0zdeq £ ZI I4o* £ S 9zo° 8 z eeee FIO’ SOO I Izo° zo’ £ S eeee sees ‘ UOIPLACIXO €£o° I bro I eee sees eens eee sees eee Lto: I gto: £ gro" z eeee eens 600° I Axequswia,ddns aeee £10° I tI0° I vIO* (4 aJeqesiey ‘I uoeAvoxe Arewuoweiddng pedureys yoeqa Plog proysdoq pedureys yooya ivsury ps0jydoq (SOI[e@} Ul UdATS JaAo] Aq 9dUdIINIIO 99e}U9919g ) peyseul-p10d eprso],y ISM, pedure3s 2249 FD ure[d UrpUes yy peduieys esoy eyes pedureys pezeorjdu0y) smoipuy 4S pedurrys pezeorjduroy IOATY MONT foes fees sees nog Bee = ° 35> 3 ao aye ae ag g n NOs reiledieh yor co hie} h@) Bea [ory HS 2.8 ee as eS a, p B [ory UIE purysy uspos oT wh WO s[eAo'T WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—WILLEY SI scattered throughout the pits from top to bottom and not showing the high frequency for the lower levels as is the case with Deptford Linear. Gulf Check Stamped seems to occur sporadically and at all depths except in supplementary excavation 2 where it occurs only in the three bottom levels. West Florida Cord-marked and Deptford Simple Stamped are found occasionally in these pits, but neither type is numerically strong enough to give any patterning of occurrence by levels. Four sherds of St. Simon’s Fiber-tempered Plain ap- peared, three of these in the lowest levels of pit I and supplementary excavation 2. A single fragment of the distinctive type Santa Rosa Stamped came from level 6, a little over halfway to the bottom of supplementary excavation I. In both supplementary excavations, and one of the test pits, scatterings of Weeden Island Plain, Weeden Island Incised and re- lated incised and punctated types were found in the topmost levels. Smooth Plain and Franklin Plain were also found toward the top but were not quite as superficial as the Weeden Island types. Residual (sand-tempered) Plain was found in all levels, but it definitely had its highest frequency at the top and faded toward the bottom. Pits III, IV, and V, on the north side of the railroad cut, averaged a much greater number of sherds per excavation than pits I and II and the supplementary excavations. The midden being somewhat deeper undoubtedly accounts for part of this, but as pits III, IV, and V are all in Weeden Island Period refuse, the greater abundance of pottery in this later period is also a factor. In examining the pottery count data (tables 3 and 4 and graphs on fig. 4) for these excavations one is struck with the complete absence of Deptford Linear Check Stamped, the marker type for the Deptford Period, and the only occasional presence of Deptford Bold Check Stamped. Swift Creek Complicated Stamped is found in all levels of all pits, and maintains an inverse and complementary frequency relationship to Weeden Island Plain, Weeden Island Incised, Carrabelle Incised, Carrabelle Punctated, Tucker Ridge-pinched, Keith Incised, and Indian Pass Incised, all types of the Weeden Island Period. Smooth Plain is found in almost all levels with some indication of its being a little more common in the upper than in the lower part of the midden. Residual Plain follows the same trend as in pits I and II, increasing toward the top. In the Weeden Island levels of pits III and IV this is more marked than it was in the Swift Creek Period levels of pits I and II. Residual Plain is 60 to 70 percent of the total sherds for the top three levels of pits III, IV, and V. VOL. 113 SMITHSONIAN MISCELLANEOUS COLLECTIONS 52 o 6 o peyissepou yr TIO) "Teo? Wg eee eeee be: Sz weer 9 seer eee os: oor Cte pfs ener eee ge oS sine JOT, eeee eeee Clb: 0g weer gt weeee 600° oLb* got I 1S FOI goo* so" Séa0 I gor eee eee gEL° 88 eeee So eee seer rZ9° goz piltite “OOe: 1zo'% ***- 7e10} Wg seer eee or: $ eee z ZI eee eSé: 991 sere ze: eeee eeee es gge Soo 62&° ure [ENpIsoy poy werg ure} YOO pedureys yxoeyD Plod ptosideq (2-4,]) seee eeee seer weee wee see to- bo: 9S° see see eee seer eee wae seee I I PI seee weee eeee sees ween eeee eeee £o° eee 9S° see eeee saee eens wees seer eee £ eee 9s seee weer weee eeee eeee weer weee 2Zo° weer vS° tere eee seer eeee eeee eee eee I eee lz seer eeee weee eeee eeee eee eee Czo* eee or: eeee eee ower seee eeee eee eee z eeee gt wees weer eee eeee sone seer weee wees 600° seer eb: seer wee teee eeee seee eeee goo" eeee weee gro" eeee OFI* gio" weer seae seer seer steer I eeee ener £ ener ve wiles. eese veces Roon coe eee cece ‘e+ $£0" Stor AOD sone b eeee eee eee seer eee eee eee eee ¢ wees ener eee eeee eee ener seer eeee go° £I0° eee £00" ewer seee eee eeee eeee eens eeee seer tz Vv eeee I ayeqe1iey ‘TIT Wd pedwrys 2249 FIND pestouy noseg ulseg pedur3}s pezeorpdur0o+) pedure3s peyeorjduroy peyourd -oSply 1syxON EL, pesiouy sseg uvipuy IAT pexooig peduirys pedureys xoaqD teourTy ptoyydaq peyteus-j9 Ny Pi?fal PUNOW peyxseu-p10y Epi ens urelg wIppuRsy SMOIPUY 3S peyeorjduio,) oD ES pasiouy yoy (SOI[e}I Ul UdATS JAAD] Aq 9dUdIINIDO 98eJUIIIOg ) apyjaqvasvy ‘AT pun yyy snd fo uorpoifisspja paaysS—e€ alavy weer eeee I eee cr weee eeee eeee pezejoung aypeqesrreg pestouy ayjeqesieg) pesrouy pursy Uspes\\ sae tees eee eee ure[d purest waped SR) i NS EIN oo: Se eT NS S[PAIT WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—WILLEY 53 Of the Complicated Stamped types, other than Swift Creek, that were seen in pits I, II, and the supplementary excavations, only St. Andrews Complicated Stamped is found in pits III and IV. In neither pit is it found in the two upper levels. Six West Florida Cord-marked and two Mound Field Net-marked sherds are in level 2 of pit III, but no others of these types appear. The generalizations involving ceramic sequence that can be made from the Carrabelle excavations are these: In the group of pits south of the railroad cut (pits I, II, and supplementary excavations 1 and 2) Deptford Linear Check Stamped, in rather small amounts, underlies, and is earlier than, the main bulk of the Swift Creek Com- plicated Stamped type. Occurring in small amounts, Deptford TABLE 4.—Sherd classification of pit V, Carrabelle (Fr-2) (Percentage occurrence by level given in italics) a] ae =e c f r= =) Sef fen ¢ 4 8 aa “oO Se | “_ Sy 3 D a oD or Aa a ay ® m os = =2 285 oH Ay S Ss F £ so 33 as Os & Sis s § a Total 73.8 2 2 39 2k ws ° 4s] a herd SS aa = SC e 3 SAE 3 a a a = v & o Ss Levels 3% So On olay no ge a % 5) level I 2 2 A006 I 10 AAc 5000 47 62 032 -032 : 016 I6I C -758 sisare 2 2 2 I I 4 20 39 -051 -051 ~025 -075 I02 -743 acta 3 2 : 6 II 2 84 I II5 077 -017 052 095 O17 -730 008 oon 4 2 Lae 25 I 41 arate 74 067 027 -337 013 -546 seve 5 14 2 15 2 33 424 060 -454 060 Zev 6 5 sareve I 5 II -454 ante -090 -454 eee itstotaler.q.ee. 334 Linear Check Stamped and Deptford Bold Check Stamped are found in most all levels of all excavations. There is a suggestion that Deptford Bold Check Stamped has a longer range upward in time than Deptford Linear Check Stamped. Also, Deptford Bold has an earlier inception than any of the Complicated Stamped types. Three complicated stamped types, Crooked River, New River, and St. Andrews, are contemporary with Swift Creek in pits I and II and related excavations. In the group of pits north of the railroad track (pits III, IV, and V) Weeden Island types are superimposed upon Swift Creek types in much the same fashion that Swift Creek overlies Deptford in the southern excavations. Of the minority types, New River and Crooked River Complicated Stamped do not occur in the northern group of pits. St. Andrews Complicated Stamped does occur, as 54 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I13 does Deptford Bold Check Stamped. As there is little doubt that the rubbish sectioned by the northern pits is later than that of the southern pits, there is an inference here that St. Andrews Com- plicated Stamped is slightly later than New River or Crooked River Complicated Stamped. Also, Deptford Bold Check Stamped has a clearly later time range than Deptford Linear Check Stamped. Con- versely, Gulf Check Stamped, of which only one sherd occurs in the northern group of pits and that in a bottom level, would seem to be temporally related to the Deptford Period and the earlier half of Swift Creek. Franklin Plain belongs almost entirely to the earlier half of the Swift Creek Period. Smooth Plain is late Swift Creek and Weeden Island in time; and Residual Plain steadily increases from almost no occurrences in the Deptford Period to a numerically predominant position in Weeden Island. The position of West Florida Cord- marked and Mound Field Net-marked pottery is indefinite, but the small amounts found seem to be later than the Deptford Period. Although Fort Walton and Wakulla Check Stamped sherds were found in the surface collection from this site, none were recov- ered from the excavation. Classification of surface sherds from Carrabelle Fort Walton Complex Fort Walton Series: Fort Walton Tncised.. .c 308i a. Ssudes ce eetes cee 6 ake sJacksonWelain...s<% daste. pS 2.6 e.8 60 as = om fp 3 [ov UIeTd pue[sy Uaps9 MM © St. inl OSD) > 68) s[eaoT WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—-WILLEY 63 Incised and the two Carrabelle types. Crooked River Complicated Stamped and St. Andrews Complicated Stamped, related in tech- nique of ornamentation to Swift Creek, have a somewhat inter- mediate distribution in the rubbish, not appearing at the top or bottom. The plain types, which make up the bulk of the pottery, per level, do not show clear-cut vertical distributions. Weeden Island Plain occurs all the way to the bottom, although it is possible that it de- creased slightly from top to bottom, an action paralleling Weeden Island Incised and related types. It would appear that Weeden Is- land Plain, at least as far as the Mound Field midden is concerned, had its inception somewhat earlier than its related decorated types. Smooth Plain shows no significant percentage changes or trends from top to bottom. Residual Plain, the category which includes all undecorated, sand-tempered body sherds, does decrease gradually from top to bottom. This decrease can most easily be attributed to the opposite and complementary behavior of Swift Creek Compli- cated Stamped. The Swift Creek mode of decoration was, in many cases, an over-all application of stamped designs to the vessel, while Weeden Island and related types restricted the decoration, as a rule, to the upper portion of the vessel body. This gives a greater plain sherd count for Weeden Island Period levels. Pit II is little different from pit I. The same relationships obtain between the incised types and Swift Creek Complicated Stamped. There is a suggestion that the rubbish of pit II began and ended a little later than that of pit I. Only in the bottom level of pit II is there a really high occurrence of Swift Creek, while Wakulla Check Stamped, a type characteristic of a later phase of Weeden Island, occurs in small numbers in the three top levels. In brief, the Mound Field midden excavations indicate a Swift Creek Period and a Weeden Island Period occupation, falling in that chronological order. Classification of surface sherds from Mound Field Weeden Island Complex Weeden Island Series: Wicererulsl ania EncisGds ai Sthitive-sthercisie sete clare esieee cienerele 2 Weedencislandy Plgath. 5 a1}. x ds'd viele Says wn cisteiaiereln tg om aie Senate 13 Santa Rosa-Swift Creek Complex Complicated Stamped Series: Switt Creek: Complicated, Stamped o's. Fes, cceke ec ce ele alee 4 Crooked River Complicated Stamped................... I St. Andrews Complicated Stamped ..c... 00600 cs ceccc ene I 64 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I13 Miscellaneous Smooth: Plain % 2 2.473 Gah epeiacok Rac nee crac 2 Residual Plain’ 35.2 ace see etiam e holes. © ere est cicens 106 Total sherds.... 134 SOWELL, BAY COUNTY (ByY-3) Description of the site—The Sowell midden and burial mound are on the west side of St. Andrews Bay, about 1 mile west of Bear Point, Bay County. This location (map 5) is approximately 60 and 70 miles west of the Carrabelle and Mound Field sites. The midden and mound are among several in the immediate vicinity. The mound was excavated by Moore (1902, pp. 167 ff.) who de- scribes it as follows: The height of the mound was 4.5 feet; the basal diameter, 50 feet. A great depression whence the sand for the mound had been taken was at its southern margin. Moore recovered several burials accompanied by a mortuary de- posit of pottery.1? He makes no mention of the associated shell midden. The midden site, which today is clear of scrub oak, fronts along St. Andrews Bay and lies a little over 1.5 miles below the new St. Andrews Bay bridge of U. S. Highway 98. The area of shell and potsherds extends back from the water’s edge for 40 or 50 meters. Along the bay shore its extent is 100 to 125 meters. The mound site excavated by Moore lies a few hundred meters distant in a dense scrub oak thicket. E xcavations—Unfortunately, the midden site at Sowell had re- cently been gutted by commercial shell-gathering operations, so that our 1940 excavations were limited to a very few places in the site which had not been disturbed. Pits I and II were both very shallow (see map 5 for pit locations). Both pits were 3 x 1.5 meters in surface dimensions. Pit I was excavated in three levels, the first two of Io centimeters each and the last of 20 centimeters, mak- ing a total depth of .4o meter. The pit was located on what had apparently been the edge of the thickest shell and midden deposit. The two upper levels showed sand mixed with fragments of shell and gray organic stain. The bottom level gradually shaded off into sterile reddish-brown sand which gave no evidence of former dis- turbance. Pit II was carried down in five 10-centimeter levels. It was almost identical to pit I in the nature and depth of the refuse. 10 See section on “Review and Analysis of Gulf Coast Sites,” Sowell site, p. 231. WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—WILLEY 65 Both pits averaged about 40 to 50 sherds per level except for the bottom levels. Pit III (1.50 x 1.50 meters square) was located in a part of the AREA OF COMMERCIAL EXCAVATIONS ee =_— ~ (e) EDGE . OF BANK y BLUFF 2, NO. 2" Hf USGS ~ BENCH -MARK qyniol =_ yit BAY pb SLUFE 2' ser-up" USGS BENCH-MARK _- 1! = BE MAINING UNEXCAVATEO ANDREWS 3 a pir st ’ ' . ST. SCALE IN METERS Map 5.—The Sowell site, Bay County (By-3). midden that appeared deeper than those sections where pits I and II had been placed. Some small trees at the spot had discouraged the shell excavators from clearing out this portion of the site. The first two levels of pit III (o-.20 meter) yielded over 100 sherds and 66 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. II3 the soils were a mixture of humus, broken shell, and sand. In levels 3 and 4 (.20 to .40 meter deep) sherds continued, but the sandy refuse showed very little shell or organic material. In level 5 (.40 to .50 meter deep) the soil was sandy, and there were very few sherds. Thinking the midden was giving way into undisturbed sand, the sixth level was made deeper (.50 to .80 meter deep), and, rather abruptly, dark midden and compact shell showed up again with an increased number of sherds. In levels 7 and 8 (.80 to 1.00 meter deep) the darker midden and shell continued, but in level 9 TABLE 6.—Sherd classification of pits I and II, Sowell (By-3) (Percentage occurrence by level given in italics) 3 % cue g 8 s 8 as «= “4 = 7 S n 7) 77) o oO ied I Ay vo Sq Sg a 5 a0 do se S22 = 3 5 os = = =| Sas 9 88 88 Ge bo ste & © aimee He pao gS es ee ee teh ces 8 ia ae wo o Levels Go sto ae Se ee Oe ee ee an [a4 P level Pit I, Sowell I 6 2 2 3 eaten I 13 elateye Ba00 48 I 76 074 -02 -025 CE» Sanae -012 aS B008 soda -631 -012 sce 2 ae sa0¢ act ae Sere By ogac ayatate 26 2 40 alta’ sane pane sao: 005 Gnac AAC = anon Ss08¢ +650 .050 S00 3 S00 % 9005 coud ye ound I 3) i eiseas 8 SARS. 6ad0 eisie S375) Wetels at BEVE Stic Pit totall eee et Pit II, Sowell I 2 I eats I 2 600 II meats 5 61 I 84 1022)" s0IT Saxe SOIT) 022) sane Sey Asoo 3059' .072) =. on one 2 I eter I seis : 10 S050 TQ) “swiss 35 028 : 028 114 ~285 Steir 542 ABar 3 ae Ac de 3 4 i 28 36 S00 OF 084 eee -027 V7 4 7 I areas 3 Seles 2 I Aonn 38 I 66 - 106 .O15 +030 5000 -045 raters -030 -O15 shioc *575 -O15 5 solar I 2 I 5000 non6 I I soon 19 3 28 -O7I AICO daype nag eetate .O71 ARF qaoe 678 .213 sahc Pit total .... 249 (1.00 to 1.10 meters deep) shell began to disappear, and the midden color grew fainter. The bottom level (1.10 to 1.20 meters deep) ex- posed reddish sterile sand below gray-white sand. Only one profile is shown from pit III, the southwest face. This cross section is typical of the pit (see fig. 8). Pit IV (1.50 x 1.50 meters) was located at the opposite end of the site from pit III, near the water’s edge. The top level (0-.10 meter deep) was in sandy midden and broken shell. From level 2 to level 6, inclusive (.10-.70 meter deep), the midden was black and mixed with whole shells. In level 8 (.70-.80 meter deep) reddish WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—WILLEY 67 subsoil showed up very abruptly. The pit was carried down another .I0 meter, but the refuse was obviously no deeper. In each level, with the exception of 8, approximately 100 sherds or more were recovered. Stratigraphy—Of the four pits excavated at Sowell only pit III shows any appreciable change in pottery type frequencies from top to bottom (see table 7 and fig. 9). The type Wakulla Check Stamped displays a gradually expanding upward trend in the midden. DEPTH IN METERS -5O 400 450 44 HEAVY HUMMUS, FINE SHELL, SAND L/GHT BROWN SANO ANDO FAINT ANIDDEN STAIN -2£O 40 -6O LIGHT BROWN SAND AND DARK AIDDEN STAIN -GO SAND, 8LACH ANMODEN AND SHELL 400 WHITE SAND AND MIDDEN STAIN CLEAN WHITE SAND | RED BROWN SAND SOUTHWEST: FACE Fic. 8.—Pit III soil profiles, Sowell (By-3). 420 Weeden Island Incised and Weeden Island Plain occur with slightly greater frequency in the middle levels. The four sherds of Swift Creek Complicated Stamped that are present come from the three bottom levels. Fort Walton Incised and Lake Jackson Plain are both found only in the upper levels. In pit IV there is also evidence for Fort Walton Incised, Lake Jackson Plain, and Pensacola Plain being late (see table 7). There is, however, in this pit little or no substantiation for the increase of Wakulla Check Stamped in the upper levels or of the presence of Swift Creek in the lower levels. VOL. 113 SMITHSONIAN MISCELLANEOUS COLLECTIONS 68 OFISVIN/ oo NOLTIOM LA DISAIVELS HIFTHD VITNMHYM SCALE /OOT7e OFSIDN/ ONVIS/ NIGFIIM CFA OLS AWOD WIFAD LIIMS LEVELS Fic. 9.—Graph on pottery stratigraphy, pit III, Sowell (By-3). ARCHEOLOGY OF FLORIDA GULF COAST—-WILLEY 69 WHOLE VOL. Sig ‘"***** [e303 1g ¢ seee bad eee wee 600° eeee nae Or eee eeee [e304 Hd sees ¢ko- I CL weer or zr) o peyissepuy 809° v1 6oF* L 16S" ure[g [ENpIsoy ureg woourg Pow BEd peyteut-p10y epritoyy aM (&-€g) 62ZI* eee eee eee eee seer ener eee tee £ seee eee seer seer sae eee seee . . wees sees eee seer tees JA O gSo° gSo° eee see weer eee seer weer Zz I P eee ge eter seer eee wee seer weee I eae geee eee eee eee seee eee 9Zo° gfo- seer LT seee eee seer eee eee v z _ = v 2 fo) ep) ae ee H » -_ Ay urelg pestouy Pues], Uapad\y pedureys pereorpdur0oy pedureys 3PPqD EAPM pesrouy singsi9}eg 4S peyqourd -3Splry 1990N LT, pestouy sseg uvipuy perejoung aTeqetieg pestouy aTjeqetieg purysy wapsa Pet] WMS (Sole UT UaATS JoAsT Aq 9oUdIINDD0 a8vIUIII0q ) 1]2m0S5 ‘AT pun TIT snd fo uoyvoiyfrssnj9 pasys—ZL aay L, pedure3s yoaqO wosT isiais eses uleld pjoorsuag Uleld uosyoef oye] pestouy uoyeM 04 wa mow, WAT Sie) AL NG 7tS7 208) S[aaoT 7O SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 113 Pits I and II are both shallow, and all types seem to be fairly well mixed through nearly all levels (see table 6). In pit II, Wakulla Check Stamped does show an increase at the top, and the only Swift Creek sherds come from the two lowest levels. This, however, is offset by the presence of Lake Jackson Plain and Pensacola Plain sherds in the bottom of the pit. The case for ceramic stratigraphy at Sowell, based essentially upon pit III, is not as conclusive as at Mound Field or Carrabelle. There are, nevertheless, indications of a number of the same sequential changes in pottery types which were found in the two previously excavated sites. Sowell was the first site excavated at which the type Wakulla Check Stamped obtained in dominant percentages. Its high frequencies of occurrence in the upper levels of pit III accord with its small occurrence in the top three levels of pit II of the Mound Field site. The small amounts of Fort Walton Period pottery coming from the upper levels of pit III at Sowell are in line with the late position in which sherds of this complex were found at other sites, and their presence argues that the area of the site sampled by pit III has not recently been disturbed to any pro- found depth. The physical strata revealed in pit III are not analogous to the sequences of soil types seen in the Carrabelle and Mound Field pits. The band of midden and compacted shell at .50 to .go meter in pit III at Sowell was not associated with an intensive Swift Creek occupation, and the overlying refuse, rather than being rich brown or black midden, was sand only slightly discolored by organic debris. The pit III deposition can best be interpreted by consider- ing it as a continuous cultural accumulation in spite of the lighter sandy zone superimposed on the deep shell and midden layer. This sandy stratum is not clean sand, and it does not appear to have been water-carried or wind-blown. The pottery sequence at Sowell overlaps with the later stages of the sequences at Carrabelle and Mound Field. Whereas both Carra- belle and Mound Field stratigraphies close with a ceramic period characterized by a mixture of Weeden Island and Swift Creek types, the Sowell pit III sequence shows Swift Creek Complicated Stamped only in the lower levels. In the upper levels Weeden Island is associated with a new stamped type, Wakulla Check Stamped. Fort Walton Period sherds occur superficially. WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—-WILLEY 71 Classification of surface sherds from Sowell Fort Walton Complex ASEM VV ILO EE IE ISEM cere Grice elo ecalcis sce rowenta cana t.ca-soe ¢ vis 4 Weeden Island Complex Weeden Island Series: Weeden [slandli@lncisedi:.2%cio0) tote so. wiciele steletolepeiate,o'elelsrelers.e 3 Weeden, sland Plaine hacer «rrals «cis, Stckelatatal aig ttaratel a fovelatuls Al Ger ae ENEISEGH sre ye,sisclelosslecainaie wayeiniee wisp oeas, « 6 GarrabellenPunctateds suc. tes croc vecle sloersiare novels spares 6 IGICHWUNCISEM He Sete lles corre n alee S Oe crore oeavertte sets ave erates 2 Antiekent dRidse=piniched Wl tsee-acry sites asic clelcisrlete etsvelersrslo-cta ever 2 WV lam CheckesS tamipedmeria ce cisercretetster crisis et levelerelereieiore’s, sore)" 117 Swit ‘Creek Complicated) Stampedins. 65. teeter de oisis levee 9 Deptford Complex Westiorca Bold Check Stamped)... a +) eiisieccpseibrelecie eis eiere si6e 2 Miscellaneous Sino JME ayaa custo Gad Gogo mer aptoc tC ocd oc Ome Como man 24 RM vat ies RCN ge reverse fa, otis alers otreish ew areroslSatea wieithe aebesmeiias ied afetaie ale 6 RES ACT Mea iti pe t= ci aiayeveroveionys lave ss vara: fetes o atelaceyereveteiehe ole sla csverels e 270 Indeterminate stamped), cia. n Btd Bt td S g 2 oO eo) a. Ef B Ores ey & 5 aa Br is) o™ 38 i} ee o a. ur s (Sole UL UdAIS JOA, Aq 9oUaIINII0 98eIUIII0g ) (9-40) “ol Og “TIT pun IT sng fo uoynrfissoj9 pasys—O6 @IAaV L pastouy purest uaps2\\. Uield pues[ uspse\\ pestouy uo}eM 14107 S[2Ao'T 84 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I13 a sequence is consistent within itself and checks fairly well with the stratigraphic data from the other excavated sites. The align- ments of the levels of the three pits follow the frequency patterns of the various marker types. The occurrences of Weeden Island Incised and Carrabelle Punctated, Wakulla Check Stamped, and the Santa Rosa-Swift Creek types were matched across from pit VI to pit VII, resulting in level 1 of pit VI being correlated with level 3 of pit VII. In a similar manner the occurrences in pit IV were compared with those of pit V1; and level 1 of pit 1V was corre- lated with level 4 of pit VI and level 6 of pit VII. This gave a total of 19 levels with considerable overlap among the pits. Percentage of types for each level were then computed on the basis of the “greater” levels where there were instances of overlap. For instance, “greater” level 6, composed of level 1, pit IV, level 4, pit VI, and level 6, pit VII, combines the sherds of all these levels, totaling 371. The Weeden Island decorated types for this combined level are then totaled to 19, an occurrence of .048 percent of the combined or “greater” level. Gross and simplified as this system is, the various types and type complexes fall into easily recognizable trends of vertical occur- rence.’ Two periods, first the Santa Rosa-Swift Creek, and later the Weeden Island, are substantially documented. The Fort Walton Period is not well represented, but enough sherds were found to prove occupancy of that time, and these sherds are similar or identical to the vessels which Moore found with the burials in the big pyram- idal mound. The Deptford Period is only suggested in the bottom levels. Deptford Linear Check Stamped, the marker type for the period, was not found in significant quantity at Fort Walton. The Carrabelle stratification would indicate that Deptford Bold Check Stamped and Deptford Simple Stamped have a later occurrence than Deptford Linear Check Stamped, and, also, that they are found in small numbers in dominantly Swift Creek Period levels. Such would seem to be the case at Fort Walton where only the “tag-end” manifestations of the Deptford Period were observed. 12 The greatest weakness, in this case, is not the inspectional and subjective matching of the levels from one pit to another, but the relatively small numbers of decorated sherds, and total sherds, in most of the lower levels of the pits. Thus, the high occurrences, expressed on the graph in figure 13, of Swift Creek Complicated Stamped and Basin Bayou Incised, in level 19, are based upon one sherd each. ARCHEOLOGY OF FLORIDA GULF COAST—-WILLEY 85 WHOLE VOL. reeeees 1e]03 Wg peyissepoug [a etPrso a ure ure[g yyoows pedwrys aduiig psojyjdoq eee eee. I I I seee sees see eons eee were eee III’ see eres eee soe eeee eee III* eons eeee sae I see aoe eee sane sees wees I ones see wee sneer OOr* seee oof Oor* eee see eeee eee eeee weer weer I eee £ I weer eee eeee eeee ones eri’ see eee eeee eee weaee weer eee eeee eee ener I eee sees sees weer see sae eee eee eeee eee 060° weer weer seer 060° eaee weer seer eeee eeee seer I sane sees I eres see eee wees eens eee III’ sees III® see. sees seee lr’ eee JII* eee I ae I eeee eens tees I ones I eee 1S0° eeee weee eeee Ifo" eee eee Czo° ener eee z eee . . seee z eee sees I eeee seer 60° seee wees 2zo° Cro* weer seer eeee eae seer on wees Zz z wae eee sees eee seer cece tZo: eeee cose tZo* Z£o° eee cece z eoee eee z I sees Czo* Czo' Szo° sees eee eens Szo* eee I I I I I I seer ecee eee I 90° cece coe Bono Teo" Sto: eeee sees gor’ eee sees . $ Zz eee . I eeee eee eeee eee eeee ¢ seer sane eeee seee wees sees eeee eens eeee eeee seee gto" . . . . . . . . . . . . . . pestouy noseg uiseg pedureys pedurys nofkeg 10;esl|V padurys esoy vyueS peduir3s payeordur0od IDA peyxooig pedureys yooyD BINAEM peyeordu0d peduirys PIOd pr0jadoq poxreul-p.10d eploly SoM 1d UE smoipuy “3S peyeorduros) 3P2tD IMS peyepung aqpeqesseD pedureys 4x99qD ure (soreyt Ul uaATs Jado, Aq 9oUIaIINIO a8vjwUs010 ) (9-¥Q) “olny 140g ‘AT 14 fo uonoryissD]d pAayS— Ol AIav |, pesrouy ayjeqerie) pues] uepse\\ pedmrjs ; pexeyD wosT : ulrg ~ ejoorsuag Senet SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I13 Classification of surface sherds from Fort Walton Fort Walton Complex Bortavwalton. Incised&. .ccccieemec eccrine arena I Pensacola Series: Pensacola Inciséd’\.< scitiaiccine cron ois aietcieceie ei ere oe meee I Pensacola JRed) acs cieckar eien e e en ee chee ie ce rae ut Pensacola Blain: wa cecorce oe eee Oe O roe ioeeneectiaoer 10 Weeden Island Complex Weeden Island Series: Weeden Island’ Incised.cticncs. ce cen ot ecoee eee 5 Weeden: Island. Plain... ji. Gyo nn18e ware. claievaltohales cfotet spermine 13 Garrabelle=Incised: /..cstiieciecs «cco s ee eee ee eee 3 Carrabelle Punctated! 223 c.o- is cots oc ten 1s ce eit eee ao Keith tr LnGised sities e-cvdtorac tea acscra tia orerevetoreiaeie stetete trae rene teneneas I Tucker Ridse=pincheds seccs os soles coe sateen tree 2 Wakulla Gheck *Stammped!aac sel. serie cic se cists otoielet teil ete 19 Santa Rosa-Swift Creek Complex Swaitt) Greek Complicated Stamped. > c1-)s- em ceecieeetel eteeiee 13 Basin Bayou: [ncised ss. sicscre.s esc’. aes ae sine dae I Deptford Complex Deptford Series: Deptiord Bold, Check Stamped: ©. <5 3...0%.-0 << te eee eee Deptiord ‘Simple’ Stamped ois,0 ays.4:.c.0 o/.¢.0 5.0 ares eee I Miscellaneous Plaine Plated Habricelimpressedsjscmieeicrieeiseier ieee I Plata Red” sks. ice /zicter ses, Sores elo s aioe eis iasevs sore etal ieretote espe I Smooth Plain /s.542 seaie on haere a's Seiaieais ecaateoaiee een I Residual: ain G cee eteic eietcreve) severe) oisictevevelevorslstorsaicienstelenentectors Bo / Winelassified:.” 5 f4cu.s ccc scuiciareforsians oloyststertivie-n ete louise. een eae ee 3 Total sherds.... 264 TABLE 11.—Sherd classification of pit V, Fort Walton (Ok-6) (Percentage occurrence by level given in italics) iy ae) ‘ “ a Ss =] (2) -_ 5 E S$ £ $8 3 gf 3 Ss a ov 4g) os) ww eke = Ay Bel oe 4 ee Pe ee eee a4 = 3 xe) = 29 Se Be BS) ho =o 3s && a2) 5 3 Total eo 24 Ga £2 22 G8 c8 92 9. 2° eee ue ¢ — on uw & ow O-= (=) a wn iS) 3) o ar = ay & o hi Leeds Go) ae +S! OF ae et eo veo Ay 29) fA level I af sane Ae cone 2 Te lee 50 S000 GNC 37 40 a8 Soo) celta) 02. -925 2 rr sc I I Bone 2 I 2 50 61 -OlOe eto -016 ~=.065 g 033 -016 -033) eaean ARac 3 30 I Aaa6 2 aisle I eee I I 23 29 -034 069 ae -034 5an0 -034 -034 + =.793 aries 4 sete alc ae pices aisiaie eienate I 8 9 Boe : -II0~—-.890 Pit totale 139 ARCHEOLOGY OF FLORIDA GULF COAST—WILLEY 87 WHOLE VOL. teoe* 1BI03 41g oSz° oS: sees sees see fee sees sees see seer sae sees weae eeee see eee see see sees oS2z" I A eee eee eee sees sees eee see tee tee sae sees eae sees sees sees see sone I tees sae ZI see cos: eee see seer 1a te eee eee eee wee seee teee III’ eeee eee. eee seee TIT eee TEI" seee wee tees ¢ sees wees eee nN see sense see see eee sees I seers sees sane tae I toes I see see. Il Cie 982" 120° steer saee TZo* 1Z0° wane eee eee weer eee Cre eons steer sneer eee see eree 1Zo° see eee ¢ a I sane see I I see eee seas sees sees ¢ sees eee eee eee eee eee z eee tae OL S9e° evr: eee eeee err’ err: eeee eee see see seee eee err: eeee eee saee sneer err’ oer eee weee eee z eaee tee se ee eeee eee seee sae I eee eeee see eeee I eeee eee eee see 6 Lto* 999° sees eee seee eeee sees wee t60° eeee Lko° seaee tee Lto* Lto* eee tee shee weer Lto’ I . tee . tee eee sees sae sees wees I . see sees I ebZ: If’ eee see tees sees oes ZSo° go" seas eae sae eee see eae 820" see sees sees eee sees wae ¢ tee eee wee eee seee FA I seee sees tee eeee eee eee I teee eeee nee see eee z gro" CrZ: £fo° stee eee eee sees seer eee gs ££o° 910° 910° 910° eee seer fo" wee weer 910° seee eeee eee wee wee sea sees sae I eee eee 669" ge eee tee goo" see sees wees Zro° goo* see ZIo° ane sees see fer: eee Iter’ obo’ eel’ ofo° Coo* tee tee tee Co0* eene wee see eee foo° III’ eee seer 0Z0"* wees ozo’ ras mS BR =e N 3 Lame) N ° WAY N Bm VASA = Se 2 So * 5 ~» S n Xx S w ~™ ° . . . . . . . - . . . . . O05 . . . . . . . . . . . . . . . I s]aAoy urelg % pestouy VR: ejooesuag PpeyIssejouy ureld qooms : peg eld pedurrys ptoyjdagq pexrteul-p109 EPHol 39M. pedmr}s noseg 10}es[ Vy pestouy Geese peduejs esoy ejueS pedurrjs % 3P94D FLAS peqourd -2Sply ieayony, pelejoung eT[eqesied aT[eqeireg pestouy % PURIST Uapss AA pesiouy UOYe AA 107 urelg [enpisoy nodée aidunig pedmerjg yosya PIO ptojadog Orel Gyyaery peduieys pazeorjd -W0D IBA payoo1g pedureig pajzeord “WOD Y921Q IAS PURIST UspedA, (Soret Ul UaATS JaAa] Aq 9dUEIINIDO a8eyUII0I0g ) (9-¥O) uoND AA 4404 ‘TA nd fo uoyvoyfissnj9 pssyS—zI aTdv L, VOL. (tng SMITHSONIAN MISCELLANEOUS COLLECTIONS 116 soreees 18903 BI ofr" peyissepou ys) 809° v1 ure[g [ENP!sey ule] gq qOoUrS ee pauryg urelg poessaiduy o1iqe sees I tee I seee wae I nae eeee eens see saee sane weer ofr* seee sees weer Sto- weer eee eeee eeee seee weee seer £ ener eeee weer I weee eee eeee eeee seee eens sees eee Zo" co: eeee cco: eee eeee eeee Zzo° ewes eee sees wees I z sees z eee eee wee sees eee seee seee steer eee weer I I £ eeee EA wees wane eeee eeee seee eeee eeee gro g00* eeee eeee eZlo* 910° eeee eeee eeee eee seee weer tA I eeee seer 6 fA wo + a) i=) co i So 8 i) I i=) ONS « ir So oO. 4 as So ‘oO ™) < Cy ~ i=) pesiouy & pesiout sseg ueIpuy peyseul-19 Ny PIPPI PUROTL pexteul-p107d EPEC THeAS SN uel upper pedwurys nofeg 10}yesl[y nokeg ulseq pedureis esoy bULS “WOOD 4921] IAS peduirys OUD EMAe MA Wy payejoung & aeqeiied pestouy aJeqeriey pedureys pazeord (Solje}yI Ul UdATS JOAd]T AQ 9dUeIINIDO 99e}U90I9g ) (9-Y¥O) uoynYy 1407 TIA 34 fo uoynoyissnj9 paayS—€I FIV], pesiouy purysy uspaa MA purysy, wapec AA pedurys yosyQ uos'T urq % elooesuag Il oI s[aaoyT WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—WILLEY 89 GULF BREEZE, SANTA ROSA COUNTY (sA-8) Description of the site—On the south or Santa Rosa Sound side of the peninsula which separates Pensacola Bay from the sound there are a number of midden accumulations. Four of these middens, visited by the survey in the summer of 1940, are grouped together in the space of about 2.5 to 5 miles from the western end of the pen- insula. One of these sites, about 4 miles from the end of the penin- sula, has been designated as the Gulf Breeze site (map 2) after a little summer resort not far from the Pensacola Bay bridge. It is on the property held permanently as a United States Naval Reserva- tion. In 1940 a temporary CCC side-camp, from Foley, Ala., was established near the spot. Actually, along the shore of the sound the midden sites are so numerous that they virtually consist of a continuous occupation. Describing this country in 1883, S. T. Walker writes (1885, p. 859) : In proceeding east the first shell heaps are met with at Dr. Rotherford’s place, about 2 miles east of the old Government Live Oak Plantation. Im- mense beds of shell and the usual indications mark this as the former resi- dence of a large population. The slopes of the hills are covered with irregular beds of shell from 2 to 6 feet in thickness, which occupy an area of several acres. As near as can be determined this is in the general vicinity of the Gulf Breeze site. This site is on the low bank back of the beach, some 2 meters above sea level at high tide. Outcroppings of shell and black midden, about .50 meter deep, can be observed along the top of this bank which is undergoing erosion. Like most of the land along the sound it is thickly grown with scrub trees and underbrush, and the best possible estimate on the extent of the immediate deposit would be about 40 meters in diameter. Although the surface was overgrown, it was possible to gather a sherd collection from along the bank. Excavations —Six test pits were put down at the Gulf Breeze site within a radius of less than 10 meters (map 7). Pit I (3 x 3 meters) was representative of the others in soil content. The upper soil zone was composed of brown sand, humic stain, and a little shell and midden. Below this top layer was the main midden and shell zone. This was, in turn, underlain by a gray, midden-streaked transitional zone; then by gray-white natural sand; and, finally, by reddish sand. The tangled mat of many roots (living and dead) of trees and bushes was noted down to a depth of at least .50 meter or part way through the concentrated shell and midden layer. 8 gO SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I13 Sherds were plentiful in all levels of pit I down to .70 meter below the surface. In the eighth level, it was necessary to take the pit down to 1.10 meters to recover sherds from a pocket of midden which extended downward from the west corner. This pocket, upon ESTIMATED LIMIT OF MIDDEN TEMPORARY SET-UP —— > N oO 0 SANTA ROSA ue SCALE IN METERS SOUND Map 7.—The Gulf Breeze site, Santa Rosa County (Sa-8). further excavation, proved to be an old refuse pit containing black soil, shells, and several sherds. All the sherds taken from it were cataloged with the material of level 8 (.70-1.10 meters), although the actual limits of the refuse pocket extended to 1.80 meters below the surface. There were not many sherds in pit II (1.50 x 1.50 meters). Seven levels were removed, all .10 meter in thickness except the last which extended from .60 to .go meter. Pit III (1.50 x 1.50 meters) yielded WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—WILLEY OI more sherds than pit II and was excavated in 8 levels, down to .8o meter below the surface. Pit IV (2 x 2 meters) was excavated to .70 meter below the surface for a total of seven levels. Sherds were fairly numerous, averaging about 50 for each of the upper levels. Pit V (2 x 2 meters) was carried down to .go meter in eight levels, the last level being .20 meter in thickness. There were almost as many sherds in this pit as in pit I, an average of well over 50 per level with the exception of the lower and semisterile levels. Pit VI (2 x 2 meters square) went to a depth of .85 meter below the surface, the last and sixth level going from .50 to .85 meter. The sherd yield was about the same as that of pit IV. Stratigraphy—An examination of the tables for the various ex- cavations at Gulf Breeze (see tables 14, 15, 16) does not reveal any significant vertical changes. The refuse was thin, although no thinner than that tested at Carrabelle, where definite stratigraphic results were forthcoming. Small amounts of shell-tempered (Pensacola Series) and other Fort Walton Period types are seen in the topmost levels of pits II, III, IV, and V. Occasional shell-tempered sherds were also found in the lower levels in pits I and VI. Pottery of the Weeden Island Complex is conspicuous by its absence. Only one or two pos- sible Weeden Island sherds were recovered out of all six pits. No Deptford Linear Check Stamped was present and only a half dozen Deptford Bold Check and Deptford Simple Stamped sherds were scattered through various levels of the different pits. Swift Creek Complicated Stamped ran almost from top to bottom of all pits. In association were the incised and rocker-stamped types of the Santa Rosa Series, including Santa Rosa Stamped, Santa Rosa Punctated, Basin Bayou Incised, and Alligator Bayou Stamped. In pits IJ, III, V, and VI there is a suggestion that these last-mentioned four types may have a slightly earlier range than the Swift Creek Complicated Stamped ; but on present evidence this is so vague and inconclusive that it can hardly be considered. The site would appear to have been occupied at a relatively early date, during the Santa Rosa Swift Creek Period. The several shell- tempered sherds, the majority of which were found in the upper levels, imply some Fort Walton Period activity at the site after an in- terval of no occupation. VOL.. 113 SMITHSONIAN MISCELLANEOUS COLLECTIONS 92 Sor eee ewee ofS eee ewee 1Zo° ofr’ 11) f ger £ gro" I OrI0* peyrssepou 1830} Hd 999° z eho: 6 brZL: [e304 UT 69° os urerg [eNpIsoy ureTg yours waee sees eee peduirys ajduig p10jz3deq pedurezg yoaqa Plog pr0j3doq eee eeee eri’ eeee tee sees I aaee see eee wees ones snes tees sees aeee eee aeae wear . aoe sees sees sees sees wae eee eee sane sen eeee azaaIg jny gfo° gfo° sees zo: I £ I sees eens eee eeee see ZIo° eens wee eee I eeee €10° os 420° €10° I see z I 9Z99I1g JN TJ ured UTPyUe IY pedurys noseg 10}esy[y pesiouy noseg uiseg pedureys esoy BIULS eee pedureys pozeord -W02d SMIIPUY 3S pedure}g pezeoryd “WOD Y924D INAS pestouy aqfeqerieg pestouy purysy, waparA, ureld purjsy wapsa (SOLe} UL UDAIS JaAaT Aq doUdIINIIO 38vIUIII0q ) (§-DS) azaaag f{jn5 ‘TIT pun J syd fo uoynsyissvj9 pasyS—vi eee eee wane pestouy toqivyy Ayayes alav |, LAGE ejooesuag wees wee pestouy eloorsuag sees eeee ule uosyoef aye] eee eee eeee eee seee eee pastouy Uo}eM 1107 a Tm So SNS cle Tarn tee Meo) S[9AIT ARCHEOLOGY OF FLORIDA GULF COAST—-WILLEY 93 WHOLE VOL. voz tee e twee peyissepouy, 1230} Wd E30} Hd oor’ £ S9e° f o1b: eee sees ssee eeee see eee stee tee eee 0o'r eeee eee saee fees taee see. eeee seee tee I eeee eeee wees Cor’ oeee eeee 250° f9e" weer wees eee eeee seer tA eeee eeee I S seer sees tfo° eee eee ZIo°* ZIo° eee weeee OgrI° eee O9I* z eee tee 1 I eee eee Z eee II §10* eee eee 9So° weer eee gro* eee LZEo° ¢Zo° I eee eee {o wees eee I seer z v sees eee eee §10° §10° eee gro" eeee eee 060° eee eee eeee I I eens I eee eee ¢ eee eeee eee eens seer 1Zo* eeee eeee eee eee eeee wees see sees eee I tees eae. eeee eee 9z901g FIND “AI Ud seae eeee eeee weer eeee err: eeee geb- eeee eeee eeee wees eeee eens wees I wee ¢ eeee seer 960° wee seer eene ££o° 62rI° eeee Sze’ tees #90" £ eeee eens eoee I ¢ eens ‘L seer z az9e1g FIND ‘III Hd ule Y}OOWS pedure3s IS proydeg pexreu-pi0o*d BPlopy SOMA uel urpuery pezejoung esoy eyUeS pesiouy nofteg uiseg pedurrys PBsoy eyUeS pedurr3s pereorjdur0og SM3IPUY 3S pedurezs duioy 421 ILAS payeory a[dur (Sole} UT UdATS JaAa] Aq ddUaIAINIDO 98eIUIII0g ) (§-DS) azaa4g finy ‘AT pun yyy snd fo uoypoifissvja pasyS—sS1 a1aV J, Brel. ejooresuag LUG uosyoef 2xeT mo & mo + Mm wo s[aAo'T zge settee? 72903 HT VOL. 113 SMITHSONIAN MISCELLANEOUS COLLECTIONS 94 see sees 00°I ones sae sees sees sees sees sees sees sees sees eons eece env. seee seee €£9° ener seer 1Zo° eeee £zo° weee £zo° £zo° £zo° weer weee eee eeee aad eee ce I Tr eee seee gto: [So° §10° seee tee ZEo° weer §10° seer Zfo° wseee 2 gro° §ro" $s eee ce £ I see weee z eee I eee A seee 8 I I steer seee 4e9° 920° eeee eeee eee 610° seer 6£0° 6f0° 94o° eeee VAG te eter weee 1s eee ze v eee eens seee I wee cz UA P eeee 9 wees seer eee £ro° CLZ° fro0° seer seer eee £10" £10" fro° Zzo* eeee seer 260° £ro° eeee I gs 1 eee weer seer I I I Z weee seer Z I seee z seer £90" IZZ° 820° wee eee wee 20° eee seer eee ener eee $90" eeee eee g ote I € eeee eaee eee eee 0) ORF In) S86. ezaeig FIND ‘TA Ud ggh scree? 78303 Hg eeee weer were oS: weer wees eeee eens ooS* esee eeee eeee eeee weee eene eaee z eee eens I sees eeee sees sees I sees eons sees aaee tees sees eeee weer weer 003° Oor* eee eeee eeee oSo°* weee sees eeee eee eeee oSo* eeee eeee oz snes gi Zz sees eeee eens I sees sees sees sees eens I seee 2fo° 903° eso° seer eee ease seee eeee eeee eeee 2fo° 2fo° ¢90° eeee eeee Ie I Sz I eeee wees eoee sees eens sees eeee I I A weer PIO" &SZ: 9So° g§zo° eeee eoee zbo* PIO* eeee Fio° 9z0" eeee ebo* eeee weee ba z sees sees ¢£ I eeee I Zz eeee € sees wees sees or0° 929° 9zo" seee sees OI0* oro" oOr0° 10° eto: o10° eens €Zo: aeee seer $6 I t9 ZI sone sone I I I € ¢ I sree Z sees sees eZ: 160° sees sees seee eeee seer sees 9z0" 9o° sees 90° £10" sees EL eeee zs Z weer eeee eeee eee eeee eeee z 9 eaee ¢ sees sees 098° zo’ sees eeee eeee g00* eens 910° g00° wees eeee 9So° goo" eeee Cz1 eeee OII £ eeee eeee eoee I eeee z I eeee eoee ZL sees £10° zg" seee eeee eeee eee eeee £10" sees £10" eeee eens oto: £10" £10° £Z eee 09 eeee eeee weer eeee seee I eens I cone eeee £ I I I azaeig FIND ‘A Wd O.0 CRP im) NO! AN) 160 S[aAo'T Uleld Tenpisoy uleld pesrouy LAGI ejoorsuag uosyoef aye] Ureld UIpyuesT nofkeg uiseg rf) iS B POGISSe[IT/)) Poa rela peduareys ajduig psoydeq pexreul-pi0d plop SoM pazejoung esoy eyUeS pedureys nokeg 10}es1[V pedureys esoy eyULS pedureys peyeordmoy SMIIPUY 3S pedure3s pereorjduros) aoe so TS (SOI[e}I UL USAID JaAV] Aq 99U91INIDO 99e}U9919 4 ) (§-DS) azaaag {[n5 ‘TA pun y syd fo uoypoiyisspj9 paayS— OI ATAV WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—WILLEY 95 Classification of surface sherds from Gulf Breeze Fort Walton Complex GOREN alt One lnCISed racic came cm cote celal sieisloeislels elas whe eMers I Pensacola Series: Pensacola MCSE jade sro ts sit aisie Seals arate oeperaraieletae acaete Rs. 5 Bensacoraeebel aime erp te cieloves leis storie cieisreecrotteleNate siers shar stoic 36 Weeden Island Complex pucker Rid se-pitichtedhy ype cir see eucntolelorasterehe) steleleieters efers el Siete\e I Santa Rosa-Swift Creek Complex Santa Rosa Series: Allieator Bayou. Stampedis ss - 5 cts coe tains avclaye aie. oa se orice I SanitamRosay Stamped crc cette cera solcietelsiate\eieleials 4 Complicated Stamped Series: Switt. Creeks Complicateds Stampede secs cele icles ss 1elerele 10 Se Andrews Complicated’ ‘Stampeds sss foc cece ccc cee I Wiestebloridas Gord-matkedencs sa ccie ceca cte cei diedetieicis s crelers I (Early Variety.) Lied IBEW odocuonjomsumo ase uucobn SH beeaUoon bgouddodade 2 Miscellaneous MEM estclttall ppb) elite vayevers, ters fecaus)clicisverars sia) reve cones aaa te fellate eveteha Voie! « /orelet 56 NOtrclassitiedamperais watevee cise evacciere arouse iciecavonslolteycnecthcrsy mrexehctetieveforolsnals 9 Total sherds.... 128 LAKE JACKSON, LEON COUNTY (LE-1) Description of the site—Four and one-half miles north of Talla- hassee, Fla., in Leon County, there is an impressive mound group (map 2) of six, or possibly seven, mounds.1* The topography of the immediate setting is a natural lake shore or valley backed immediately to the west of the site by sloping but high bluffs (pl. 3, top). The lake valley stretches northward for about 1o miles and is surrounded by similar bluff and rolling hill formations. At present the lake is little more than a shallow marsh, and, according to old inhabitants of the region, it has been receding for a number of years. Oak and gum trees form a screen along the lower slopes of the bluff around the lake, and cover, or partially cover, some of the mounds. A large part of the area immediately surrounding the mounds has at one time been under cultivation. Physiographically, the region is a part of the “Southern Pine Hills,” a gently rolling red clay formation which begins some 20 to 30 miles inland from the coast. Jackson is one of several lakes in the 13 Previous to 1940,.the only direct reference to the site in the literature, to the best of the author’s knowledge, was Boyd, 1939, p. 272. 96 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. II3 Tallahassee region formed by the solution of underlying limestones. In general, the Tallahassee region has a greater physiographic re- semblance to the red clay hill country of parts of south Georgia than it does to the sandy coastal strip of Florida. The superiority of the soil in this hill country, as compared with the sandy soil of the coast, is attested by modern agriculture which is flourishing around Talla- hassee. This present-day condition undoubtedly reflects the aboriginal situation, and, in part, explains the presence of such a large mound center in this region when there are none on the immediate coast. The mound group is located on the southwest shore of the lake on a small inlet known locally as ““McGinnis’ Arm.” All six or seven of the mounds can be enclosed in an area measuring approximately 450 meters north and south and 375 meters east and west. Six of the mounds are, or were, in all probability, truncated pyramids. Mounds 3 and 6 still preserve this shape, but the others no longer retain their original contours which have been blurred by erosion accelerated by promiscuous digging. The accompanying plan (map 8) exaggerates the shape of all the mounds, as measurements were taken from points estimated to be the corners of the bases and summit platforms. Mound 2 is much the largest of any of the six, having an estimated height of about 8 meters and base measurements of 65 x 48 meters. What may have been a ramp slopes down from the summit on the northeast side. Viewed from the top, the mound appears to be five- rather than four-sided. Together with mounds 3, 4, and 6, mound 2 forms a quadrangle which is roughly oriented to the cardinal points with mound 2 being on the north. Mound 4 is second in size and is on the west. The base of mound 4 measures 34 x 43 meters, but the height is only about 3 to 4 meters, Mounds 3 and 6, on the south and east, are each about 3 meters high. Mound 6 is particularly well formed, oblong and steep-sided (pl. 3, bottom). A little to the south of mounds 3 and 4 is mound 5, also approxi- mately 3 meters in height. The summit of this mound seems to be triangular; however, it has been badly scarred by excavations into the top and sides. To the north, a good distance apart from the other mounds, is mound 1, the smallest of the pyramidal mounds. It measures only 21 by 20 meters at the base and is approximately 2 meters high. The old excavations in the sides of the mounds reveal something of the nature of their construction. All are topped with red clay mantles, and in some, successive clay mantles with intervening beds of sand or midden fill are discernible. Near mounds 2 and 4 are large de- pressions suggesting borrow pits. There is little doubt that the six WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—WILLEY 97 mounds are of the platform or substructure type, and future excava- tions, if they are carried out, will probably reveal house patterns on the various clay mantles. 166 METERS To M07 7O0M. BASELINE OLD EXCAK ° 70 20 JO 40 $0 je SCALE IN METERS Map 8.—The Lake Jackson site, Leon County (Le-1). The conformation of the mounds is highly stylized. A few meters to the north of mound 6 is a small, low, circular rise which we have termed mound 7. Not more than a meter high, its artificiality can only be proved by excavation. Although it may be a burial tumulus, the possibility that this low hummock is a house mound is increased by the fact that a small stream flowing nearby has cut through a portion of a baked clay floor. 98 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 113 The area surrounding the mounds is, with few exceptions, covered with sherds. A surface collection was made at large over the site and from concentrations on and around mounds 1 and 2 and the area between these mounds. Excavations—Two 3- x 3-meter test pits were made at Lake Jack- son in the village area (map 8). Pit I was located about 20 meters south of the southwest corner of the mound. This area, between mounds 2 and 3 appeared, from surface inspection, to be covered with a heavy alluvial layer deposited by a small stream which once ran be- tween the two mounds and then turned northeastward to empty into the lake. It was thought that at such a spot the probable village refuse would be sealed in and protected by the alluvium. Level 1 of pit I was carried down to a depth of .30 meter below the surface. At this depth, sand wash mixed with recent dark brown humus was exposed, and a few eroded sherds were found. In level 2 (.30-.45 meter) the soil appeared the same, and no more sherds were encountered. A small test hole was put down in one corner of the pit at this depth to see if the midden zone lay below the sand and silt. Between .45 and .70 meter consolidated sandy clay and specks of charcoal were ob- served. Below this was a thinner zone of clean sand which was again underlain by clay and traces of charcoal. Under this second band of clay, clean light brown sand was followed down to 1.40 meters below the surface before the pit was abandoned. It seems likely that the midden, if it existed between mounds 2 and 3, had been scoured away by the action of the stream, or was buried at a considerable depth. Pit II, on the north side of mound 2, 60 meters or more from the mound, was situated in the midst of an intensive occupation area. The surface, which not long in the past had been plowed, was covered with sherds. In this pit the first level, .10 meter deep, contained 100 sherds. Working in .10-meter arbitrary levels, the first two or two- and-a-half levels were plow-disturbed, and the sherds found were very small fragments. Beginning in level 3 (.20-.30 meter), the soil became more compact, and the color of the midden deepened in in- tensity. Sherds were most numerous in level 3. Level 4 (.30-.40 meter) revealed undisturbed black midden. A few briquette frag- ments came out of this level, and there was a small patch of partly burned yellow clay on one side of the pit floor. In level 5 (.40-.50 meter) the midden was much lighter, and the sherd yield dropped to about one-quarter of either of the two previous levels. In level 6 (.50-.60 meter) only a few streaks and spots of dark midden stain WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—-WILLEY 99 showed up in the red-brown loam. It is possible that some of the spots may be post molds, but the excavated area was too small to determine an alignment. Level 7 (.60-.85 meter) was a small test in one quadrant of the pit. In this test, a yellowish calico-clay marl was revealed in an uneven stratum below the loam, but no sherds or other evidence of human occupation were found. TABLE 17.—Sherd classification of pit II, Lake Jackson (Le-r) u S 2) se) Sms cs wilt ee g ret [3] rS) ~ — rt So) MRSS ot alse aie on 3 Total 2 Oe 2a ag | oR a sherds ro os vO £ ow = as 1S) 58 A a5 goa ic on & by Levels Bo 4 Gi ey A 4 =) level I 7 3 2 5 aye 85 3 100 2 6 2 I 50 136 2 146 3 15 7 2 I I 18r 3 210 4 13 5 I 2 155 I 177 5 6 56 of 34 50 40 6 4 7 23 : 34 Pithitotall iiec.cca caste 707 Stratigraphy—The pit II excavations showed the same pottery types as the surface collection from the site. The rich but thin midden gave no indications of a percentage stratification of types (table 17). The site, on the basis of the tests and the other collections, belongs to the Fort Walton Period of Gulf Coast occupation, and it is undoubtedly one of the largest mound centers of the late period in northwest Florida. It is possible that earlier materials than those of the Fort Walton Period may be found at the site with intensive excavation. Classification of surface sherds from Lake Jackson Fort Walton Complex Fort Walton Series: HOrtenvWialtorme incised nts siere salersherls Stators Sale telawiesieeier onus 21 WsakemlacksOnb bu laiiieyvaces restore, olen me mere iancsteraeic eke reais 5 SALE LU Mla Om SENCISEdL 2 <5 shea facts can Sache te bia erarta ae alee as 2 SS PMR ctehs It oonMCISEd sonic sata rtisint este seit Ue serene ane Aue ics I Miscellaneous Brushed (possibly Walnut Roughened)...........0.0...00+6 I IRGHGell BEM Cae daoodd sab on Co ce Ob See Tee me cord aa Eroe 188 (These are undoubtedly body sherds of the Fort Walton Series. ) or Total sherds.... 218 100 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 113 CONCLUSIONS In briefly analyzing and characterizing the successive culture periods of northwest Florida and their ceramic identities we begin with the earliest (see fig. 14). This is the Deptford Period which is charac- terized by the appreciable occurrence of the type Deptford Linear Check Stamped. This type is the most distinctive of all the South- eastern check stamped types, and its appearance cannot easily be con- fused with other types. The evidence for the Deptford Period comes from the stratigraphic tests at Carrabelle where Deptford Linear Check Stamped is the dominant type in the lower levels of four of the tests made in the Carrabelle midden. It is associated with Deptford Bold Check Stamped, Deptford Simple Stamped, Gulf Check Stamped, and small amounts of Swift Creek Complicated Stamped. At Carrabelle, and in other excavations Deptford Bold Check Stamped, Deptford Simple Stamped, and Gulf Check Stamped are found in later contexts where Deptford Linear Checked Stamped is absent. The Santa Rosa-Swift Creek Period is distinguished by the ascend- ancy of Swift Creek Complicated Stamped (Early Variety) ; and the coeval appearances of St. Andrews Complicated Stamped (Early Va- riety), New River Complicated Stamped, and Crooked River Compli- cated Stamped (Early Variety). Deptford Bold Check Stamped, Dept- ford Simple Stamped, and Gulf Diamond Check Stamped occur only in small quantities. The type Franklin Plain enjoys its greatest popu- larity along with the rise of the complicated stamped pottery. Plain, sand-tempered body sherds, never a large part of the sherd counts from Deptford Period levels, are more common during the Swift Creek Period than previously. At Carrabelle, levels overlying the Deptford Period and underlying those showing Weeden Island deco- rated pottery best represent this period. The lowest levels at Mound Field also fall into the Santa Rosa-Swift Creek Period. Around St. An- drews Bay and to the west the make-up of the Santa Rosa-Swift Creek Period differs somewhat from the pottery complex seen in the middle levels at Carrabelle and the lower levels at Mound Field. In addition to the early complicated stamped types, Santa Rosa Stamped, Basin Bayou Incised, Alligator Bayou Stamped, and Santa Rosa Punctated appear. These types are also found to the east of St. Andrews Bay, but seem to be more common to burial mounds in which they are found as far east and south as Pinellas County. The Santa Rosa and Alligator Bayou Stamped types are the result of a technologically dif- ferent surface treatment, being produced by a rocker rather than a M3 ~ PITAGE FRIPERIODS are = WEED: in rksville pot- Z WALTON. yle of Ohio. ect this gen- x ad the bulk Santa Rosa- ‘ elle Incised, sed, Tucker q eginning of YEN ISLAND IZ ist numerous - limited sur- Swift Creek period but is = 1ew Weeden Swift Creek c 3] : i > are calling ‘ »r Carrabelle 8 en Island II 5 rc acterized by | typed which eeden Island nSDEN /SLAND I is definitely ugh there is ame ikulla in the :d by the oc- a ensacola In- ell-tempered, ¢ Sort Walton, of the Fort G& ~ ~ > i Pa = nig gi iS al lg ea glia Sl cl ly ya ul ina 4 G | ROSA-SWIFT CREEK DEPTFORD fest coast. Sti. els. Only at Fort Walton ibundance. e I pits AND LEVELS POTTERY TYPES BY PERCENTAGE FREQUENCIES PER LEVEL PERIODS DEPTFORD L/NEAR SWIFT CREEA SANTA ROSA WEL CHECK STAMPED COMPLICATED STAMPED DECORATED TYPES een WANT A CHECK FT Wal. ee 4 FT WALTON, == aT WEEDEN ISLAND IZ = C___} = J WD SS ' cp Z sacs et WEEDEN ISLAND I eae TT ITT) rs 5 enpissseete CIT TITTY) CTI TJ ky CTI ofroon man k Z bs | 4 BOTT ETO SSeS Sows sw 58) : < < PaPPee eee eee eat e, hi 5 AaB nor CT an FS) 00 (St Od ao 0 ee ad Cc S a ? SANTA ROSA-SWIFT CREEK 6 Cet ee eee feeatere a Os a K a 5 w fey PTT TTI ITIT LZ x ssuasenene - Crrrrsrryorsrsrs4rirri crirtTrtr4 g g S4 SCALE 25 % S QPS DIY LEE ROR ILE CD DON RATT LR PRT ETS EATER) Gara] . : A . oe ield, Sowell, and Fort Walton are interdigitated. . 14—Graph showing interrelationships of ceramic stratigraphy among various sites of the northwest coast. Stratigraphic pits at Carrabelle, Mound Fie a gitated WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—-WILLEY Io! flat stamp, and they bear obvious relationship to the Marksville pot- tery of the lower Mississippi Valley and the Hopewell style of Ohio. Basin Bayou Incised and Santa Rosa Punctated also reflect this gen- eral relationship. The lower levels at Fort Walton and the bulk of the midden at Gulf Breeze are representative of the Santa Rosa- Swift Creek Period in the west. The appearance of Weeden Island Incised, Carrabelle Incised, Carrabelle Punctated, Indian Pass Incised, Keith Incised, Tucker Ridge-pinched, and Weeden Island Plain defines the beginning of Weeden Island. Plain, sand-tempered body sherds are most numerous during the Weeden Island Period, in part the result of the limited sur- face decoration area of the Weeden Island style vessels. Swift Creek Complicated Stamped is still present in quantities in this period but is undergoing changes as the result of the impact of the new Weeden Island influence. This modified type is referred to as Swift Creek (Late Variety). An early Weeden Island, or what we are calling Weeden Island I, is seen at Mound Field and in the upper Carrabelle levels. At Sowell, most of the midden seems to be Weeden Island II in time. This second stage of Weeden Island is characterized by Wakulla Check Stamped, a small square check stamped typed which occurs in great quantities compared with the decorated Weeden Island types. Swift Creek Complicated Stamped (Late Variety) is definitely on the wane during the Weeden Island II Period, although there is evidence of some overlapping of Swift Creek and Wakulla in the middle and upper levels of the Fort Walton test pits. The beginnings of the Fort Walton Period are denoted by the oc- currences of Fort Walton Incised, Lake Jackson Plain, Pensacola In- cised, and Pensacola Plain. The last-named types, both shell-tempered, are common to the west of St. Andrews Bay. At Sowell, Fort Walton, and Gulf Breeze a small proportion of the sherds were of the Fort Walton types. These came exclusively from upper levels. Only at Lake Jackson, a site purely or predominantly of the Fort Walton Period, did we find Fort Walton refuse in any depth or abundance. IV. EXCAVATIONS ON THE WEST COAST: 1923-1936 INTRODUCTION This section treats of those excavations that were made under the direction of the Bureau of American Ethnology, Smithsonian Insti- tution, between the years 1923 and 1936 (map g). Of all the sites discussed only Weeden Island has been previously published upon in any detail (Fewkes, 1924). In this account some material, here- tofore not published, is presented from the Weeden Island site. Safety Harbor, which was excavated by Stirling, is presented here for the first time as a site report. The remaining sites, investigated under Stirling’s general direction, were a part of a joint Federal Relief Archeological Survey sponsored by the Bureau of American Eth- nology and the State of Florida. Some of these had been briefly sum- marized in a preliminary account (Stirling, 1935). The location of this group of west coast sites is at the boundary of two of the cultural regions of the Gulf Coast area, the central coast and the Manatee regions (see “Introduction,” pp. 2-3). The majority of the sites are within the Manatee region, and those in the central coast region lie just north of the line. For practical purposes they fit best, as a total group, in the Manatee region in that most of them show traits of the peripheral blending between the Gulf Coast and the Glades area to the south. (See Willey, 1948a). This discussion of site excavations from this part of the Gulf Coast is placed immediately following a similar presentation of basic data for the northwest coast region. These lower central coast and Manatee region excavations differ from those of the northwest in that they are, for the most part, investigations of burial mounds rather than refuse middens. Consequently, the problems of ceramic sequence are not as effectively resolved. In spite of this it is possible, by inferences drawn from the northwest coast, to arrange this group of mounds in chronologic order. The preparation of these west coast site reports is based upon the examination of collections and upon the field notes and verbal informa- tion of the various field supervisors. The Weeden Island analysis was written from the Fewkes report (Fewkes, 1924), from unre- ported collections, and from information given me by Dr. Stirling. Safety Harbor excavations are described from verbal information given by Dr. Stirling, from a brief published preliminary note (Stir- [ 103 ] 104 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 113 SAFETY 2 HARBOUR WEE DEN 4SLAND A THOMAS COCKROACH KEY --- =. =-_- _-. SCALE OF MILES Map 9—The Florida west coast between Tampa Bay and Charlotte Harbor, showing sites excavated in 1923-1936. WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—WILLEY I05 ling, 1931), and from the collections. The Thomas mound and Cock- roach Key field notes were prepared by Preston Holder, the excavator. An additional set of notes, on the Thomas mound, covering a short period of work in 1937, were submitted by J. Clarence Simpson. Dr. Marshall T. Newman gathered the data on Perico Island and the Englewood mound, the two sites under his immediate charge. The work and notes at the Parrish mounds were maintained by Lloyd Reichard. Collections from all these sites were studied in the United States National Museum. In most cases, however, these collections repre- sent only a portion of the material actually taken from the sites during excavations. This is specifically discussed in the report on each of the sites. Pottery and artifact types are listed or discussed in context only so far as they serve for period identification of a site or to differentiate certain features within a site. Nonceramic artifacts are, in some cases, described in more detail than the pottery. Both are more fully treated in the subsequent sections of this report which summarize and define the cultural complexes and periods of the Gulf Coast. Where data permit, a brief summation of the physical anthropology for each site is included along with the site excavation report. This information has been extracted from Hrdlicka’s “Catalog of Human Crania in the United States National Museum Collections” (1940). WEEDEN ISLAND, PINELLAS COUNTY (PI-I) Introductory review—The Weeden Island cluster of shell middens and sand mounds is one of several sites on the southwestern shore of Old Tampa Bay not far from the city of St. Petersburg (map 9). Geographically, Weeden Island is a peninsula rather than a true island and projects out into the bay on the north side of a small body of water known as Papy’s Bayou. S. T. Walker’s investigations (Walker, 1880a) in the neighborhood of Tampa Bay may have taken him to Weeden Island, although J. W. Fewkes (1924, pp. 1-3), who later excavated the burial mound at the site, is uncertain if the mound group described by Walker from Papy’s Bayou is the same site. In any event, the Weeden Island burial mound which Fewkes and party excavated in 1923-24 had not been previously touched by Walker, Moore, or any of the other earlier investigators. Dr. Fewkes published an account of his Weeden Island excavations in 1924. This was entitled “Preliminary Archeological Explorations at Weeden Island, Florida” (Fewkes, 1924). Although the Fewkes 9 106 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I1I3 account is obviously not a complete record, it is not possible to expand it greatly from the materials and data now extant at the United States National Museum ** and in the Bureau of American Ethnol- ogy. Virtually all of the whole or restored vessels now in the National Museum are those also illustrated in the 1924 paper. A sherd collection now available was not illustrated or tabulated by Fewkes and is pre- sented in this discussion; however, few nonceramic artifacts are in the Museum Weeden Island collections that were not illustrated or mentioned in the original report. Rather than to present new information, the purpose of this anal- ysis of Weeden Island is to review this key type site in Florida archeol- ogy so that the original data may be reappraised in the light of more recent findings. Site description and excavations ——The excavations of 1923-24, con- ducted in the field by Stirling, Reichard, and Hedberg, included nu- merous test pits and trenches made in the various mounds on the Weeden Island property and the more complete investigation of the burial mound. About one-third of the latter was removed. From these excavations Fewkes concluded that the group represented a single vil- lage unit, and he defined four types of mounds or works comprising such a unit. The first type is the refuse pile, mainly composed of shell and fragmentary artifacts, but no complete vessels. The largest mound at Weeden Island is such a rubbish dump. Two other types (between which I can discern no real difference) are both domiciliary mounds. These are small hummocks of sand, shell, and midden which con- tained no burials or ceremonial deposits of any sort. Fewkes found no post-mold patterns or other direct evidences of superimposed build- ings on these mounds so that his statement that they were house sub- structures is inferential. There is, however, evidence to support Fewkes’ assumption in that Moore’s excavations at more northerly sites of the Weeden Island culture occasionally revealed sterile sand mounds in group association with a typical burial mound. The fourth type listed by Fewkes is the small sand burial mound. The burial mound at Weeden Island was approximately 4 feet in height and circular in outline. Fewkes was of the opinion that the mound was a natural tumulus of wind-blown sand which was selected by the inhabitants of the site for burial purposes because its soft, unresistant nature was easily penetrated by the crude implements the Indians employed for grave digging. This seems extremely unlikely. 14 Weeden Island collections are numbered 325671, 326044-326081, 360387- 369390 in the U. S. National Museum catalog. WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—-WILLEY 107 In view of all the data gathered by Moore and others on Weeden Island burial mounds of the Gulf Coast it is more reasonable to assume that the mound was raised as a burial covering over the dead.*> Fewkes describes the stratification or structure of the burial mound as follows: Three layers, irregular in thickness, often lacking definite lines of separation, can be differentiated in Weeden Mound, but two of these are very evident, situ- ated stratigraphically one above the other and distinguished by the nature of their contents. The third or uppermost is naturally modern or deposited since the locality was abandoned for burial purposes. It is penetrated by roots of trees and gives every sign of having been formed by blown sand and reveals nothing of Indian manufacture; in other words, it seems to have formed after the Indians had ceased to use this mound for a cemetery. Its depth averages about 4 inches and is fairly continuous over the mound so far as it has been excavated. Immediately below this superficial deposit came the first of two strata which are supposed to indicate two successive occupations. The shallowest burial in this stratum was little more than 4 inches below the surface. ... Down to a depth of 3 feet we have at all intervals numerous fine examples of crania. Skulls and skeletons occur in numbers until we reach the lower portion of this stratum. The skeletons found through this layer were in bunches or bundles hastily de- posited in their graves and destitute of a covering of any kind..... Apparently a considerable time elapsed between the use of this place as a cemetery in the interval between the deposit of the lower stratum and the next in order. This is shown in places by black deposits of vegetable matter. The pottery of this layer is as a rule finely decorated and better made than that of the layer below. .... (Fewkes, 1924, pp. 10-11.) Stirling has recently clarified this somewhat and has placed a dif- ferent interpretation upon the total mound and submound deposition.*® According to his observations, the bottom layer referred to by Fewkes was the old original ground surface. This was covered with a zone of humus and decayed organic matter. In the natural sand and humus of this old surface a number of burials had been made in small, round pits which were lined with oyster shells. Sherds, one whole ves- sel (see Fewkes, 1924, pl. 4, A and D, pl. 21, C), and various other artifacts, came from the old surface stratum. Over this old cemetery the artificial mound of sand was later constructed. Disregarding the thin, 4-inch weathered or wind-blown top layer of the mound which Fewkes has described, Stirling divides the mound proper into an upper and lower zone. Physically these zones were not easily dis- cernible, but the nature of the burials found in the two zones distin- guished them from each other and from the old ground surface be- neath the mound proper. In the upper zone, virtually all of the burials 15 Dr. Stirling, who was in immediate charge of the excavations, is of the opinion that the tumulus was a man-made mortuary. Mi 16 Personal communication, 1947. 108 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. II3 were of the secondary vertical bundle type; single skulls and long bones without associated crania were also present. As opposed to this, the lower zone burials were mainly primary and extended. A third mode of burial was then found in the round pits in the old surface beneath the mound. These were primary and tightly flexed. Physical anthropology—tThirteen adult male and 22 adult fe- male skulls were examined from the Weeden Island burial mound (Hrdlicka, 1940, pp. 341 and 375). Only two of the females showed any evidence of artificial deformation. This was slight-to-moderate lateral-occipital. Male skulls were underformed. The average male cephalic index is 81.34 with a range of 76.0 to 87.9. Female average is 81.84 with a range of 77.4 to 88.8. Provenience of these skulls, with relation to mound proper (Weeden Island context) or submound (Glades Culture context), is not known. Pottery—Only a few vessels and sherds are illustrated in the Fewkes (1924) report, but a very great amount of ceramic material was taken from the mound excavations. Most of these unreported specimens were sherds of which only a few have been saved. The pottery illustrations in the published report (Fewkes 1924) are representative in the sense that they show that the ceramic com- plex from the mound is quite definitely Weeden Island in culture and period. Its selection as the type site is, indeed, justified. At the same time, the illustrated collection is selective on two counts. First, Fewkes selected whole or nearly whole vessels to be photographed; second, he appears to have exercised a certain esthetic judgment, thereby slighting some pottery types. From the preliminary report one would get the impression that check stamped and complicated stamped types were quite rare at Weeden Island. The reverse, however, is the case. Hundreds of stamped fragments, as well as incised and punctated pieces, were found in the mortuary mound. Similarly, ade- quate samples of the simpler, plain pottery from the old ground sur- face or submound level are not shown. In this last connection, both Fewkes and Stirling state that pottery was found at all depths throughout the mound as well as in the sub- mound level. They are quite clear in identifying the Weeden Island style pottery only with the mound proper where it had been placed as a mortuary offering. Fewkes also maintained that pottery was not as abundant in the submound as in the mound proper and that this sub- mound pottery was of a crude sort, quite different from that of the Weeden Island Complex found above. Stirling verifies this from memory ; but, unfortunately, sufficient pottery samples are no longer available from the submound stratum to prove definitely this stratifica- WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—-WILLEY 109 tion. Only one specimen, a large plain bowl (Fewkes, 1924, pl. 21, C), is designated as coming from this old level. This bowl is of the type Glades Plain. The illustrated pottery in the Fewkes report (1924) has been clas- sified below according to the types which have been set up for the Weeden Island Complex at large. In most cases these photographs have been checked against the original specimens, now on display in the National Museum. ERO ES Fact vt sehr: cen ciate All Weeden Island Punctated. 124, TOL Os ae Reema eos Weeden Island Zoned Red. [RAI Ste eens es aire acne ang Papys Bayou Punctated. RT ANIA pathos cies stoke ora Wakulla Check Stamped. Rite enn Gao eee Weeden Island Period sherds. LPAI 24 1 Bs ol DE 8) Sse Giclee Unclassified. ESTs (62) AN chore cist reteene Weeden Island Punctated. zie hia (2) Senin Gop hevorarnecrs Weeden Island Plain. [2a eed eee ee eae Weeden Island Punctated. EM Es oa ne avec ess Unclassified complicated stamped. EMT Aes ageeaclaa le eens Indian Pass Incised. a ND) rece etme cs, an nn) Weeden Island Punctated. 12, A 1 Cea eee ar Weeden Island Incised. [Pile Tis Rl Sl Se Weeden Island Punctated. 12) List Gal Da eee eee Weeden Island Incised. [210 ci Sl as Ae ea Weeden Island Punctated. ZONE AR DF onc repels nici Weeden Island Punctated. TL aay 1 al Se i rt eeaericiae Pinellas Incised. TRYG, ori) AAR soe te hotter Weeden Island Period sherds. [Ply Fics el Grae Ae Se eeincricy chee Glades Plain. LEAL, Zeit SE rete eens? Weeden Island Plain. The sherd collection now in the United States National Museum, and with one or two possible exceptions not previously illustrated by Fewkes, has been classified according to the same ceramic types as those applied to the illustrated vessels. All these sherds came from the body of the mound. It will be noted that no Glades Complex pottery is included. However, in addition to the Weeden Island Period pottery, which forms the bulk of the collection, a few pieces of Engle- wood and Safety Harbor Period styles were also present. It is pos- sible that the Englewood types represent a continuity with the Weeden Island, and they may have been manufactured at about the same time as there is reason to believe that the Englewood Period IIo SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I13 immediately follows out of and overlaps with the Weeden Island in the Tampa Bay-Charlotte Harbor section. On the other hand, it seems unlikely that the few Safety Harbor fragments also represent this continuity. It is more reasonable to believe that these specimens come from a later occupation of the site during which time a few typical sherds of the period were dropped on the mound or possibly placed with a later burial into the body of the mound. There is, unfortunately, no exact provenience for these few Safety Harbor sherds. The following tabulation is for the Weeden Island collection: Weeden Island Complex Weeden Island Series: Weedenwlsland (Plainicpyes crateieya softs cio brie er ee 23 Weeden Usland Tncised eis: as, sieiose msaye's se 3 = Ses ee 52 Weeden Island “Punctated oc. csc. cc's a dcnnis cee eee 49 Weeden! Islands ZonediRedtar seciecicccs «cits sce II Keith Tncised iaerektiac bin: cree are las accee sc. cee eee 26 Carrabelle, dncised! iss teen Se wiclsne copa cco ee olsen eae 82 Carrabelle ‘Punctated! i. ciesn ss asic oe. eie ion ee eee 2 (Indian. Pass ncisedieiss. ceases cin etuirnsit cae ce eee 15 Hillsborough Series: Hillsborough! Shells Stampeders ccc 7 Ruskin Dentatée "Stampedoc2..iee sc se's oes oie shawie se Oe eee 15 Complicated Stamped Series: Tampa Complicated “Stamped $0:. (2.220... ee nee II Sun! Gitys Complicated Stampeders eee cesses 12 Old. Bay: ‘Gomplicated Stamped. . 3 .c. shengeb secon e/etdcey'es aUeeopye GPA Ta eto els Sis Vos dO aE 2B Shell ipracelet: se ct...c:te a dao stars mero iaie see tite 6, bye ai Slegs, eee I Sirell pendants) oi pplimimetSraeetat rere chee ace y-7< 6 oy ey ore) een cre ean 2 Miscellaneous worked shells. Coquina-stone) perroratedmsimlvenspre eye ceea ss ie ele oe ee reser ee 2 Summary.—The Weeden Island burial mound is one of a mound group located on Old Tampa Bay. The other mounds of the group are rubbish heaps and, seemingly, artificial sand mounds which may have been the bases for temples or other prehistoric buildings made of perishable materials. About one-third of the burial mound was excavated by Dr. J. W: Fewkes, assisted by Dr. Stirling, Mr. Hedberg, and Mr. Reichard, during the winter of 1923-24. In a brief report (1924), Fewkes has discussed the composition of the mound and its contents. In his opinion the mound was a natural eminence which had been used by the Indians as a burial place at two different prehistoric periods. From what is known of burial mounds on the Florida Gulf Coast, and from the comments of Stir- ling who directed most of the investigation, it seems much more likely that the Weeden Island burial mound was artificially con- structed. According to Stirling, mound structure and burials within the structure took the following form. Flexed primary burials were found in the submound level or original ground surface beneath the mound proper. These burials were in small, round pits which had been lined with shells. Extended primary burials were found in the lower zone of the sand mound proper ; and secondary burials, mostly of the vertical bundle type, came exclusively from the upper zone of the mound proper. Both Stirling and Fewkes agree that the submound stratum con- tained little pottery and that this was simple and crude. The one available specimen from this stratum, a large food bowl, is of the type Glades Plain. Nearly all of the pottery, both vessels and sherds, which came from the mound proper is of the Weeden Island Complex. Shell tools, typical of the Glades Complex, including hafted conch weapons or hammers and celts, were found in the excavations, and it appears as though these came from the submound level. Unfor- tunately, this provenience for the shell tools, which would be of im- portance in substantiating the presence of a Glades archeological com- plex underlying the Weeden Island Period mound, is not verifiable. WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—WILLEY 113 Fewkes concluded that the remains from the submound were those of a people related to the primitive Ciboneys of Cuba. This group, he felt, were driven out by a people coming from the north with a more typically Southeastern, or even Muskogean culture. Leaving aside the ethnic identifications and the suggested relationships to the West Indies, there seems to be a very definite possibility that the Weeden Island burial-mound site is a stratified one. This stratifica- tion can be expressed by saying that the old submound surface deposi- tions and burials represent a manifestation of the Glades archeological culture which centers principally in south Florida but that the mound proper was built and used during the Weeden Island II Period. The lack of more material, particularly from the submound, and of more complete records, preclude a definite conclusion on this stratigraphy. Future excavations at the site may give such conclusive results. THOMAS, HILLSBOROUGH COUNTY (HI-1) Description of the site—The Thomas site, Hillsborough County, consists of a sand mound and an extensive shell midden located on the north bank of the Little Manatee River not far from the mouth of the stream (map 9). The mound is listed by Moore (1900, pp. 358-359) as “Mound near Little Manatee River, Hillsboro County.” However, the present owner is Rupert W. Thomas (as of 1936), and the site and mound are now commonly known by his name. The region in the neighborhood of the mouth of the Little Manatee River and Tampa Bay is typical of the central west coast of the Florida peninsula. The gray sandy soil is generally quite sterile and supports only a growth of palmettos, water-oak, pine trees, and oc- casional cabbage palms. The littoral of the bay is fringed with a heavy growth of mangroves, through whose root action innumerable small keys have been built up of a heavy black silt. The bay is shallow and placid, except during times of high winds, and is ad- mirably suited for navigation in small boats. Fresh water is available from numerous springs on the mainland, and the country must have supported a large fauna in pre-Columbian times. The shell midden on the Thomas property extends for about 60 meters along the shore of the river bank. It has been almost com- pletely removed by dredging operations, and no excavations were at- tempted in this part of the site. Its location along the stream bank and its proximity to the sand mound suggest that the midden area and the mound were, respectively, the habitation place and the mor- tuary of the occupants. The mound lies 70 yards (65 meters) north II4 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. II3 of the shell midden, away from the water’s edge. It was approximately 65 feet (20 meters) in diameter and 6 feet (2 meters) in height. The form was a roughly truncated cone. In composition it appeared to be wholly of sand, and excavation bore this out except for a scattering of black, shell-less midden layers along the northern and northwestern sides at a depth of over 3 feet (1 meter) below the surface. The mound showed evidence of considerable disturbance along its north edge, and this checks with Moore’s descriptions of his excavations in the mound. The following comment is from Moore’s account: The mound, irregularly circular and rather rugged as to its surface, has a base diameter of about 58 feet, a height of 6 feet. From the southwest side of the mound an aboriginal canal, almost straight, runs a distance of 238 feet to the water. Leaving the mound the canal is 64 feet across, converging to a width of 36 feet at its union with the water. The canal, in common with the field through which it runs, has been under cultivation, and consequently is irregular as to sides and bottom. The maximum depth is now 3 feet 3 inches, though, according to Mr. Hoey, 20 years ago, when he first came to the place, the sides were steeper and the canal about 2 feet deeper, so that high tides entered the field until a dam was placed across the mouth of the canal. Beginning in the marginal part of the northeast side of mound, a trench 35 feet across at the beginning was run 29 feet in to the center of the mound, where the trench had converged to a width of 9 feet. The mound was of pure white sand, unstratified. At the very outset burials were encountered. In all, 112 burials were met with, classing as such human remains with which the cranium was present and omitting a limited number of bones found loose in the mound. E xcavations.—Excavations under the sponsorship of the State of Florida and the Smithsonian Institution were begun on the Thomas mound late in the fall of 1935 and were terminated January 30, 1936. Preliminary survey revealed no trace whatsoever of the canal recorded by Mr. Moore; however, continued cultivation and plowing in the 30-year interval could easily account for its complete oblitera- tion. Coming in from the northern periphery of the mound were evidences of Moore’s earlier trench. In addition, there were a few other miscellaneous “pot holes” at various places on the mound sur- face. Preceding excavation, a grid system, 100 x 100 feet, was staked out over the mound at 5-foot intervals, and elevations were taken for each 5-foot intersection. Two trenches, 5 feet wide and 100 feet long, were opened along the western and northern edges of the mound. Potsherds were found in great numbers on these two peripheries. No burials were uncovered, however, until excavations proceeded farther into the body of the mound. Excavation was continued in progressive 5-foot strips, along both the north and west sides of the mound, and by smaller units of 5-foot blocks. Depth of excavation, which was carried to undisturbed soil, WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—WILLEY II5 ranged from 18 inches to 60 inches, depending upon the location in the mound. Potsherds, plain and decorated, were found throughout, although they seemed to be most concentrated in two layers. Both of these layers extended from the peripheries to the center of the mound; and at the center they were located at depths of 24 to 36 inches and 42 to 50 inches respectively. The total pottery count for the mound was 7,746 sherds and vessels.1* Twenty-one specimens were partial vessels or restorable vessels, including a complete and undecorated “toy” pot and an undecorated “utility” vessel. Pottery fragments were found nearby and surrounding burials in the mound. Direct association of mortuary pieces with individual skeletons could not be conclusively proved but was suggested in some instances. The present account of the Thomas mound is based chiefly upon the excavation notes, materials, and other data submitted by Preston Holder. This covers the period of work in late 1935 and early 1936. Subsequent to this J. Clarence Simpson, representing the Florida State Geological Survey, returned and excavated the remaining portions of the Thomas mound that had been left undisturbed by Holder. This later investigation was made during the months of June to August in 1937. A copy of Mr. Simpson’s field notes and two study photo- graphs of materials recovered from the mound are on file at the Bu- reau of American Ethnology.’® In the following discussions the ob- servations and data on the Simpson excavations will be separately designated. In addition to the final excavations in the Thomas mound proper, Simpson made several exploratory trenches in what were probably nearby piles of shell refuse, quite likely the midden at the water’s edge. Burials were found in some of these exploratory trenches, and several artifacts and potsherds were also recovered. None of the latter are illustrated in the photographs or described in detail; hence, it is impossible, from available data, to determine the period of the burials or rubbish. Mound structure-—Near the northern edge of the mound, at a depth of 42 to 50 inches below the surface, was an intermittent layer of black midden, the only strikingly distinct stratum or “lensing” in the mound which could be differentiated from the sand. The most intensive section of the above-mentioned lower pottery stratum, a small zone 28 x 36 inches in diameter and 4 to 5 inches thick, was found associated with this black midden layer. 18 As recorded by Holder during the 1935-36 season. 19 Simpson has published a very brief account of the 1937 work at the Thomas mound. (See Anon., 1937, pp. 109-116.) 116 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I13 Concerning the possibility of an inner mound the field notes state: While suggestions appeared in the field to support a thesis for the existence of two mounds, one of which was an older underlying mound about 20 to 30 feet in diameter with a rise of perhaps 2 feet which had subsequently been covered over by the present mound, it is doubtful that the evidence recovered will warrant such an hypothesis. The evidence of definite (physical) stratigraphy was dis- appearingly faint. As will be noted, the ceramic evidence from the mound does not help verify the presence of an inner mound culturally distinct from the upper or outer mound. Burials —There were 137 human burials in the portion of the mound which Holder excavated, almost the entire western half. These are in addition to the 112 burials which Moore had previously recovered from other sections of the mound and the 170-some re- moved by Simpson during the final excavations, a grand total of ap- proximately 419. Careful burial data were assembled upon 112 of the 137 burials discovered by Holder, the remaining 25 of this group being hastily removed at the close of his operations at the site in order to facilitate excavation to the lowest levels of the final blocks. Most of the burials which Holder found were in the southwestern quarter of the mound. In addition, he uncovered a smaller burial concentration in the west-central to northwest section. Most of the primary interments which he recorded were found in this west-central group at a depth of 50 to 62 inches below the surface. Only one secondary burial was observed by Holder at this lower depth; and conversely, only one primary burial was recovered in the upper part of the mound. The following is a summarization of the burial types found in the Thomas Mound (after Holder) : Secondary burials Werticalaibundlessene. eee ree oer 50 (see pl. 4, top, right) Fiorizontallsbundlewersseea ae eee 33 Group: vertical. bundle.2.s. .%2)..00.. 9.05.6 3 (individuals) Aberrant secondary burial............... I Primary burials Bleeds useah rok cs Seer ee eadech ete dence eee 6 (see pl. 4, top, left) Semiflexeds "ice etjas otto aoa ceeee 3 Brimatyvan (distin bed) meee eer etree eee I Cremation! J.ees cere: oeetvani oe ceewc I Isolated) skullstavteasske tetas eu ete eee 3 Too decayed for identification................ 2 Total) burials). .112 Secondary burials in addition to those classified.................005 25 Miscellaneots: bone! pile deka en. de eclosion ernie see ee ee ? WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—-WILLEY II7 Aside from possible associations of pottery fragments with indi- vidual burials no artifacts were found in immediate association with the bones. A few instances of charcoal, as though from an in situ fire, were seen in connection with burials, and occasional scraps of calcined human bone were found in or near the charcoal. One burial lay below some limestone rocks. Simpson’s data on the 1937 excavations confirm the above observa- tions. He mentions both full-flexed and semiflexed primary burials and secondary bundles of the vertical and horizontal types as well as single skulls. In several cases there is notice of cremation or partial cremation. As in the earlier excavations it was seen that the pri- mary burials were more often met with in the lower levels of the mound. Physical anthropology—Most of the skeletal material was in a very poor state of preservation, precluding a thorough analysis of the burials on the basis of age and sex, but some data were obtained. There were 18 adult males (over 20 years of age) ; 5 adult females ; 6 adults (not sexed) ; 3 female youths (12 to 20 years of age) ; and 2 children (3 to 12 years of age). Because of the very poor condition of the bones only 10 skulls, or partial skulls, were saved. Seven of these were sent to the United States National Museum. The field notes make only one reference to physical characteristics, describing three skulls. These crania showed pronounced prognathism, heavy supraorbital ridges, low and abruptly curving noses, low foreheads, and heavy and foreshortened mandibles. They are mesocephalic to brachycephalic. Associated long bones are described as heavy and somewhat short. One case of what seems to be slight artificial deformation (occipital ) was noted. The feature of a severely fractured left temporal was observed on six skulls, and there was one instance of a crushed right temporal. Pottery.—Pottery fragments were removed from the mound in arbitrary small levels from the 5-foot square excavation blocks; or they were removed from other specific locations where possible burial associations were encountered. The 7,746 sherds and restorable ves- sels which Holder took from the mound were divided into two gross lots, and 6ne-half of the material was shipped to the United States National Museum ?° in Washington for study while the other half was presented to the State of Florida. Only a very primary classifica- tion of “plain pottery,” “decorated pottery,” and “stamped pottery” 20 U.S.N.M. Nos. 384130-384291, 384312-384313. 118 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I13 was made in the field prior to the division. Further discards were made at the National Museum preparatory to storage; hence, the collection treated here is considerably smaller than the original lot ob- tained from the excavation. Eleven hundred forty-three sherds were examined and classified in the National Museum. This number is supplemented by a smaller collection of sherds which is now in the Florida State Museum and which was classified during the summer of 1947 by Dr. John M. Goggin. This second collection is appended separately below. Presumably, it is a part of the material obtained by J. Clarence Simpson in the 1937 excavations at Thomas. The Thomas collection compares favorably with that from the Weeden Island burial mound in that it is also quite typical of the Weeden Island Complex. The Weeden Island, Hillsborough, and Complicated Stamped Series are all well represented, as is the type Wakulla Check Stamped. These are marker types for the Weeden Island Period, particularly for Weeden Island II as it has been de- fined from the northwest coast. Chalky or temperless ware is more common at Thomas than at Weeden Island proper, although this may be due to selective discarding from the latter site. Biscayne types are numerically strong, and the Papys Bayou Series, which parallels the Weeden Island decorated types, but on the soft paste, occur in con- siderable quantity. The Little Manatee types, also of the soft, chalky ware, are present here as they were at Weeden Island. A significant difference between the two sites is in the plain wares. The residual type, Smooth Plain, is fairly abundant at Thomas but nonexistent at Weeden Island. Again, this may be due to selective discarding as might also be the frequent appearance of the types Glades Plain and Belle Glade Plain at the Thomas mound but not at Weeden Island. On the other hand, the Glades and Belle Glade types at Thomas may possibly be correlated with the primary burials found in the lower levels of the mound. If so, this would be somewhat comparable to the mound superposition at Weeden Island, with a Weeden Island Period mortuary constructed over an earlier burial place of the Glades culture. Such a stratification at Thomas seems less likely, however, than for Weeden Island. At the Weeden Island burial mound, primary, flexed burials and Glades type ware are said to have had a 100-percent correlation with the submound stratum. Only absolute material proof is lacking. At the Thomas burial mound there is more avail- able evidence, but it does not reinforce such a separation and corre- lation. In the first place, the lower stratum at Thomas seems to have been part of the artificial mound, not the old original ground surface WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—WILLEY II9g as at Weeden Island. Further, although the examination for stratig- raphy was based upon only about one-seventh of the total number of pottery specimens found in the mound, these available specimens were from locations in all parts and depths of the tumulus. One such distinction was noted, but this was with relation to the few Safety Harbor Period sherds which were found not far below the mound | surface and, presumably, represent a later intrusion into the Weeden Island Period structure. Thus, if the Thomas site is not a stratified one we are left with the possibility that the Glades and Belle Glade Plain types were con- temporaneous associations with the Weeden Island types and a part of the mixed ceramic complex of the people who built the mortuary mounds. This interpretation would not be too far out of line with what we know of Glades and Belle Glade wares. Weeden Island sherds were found in association with Glades and Belle Glade pottery at the Belle Glade site (Willey, n. d.). This being the case, a mixture of the two pottery styles and wares at any other site would be in order. As will be seen farther along, the type Glades Plain, with slight modifications, seems to have had an extremely long life span on the Gulf Coast in the Manatee region. There is every reason to believe that it had a temporal extension from pre-Weeden Island up to historic times. The following is a classification by types of the available pottery from the Thomas: Weeden Island Complex Weeden Island Series: Micedenmisland Plain scotia chat sive esene voemee ret es 102 Wreedenwisiand jInicised-ic85 s.cciis 26 ees kcceoncmmeeice 154 Wieedenwlsiandeunctatedmcnec-j1aceee ees sereeoe cies 187 Weeden Island Zoned: Red tial. Sioee «ceca cee aide ts 14 eertin, Tre rseiiy i.e oetertAsyeioclere oloibee eee eee eras eee cicts 12 Garrabelle incised: -frttectok se elas ol ae eieiee als II CarrahellesPunctated: aci.40.2.3'vsels « Seen eee te I Rndiae pPass> Incised cc cae necies evicwe we Rites oe eisler helene 4 AGKer Id Se=pitiGhed!! sc arapeisters sieve sistehs Store ales svers eratciemieheis I Hare Hammock ‘Surface Indented. 0.5.05. 5 0000555 c¢e0s 4 Hillsborough Series : Hilishorough: Shell Stamped’: ot. oj. < aalecroe eee 7 Payps Bayou Series: Papys, Bayoupbunctateds ernie siers crlee eriaeocie eee 19 St. Petersburg Incised..2 52%. : scacte aceon eae eee eee ee 5 WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—WILLEY 133 Glades Complex Glades Series: Gladesee aint se ae te ects soko ee et 87 IMaamielneisedieee.< citric s aire 3c 4 Onto On eIe nae Toa Se I Safety Harbor Complex Safety Harbor Series: Saretymlarborbncised ihe... sieco sie s rete ota a oaaroetes eve 21 Combination Safety Harbor Incised and Simple Stamped...... 2 Miscellaneous Syrsteyeyd cag! Ad ENE agin ino cl cia dia Bier Aare ACIS nico ocko.cpatrendig in 2 (Unclassified Incised:and/om Punctated.-.+..+...se eee sess oe 2 Total sherds.... 381 The Florida State Museum, Gainesville, collection is classified as follows: Englewood Complex Einolewoodaiylatite yrrastereercee eistleye scresicics@ = cletoyeha ever stoi steisietera ovet- ays 22 Biscayne Series: Biscaynea@hecky Stampeds samiacerscciice ociet crlerictect- bletke = II Biscayries se lain ire gta acc vse cores Oe Oe wie tole Riorchakelotecas Oars shes 4 Biscayie vOord-marked: 2 vadc os sres ede seen oes eres 2 Weeden Island Complex Wwieedermlislcricl uit ett ppris sya eyerers rovavenskete i oosvave eis \s ee oreseye crerorarcrerelatsyete I Walla Cheek Stamped crys «vielcreters c1s-ace 4) s/cieloyate jecencucnei ae eters pic 3 Glades Complex BellerGladewP lain -aines hwintod she Sanrio adinte cies lee 3 Miscellaneous WiGlaSSified’s SAiencce seh ace sels ao secs ore Sera eiatoe Resa etecrnsraT eons a Total sherds.... 49 The collection from the museum at Tallahassee is classified as follows: Englewood Complex Hee We Orne tiles tee wae, 0 Fare nears & oelersl ls SheleimGAye sia Saal a dicia % 3.8 7 Sarasotamlticisedaane ssc cecic tia totes olen oe aS oie to cle afore oi anorshee eo 3 Biscayne Series: Biscayne Checks totammned sevien sietcisige cersisie wings ciesin ee te ms 20 Biscaynieu Plants apaa ecco cs tena wee wire a alee intents 9 iS Cary w Nec aeraravateretaictetotere ciate etme tetorexer ote te reiete ore reieter vercie ee 3 134 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. TES Weeden Island Complex Carrabelle Punctated: 25044555 ss. e4< 5c eee ee eee I Wakulla Check (Stampedtiy-ee.cne see os eee eee I Little Manatee Zoned Stamped sr: vc:saionnssvncie eee mre I Papys Bayou Series: Papys Bayo tain access i006 's0. PINELLAS COUNTY (PI-2) Description of the site—The Safety Harbor site, near the head of Old Tampa Bay in Pinellas County, is situated on a little spur 25 This site is also called Phillippi Hammock or Phillippi’s Point (Walker, 1880a, pp. 410-411). It is generally believed to be the Tocobaga capital of the sixteenth and seventeenth centuries. 136 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I13 of land which extends out into the bay at a point about 1.5 miles north of the town of Safety Harbor (map 9). The waters of the bay have gradually encroached upon the site so that a long sandspit which extends eastward from the principal mound, and is now subject to tidal action, was formerly dry land and a part of the village to which the mound belonged. This principal mound is a flat-topped, rectangular shell mound, presumably an artificial platform for a temple or building. It measures about 70 feet square and is somewhere be- tween 15 and 20 feet high. At distances of 50 and 100 yards west of this large mound are two tumuli of shell. These are circular and dome-shaped and may be midden piles. The burial mound at the site lies 400 yards west of the big flat-topped mound. It is approximately 8o feet in diameter and 10 to 12 feet high.?® Previous to the Bureau of American Ethnology excavations it had been considerably dug over, and the old pits revealed that the mound had been composed chiefly of sand. Excavations —Permission to excavate was given to Dr. Stirling in 1930 by Col. Thomas Palmer, owner of the property. Excavation centered mainly on the burial mound which was removed in about one-quarter of its area down to subsoil base. It was revealed that the burial mound had been built up in successive stages of sand layers. Burials were of the secondary type and were placed without any par- ticular order throughout the mound. Over 100 were removed in the course of excavations. Pottery was found principally in a mortuary deposit on mound base near the edge of the mound. Many of these vessels were intentionally perforated or “killed.” A few additional sherds were also scattered through the sand fill of the mound struc- ture. All European artifacts, such as silver tubular beads and an iron ax, were found associated with burials in the upper portion of the mound. No comparable trade materials were found with the lower burials or in the mortuary deposit on mound base. Some small test diggings were also made in the area between the platform mound and the burial mound. This was the village occupa- tion area of Safety Harbor, and in the following lists of pottery and artifacts, materials from this zone of the site are designated as “Safety Harbor Village” to distinguish them from those of the mound. Physical anthropology.—tThe skeletal material from the Safety Harbor burial mound was examined by Hrdlicka (1940, pp. 339-340 26 All these distances are estimates given to me by Dr. M. W. Stirling, June 1947. WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—WILLEY 137 and 373). In the 27 males and 23 females studied there was no defi- nite example of cranial flattening. The cephalic index for males ranged from 73.7 to 85.5 with an average of 80.3. The cranial mean height index on males, on the basis of 11 skulls, had an average of 87.7. Among the females the cephalic index range was 74.6 to 87.7, averaging 81.8. Cranial mean height index, on the basis of 10 skulls, had an average of 86.3. Pottery.—Two small pottery sherd collections, one from the Safety Harbor burial mound ** and one from the village site,** were ex- amined. These collections represent the totality of the decorated pot- tery found at the sites, plus a few samples of the plain types. In the burial mound, the bulk of the pottery, from mound base, belongs to the Safety Harbor Series. These are the types Safety Harbor Incised and Pinellas Plain and Incised. Presumably this was the intentionally destroyed mortuary pottery of the tumulus. In addition, a few sherds of the Biscayne and Glades Series also came from the mound. As compared to the Englewood mound, the pottery from Safety Harbor shows appreciable Mississippian influence. The types Pinellas Plain and Pinellas Incised are closely related to the Fort Walton and Lake Jackson types of northwest Florida. These types were found in the Safety Harbor village as well as the mound, and the village ex- cavations also yielded some typical Fort Walton Incised fragments as well as one sherd painted red-on-buff with finely pounded shell as temper. In total, the Safety Harbor ceramics, from both the mound and village, appear to be later than those of the Englewood mound. Added to the strong Mississippian influences which are pres- ent at Safety Harbor, is the absence of any Weeden Island Period types such as those found at Englewood. These Weeden Island sherds at Englewood implied a temporal overlap with the earlier period, an implication lacking in the Safety Harbor collections. Then, intrinsically, the Safety Harbor Series types are stylistically degenerate. Their manufacture is poor; the designs are badly conceived and slovenly executed. This, coupled with the fact that they bear a very definite relationship to Englewood and Weeden Island types, suggests a decline in the ceramic art which took place with the European con- tact period. In connection with the Safety Harbor pottery, it should be noted that Glades Plain and Biscayne Check Stamped continue into the Safety Harbor Period. 27U.S.N.M. Nos. 351513-351525. 28 U.S.N.M. Nos. 351526-351536, 362378-362386. ET 138 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I13 The pottery from the Safety Harbor burial mound may be classified as follows: Safety Harbor Complex Safety Harbor Series: Safety Harbor’ imeised 2or. DIRECT CREMATIONS © SECONDARY CREMATIONS @ UNCREMATED BURIALS O CHARCOAL PITS ° oO ° ° ° ° ° ° ° > GO GLE OEE ° 0° 0% 0h 0 0 0 0000000000000090009000000 S PARRISH MOUND NO. 2 MANATEE CO., FLORIDA SECTION THROUGH A-B Map 11.—Plat map and cross section of structure, Parrish mound 2 (Ma-2). WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—WILLEY 149 with some cremated remains. There were also instances of European artifacts in association with, or nearby, other cremation burials. In addition to the burials excavated by Reichard, Wesley Parrish, living in the vicinity, reported that earlier excavators had taken three human skulls from the mound. An analysis of these data lead most easily to the following con- clusions as to the nature and history of the mound construction. An original crematory pit was excavated on the old ground surface and a considerable number of bodies were burned in this pit. Over the pit the circular mound of sand was heaped up to a height of about 6 feet. This mound building may have been accomplished over quite a period during which time the various cremated burials, both sec- ondary and in situ, were placed in the fill; on the other hand, the mound may have been piled up in a single, rapid operation with the burials made later. In any event, after it was built it was obviously used as the base for a building constructed of wooden posts. This building or enclosure may or may not have been roofed over. In one corner of the building there was either a crematory platform or an altar for the bones of the dead (see map 11). It may be that bodies or skeletal remains were cremated here purposefully and subse- quently buried in the mound. The other interpretation, that of a bone repository, coincides with the known southeastern practice of keeping the denuded bones of dead tribal dignitaries in hampers or boxes within the mound temples. If this was the case at Parrish mound 2, the charred mass of bones in the corner of the structure or temple could be the result of the complete burning of the structure which thus destroyed the bone cache that had been placed on a platform or altar. The destruction of the temple mortuary by fire may have been a purposeful ceremonial act, or it may have been accidental or the result of warlike depredations. The results would have been much the same. That the structure was burned seems indicated by the charring of what was the outer surface of the wall posts and by the scattered charred timbers found at what was approximately the floor level of the structure. That the lower portions of the wall posts did not burn completely through is to be expected in that they were im- bedded in the sand of the mound. Pottery—No complete vessels from the mound were found in the excavations conducted by Reichard, but Stirling, in his preliminary report states: A very interesting owl-effigy water bottle of a familiar lower Mississippi type was taken from the central part of the mound during a previous excava- tion. (Stirling, 1935, p. 380.) 150 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I13 Such a piece as he describes would be chronologically consistent with the Safety Harbor sherd types found in the mound fill. Again Belle Glade Plain and Glades Plain are present. The mound undoubtedly dates from the Safety Harbor Period and is contemporaneous with Parrish mound 1. The presence of the Glades Complex types raises the same question as it did in Parrish mound 1, and the probabilities favor their association with the Safety Harbor Period rather than earlier materials accidentally included in the body of the mound. Pottery vessels and sherds from Parrish mound 2 are tabulated as follows: Safety Harbor Complex Safety Harbor Series: Pinellas: Plain'y viccisrsins ess brogiea ae soos she hae Genie ee Cee 18 Pinellas Incised: 1: S8e. . esicceteenee eressiete evele dueihe ate Sete cree 22 Glades Complex Glades Series: Glades (Plain xh. asec sacs Oat tie necator CeCe ae I Belle Glade Series: BellerGlade! Plain. cvatacnce wie site co crete tee eee cee Eee 17 Total sherds.... 58 Artifacts—The nonceramic artifacts from Parrish mound 2 are of the same general categories and types as those from mound 1. Projectile points are stemless ; conch-shell cups are present but shell tools are scarce; and European trade goods were included in the mound. The latter are less common than in mound 1, but three small seed beads and a bone or tortoise-shell comb, all identical with those from mound I, were recorded. Because of the accident of preservation resulting from the burning of the structure within mound 2 a few carved wooden specimens and some rope or cordage were recovered (pl. 59, m-q). These are almost certainly of aboriginal make. Nonceramic artifacts from Parrish mound 2 are tabulated as follows: Stone: Projectile points: Unstemmed; (small triangilarsaeee cre cleo ele terereeneiee 3 (Between 2 and 3 cm. in length.) Tubular bead (CP) :cc ce 2eite sowie arecins oe os onic oles aie I (3.4 cm. long x 1.2 cm. in diameter. Hole is biconically drilled.) PJtamimety ds cdvors orecseerciete ole wetoiere Chere tele cistern eee I (Crudely worked, single-grooved. Measures 4.7 x 2.8 cm.) WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—WILLEY I5I Shell : Conchectnse OteGip pers tasiaes ciesanise ceiceectacheiseht wemrtion erelentere 4 (“Killed” or perforated.) Couch pick (UBusyCom COnIFGrHIN) so. co ess cai co es vane dees I Gincubar beads stere isa. ccaiecd etek aie ats Sots wn shesoloe elete are ays) oatorereraeare 2 (2 cm. in diameter and about 5 mm. thick.) Wood: Ganyed. traements: (charted. 24... ca'os ats nsieds «osetia ais we 2 (One piece has close-spaced incised lines. The other has deep central indentation encircled by a ridge; this, in turn, is en- closed by a spiral ridge which is ticked off with close- spaced notches. The total design on the second piece mea- sures 2.2 cm. in diameter.) Textile: Cordace or trope (chatred)ac4.o$.¢e.0cewesei a several fragments (Appears to have been made from hair.) European: BSCR CAUSE CLI EEIASS iaia's oie 2 cts a aroha acavare eieiciciers wate MOMs we wiaresatenee 3 (Light blue and white.) ESAS ECE AME ens Sand aetna saree ateyars aes esiseiA se eislalsisibis vlwiste er tea reece a's Wie ae oe clare see's ewe ave 3 164 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 113 Secondary: Bundle. (pls.4; bottom) is 5/208wctratapaaste sone, tr< ols Stamped sarsnc-sciae tc cloe asic oats lose oc tels siereieeuiane I Miscellanous Smoother se laine a5 sepa Ae ee CT ee etree Soe 2 estdtial Mclain» oer eee ak Cire ae eter ante eintelere ie ate rater 2 eine lasSiited tet att ase oe cis a Sete eee ers meee I Total sherds.. 9 The very limited evidence indicates the Weeden Island II Period. Rocky Bayou, west (Ok-3).—Moore (1901, p. 455) describes a mound on the west side of this bayou about 1.5 miles up from the mouth (map 14). The area was visited in 1940. No shell midden was discovered anywhere in the immediate vicinity and no surface material was found. The mound was a sand tumulus, circular in outline, 2 feet high, and 28 feet in diameter. Moore discovered three secondary burials 212 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. II3 of the single skull form and a mass deposit of intentionally destroyed pottery. Scattered pottery caches and other artifacts were found at various points in the mound. A stone equal-arm elbow pipe and a pointed-poll stone celt *? were among the latter. Most of the pottery was perforated or “killed.”’ A plain, gourd-shaped, flattened-globular bowl is illustrated by Moore (1901, fig. 47). This specimen is prob- ably Weeden Island Plain. This identification would seem to be cor- roborated by two vessels in the Peabody Museum, Harvard Uni- versity. These are cataloged as Rocky Bayou and are, respectively, Carrabelle Punctated (No. 56370) and Weeden Island Plain (red painted) (No. 56729). It is most likely that the mound dates from the Weeden Island Period. Black Point (Ok-4).—The Black Point mound excavated by Moore (1901, pp. 454-455) is one-quarter of a mile northwest of the point (map 14). This was a circular, flat-topped mound, over 80 feet in diameter. The height is not given. Trenching revealed check and complicated stamped and incised sherds in the sand fill but not burials or mortuary deposits. Moore considered the mound as domiciliary. This mound was not relocated in the 1940 survey, but a sherd col- lection was made along the bay bluff about 1 mile west of Black Point. Sand-dune formations are developing on the beach in front of the cliff, and sherds and shells are found scattered over the dunes where they have eroded and blown out from the original midden deposit. Our collection from this site was augmented by another given to us by William Meigs, of Niceville, Fla., who also accompanied us to the site. Surface collection from Black Point: Weeden Island Complex Weeden Island Series: Weeden Island Incised «..)./i05:<..<'s <.«-«jo'siclais-» ele ote le 15 Carrabelle: Incised’ 20 23\2 sayeie cau actos lets oo ee eee 49 Carrabelle>Punctated :...2:...0me scien cease cee ee oe Oe 15 Keith “Incised ec2.sias 8 oh ee oe ake decisis «oe ose CREE 4 Tucker Ridge=pinched) oie smi. eiusss Salers oie wen eee 5 Weeden Island. Plain. scise.c.c.01eoeesesearsisie else oe 5 Wakulla ‘Check ‘Stamped: :)....... swti os <1 sie cane ss ss ee 146 St. Petersburg Uncised tir oie cee eens sco coon eee 2 Complicated Stamped Series: Swift Greek Complicated ‘Stamped: . .2<<\:% s7. 2. see ee 3 St. Andrews Complicated Stamped... ... «s/2...ccteeeee I 42 Apparently this celt was given to Peabody Museum, Harvard University, where it is cataloged as No. 18580. WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—-WILLEY 213 Santa Rosa-Swift Creek Complex Jae TA TiN MENG OOS GOO ooo OO oO COCR MOB EE CnG TEBE OC OOnCn ane I Miscellaneous Smooth Play Wesfeaie atoisvs ore eis cinta ace o Sates ates dd SeeeeOnte. Seale eas 6 Pl aitne ted ttrararate stateieet ciara si aust aale Siciccsts oleae orto ce aan Ries I ReSidittal sy lattimrrctsscts tests, sho Vote esse aero ee eee 310 PsISeteCtnIALes StAMIPed » % .. «cade a.cldsisre vie vis lace cielo elowies » aeRO 20 Othenmanclassthedy &viee sioner ote etince cece cee 19 Total sherds.... 602 There is little doubt that this large collection fits the Weeden Island II Period. Site between Florosa and Mary Esther (Ok-5).—This is a midden on the mainland shore of Santa Rosa Sound near the mouth of a little stream which empties into the sound (map 14). The exposed shell midden can be traced along the bank for approximately 200 meters east of the stream. In no place does the deposit appear to be very deep. A small and inconclusive surface collection was gathered from the site : Fort Walton Complex PN eeu eee U aT a ey cte Norse ea aiove Sani ta ne cig eames wise ee ore oes I Weeden Island Complex Wialculilaa Check tStampedits vasreste acne sie sale ected slesie ace siate eleare.c 3 Miscellaneous Resid dal Wlaanil strc mtusestets eecrnatee who atewen tej oe Sal pois hemix sions sere 18 Wiriclassificdwynt ie) Gacraeee oesic Set cars bao does ceblocciecis slap iee « 5 Total sherds.... 27 Fort Walton (Ok-6).—Our 1940 excavations at the Fort Walton site have been treated separately under the section, “Excavations on the Northwest Coast: 1940.” A brief résumé of the operations of Walker and Moore at the Fort Walton site are also included with that treatment. The 1940 excavations were, however, concerned with the village midden at Fort Walton, not the large temple mound. The present review and analysis is based upon this mound and its contents, particularly as these were revealed by Moore. The Fort Walton temple mound is rectangular, flat-topped, about 12 feet high, and has a graded approach. It has been built in layers or stages. Sixty-six burials were recovered by Moore from the summit and sides of the mound; none of these had penetrated to any great depth and all were, apparently, intrusive into the mound. Sec- 214 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I13 ondary (single skull and bunched) and primary burials (extended) were encountered. Pottery and other artifacts were found with individual burials. In some cases bowls were found inverted over crania. What appeared to be ceremonially destroyed caches of pot- tery were also found in the mound apart from burial associations. Nonceramic artifacts include: Chipped-stone points and scrapers, stone hones, hammerstones, discoidal stones, stone chisels, stone and bone beads, shell beads, and shell ear pins. Occasional pottery vessels were “killed” by perforation. Loop handles, polished black ware, and shell temper are three general ceramic traits which held Moore’s attention. The illustrated pottery (Moore, 1901, pp. 435-454) is clas- sified as follows: Fort Walton Complex Fort Walton Series: Fort Walton Incised....... Figs. 15, 16, 17, 22, 23, 24, 25, 20, 30, 31, 40, 41 Poimt Washington ‘Tncised.. .:\c.0 0s. ence snes aoc ene eeee Fig. 18 Pensacola Series : 43 Pensacolavsineised) iene cescc cent Figs. 20, 21, 32, 34, 30, 37, 39 Pensacola Plain meses ciaisisieeine Selo oe ee Fig. 27 MoundyillevEnp raved i uvicdme,stciaceee cen eons eee Fig. 19 (?) Fort; Walton effigy shandles. <:4 .:... 0. .%2. «0s cease ae Fig. 42 Other St. Andrews Complicated Stamped... «<2 2/0. «sss eee Fig. 44 Sherd (with® Scat’) deSiguiz’. ./c.../s\.0,+.0.c cet siete «tee eee Fig. 45 (A similar element occurs on a vase from a Weeden Island Period site at Strange’s Landing. See Moore, 1902, fig. 117.) The burials in the Fort Walton flat-topped mound are all undoubt- edly of the Fort Walton Period, and the mound almost certainly dates from the same time. The pottery identifications, as these concern the Fort Walton and Pensacola Series, are open to some question, as it is difficult to recognize temper type from a photograph; however, some of the original specimens were checked, and it is certain that both the grit- and shell-tempered types occur at the site. The St. Andrews Complicated Stamped sherd is quite clearly from an earlier horizon and was probably an accidental inclusion in the mound fill. This, of course, is not unusual, as the surrounding midden area shows Dept- ford, Santa Rosa-Swift Creek, and Weeden Island types as well as those of the Fort Walton Period. The “cat-face” sherd is more difficult to place. It may be either Weeden Island or Fort Walton. 43 U. S. National Museum collections Nos. 58789-58798, contributed by S. T. Walker, consist of nine Pensacola Series sherds. These came from the same mound. WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—-WILLEY 215 WALTON COUNTY Okaloosa-Walton county line (WI-1)—A midden of clam shells is cut through by Florida Highway Io at a point 2.2 miles east of the ov ¥ el ST ANOREWS BAr Map 15.—Site map of Walton, Holmes, Washington, and Bay Counties. Sites indicated by numbers. Okaloosa-Walton line (map 15). The location is on a cliff above the bay beach, and the total area of the shell is about 50 meters in diame- ter. The refuse does not appear to be either deep or to contain many sherds. 216 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 113 Surface collection: Deptford Complex Deptiord Bold) Cheek eStampedteesateca sc acietine ee ieee tate 9 Deptiord Simple ‘Stampedsjecercn cee ocr es ae ste sero eee 3 Miscellaneous Residual! Plata’ 52:2 aissiateecscisios stores cis save osleelece ree Terteeenne 12 Total sherds.... 24 The small collection belongs to the Deptford Period. Villa Tasso (WI-2).—On the north shore of Choctawhatchee Bay, near the Okaloosa-Walton county line, is the small resort, Villa Tasso (map 15). An unevenly distributed shell midden extends for about 200 meters west of the main lodge of the resort. The deposits are in the form of small piles of clam shells, rising .25 to .50 meter above the surrounding sand. The bay beach is 40 to 80 meters distant. Surface collection from Villa Tasso: Fort Walton Complex Fort)) Walton™ Incised ec .fo siete eicicse clsiscsis 6.5/4 0.61 ererevetenedels soot 2 Safety Flarbor. Incised*s a acceas eclece oa sri de) chore) shovel ree ene I Weeden Island Complex Weeden Island Series: Weeden dsland WRlain.Goiisciies ccc esse eee ete eee I Garrabelle Encised. .eisjer otc cic «ores ais aro orsterote eistehe sree eee I Garrabelle Punctatedy jc.acclac coe cis coe oete cee 4 SwilttGreeky Complicated Stampedacss sce seers eee eerie 2 Miscellaneous Residual» Plaine vv; cays: oh w-sisile dos ee oeaaceeD Oe Crees 87 Total sherds.... 08 The collection indicates both Weeden Island and Fort Walton Periods. Big Hammock (WI-3).—Four miles east of the Okaloosa-Walton county line, on Florida Highway 10, is another midden of clam and oyster shells (map 15). This site is also cut through by the modern roadway, and the thin refuse deposit is revealed in cross section. The location is not more than 20 meters back from the bay bluff. Somewhere in this same neighborhood Moore (1918, pp. 532-533) excavated four sand mounds. These were all low, circular mounds. One contained evidences of burials, the others simply an occasional sherd or artifact. From the descriptions none can be identified as to cultural period. They were not located during the 1940 survey. WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—WILLEY 217 A sherd collection made from the midden in 1940 is classified as: Fort Walton Complex Pensacola eUnciSederct.rn aic.s save cis eee ecg sievete ctuiers e etietoleto avete ckecthaldl 2 Weeden Island Complex Weeden Island Series: 15eit 3 dled baveriey ota Ie ete eco Me tiotcwercoe c aoe ootice. co mae ma oe 2 Mnckernskid @e-pinchedt. jeseci ls ciniek ine co.ctis ton erent: I WrakullanGheek« Stampedua sachs mera ctemmniners acre a cieae eoeryocine 8 Bi HCteeS ULE: LACISERs « a cvcte sis crerntiets Sais © ot sia sae e ate sy siateiayeie sr I Miscellaneous HReSt ita le latral byse versa cies creates ete cre woreccioteis fale, os Staveveneiste alee ie se 42 Girtclassitiedd gues cvevereterste ore evsiate ecole Gr evone ail lorapscchate et erscopetovernyorsyalisispe¥e fereus 7 Total sherds.. 64 Predominantly Weeden Island II but with some evidences of Fort Walton Period occupation. McBee’s mound (WI-4.)—One mile northeast of Piney Point and .6 mile inland from the shore of Alaqua Bayou is a sand mound in the midst of a thick hammock (map 15). The property is owned by Alton McBee. The diameter of the mound is about 10 meters and the height about 1 meter. It had recently (1940) been excavated by unknown parties. A number of sherds were found in the fresh exca- vations. Mr. McBee is in possession of a stone celt which came from this mound. The artifact is 10 cm. long, 4 cm. thick, slightly flattened but with a pointed poll. A surface collection from the McBee site is listed below: Fort Walton Complex Fort Walton Series: Hort Waltons dncisedes wane scr cere cee cee Oca e ease 9 Makes TACKSOIMME Malt set ae, ants A iel< spa Cc cia gies orotate 3 Pensacola Series: Rensacolamincisedy s.tvaunicmrctce Hor oye eee ee 6 Hensacolawe latipes. cctererccecerctans che lroniees ore ase aren tetorerepet oie aee 17 Miscellaneous TE ebay Lee etG tae ean tO ASE ee OR NG OO DCS OOM DOTA SoG Ocoee I Smoothie lattioacnn cys o esiaves cyeteie ee level eerue terete oie aerate vets s 2 ReescttleTrllmeenati cia tt rcvc ac ina, aay elo ute tie miapalehe ee MS ae Taos a ea 30 Total sherds 68 It is questionable whether the McBee site is a village area, a domi- ciliary mound of some sort, or a sand burial mound. No evidence of burials was seen or reported. At the same time, shell refuse, generally the indicator of a village or living area, was absent. The site is defi- nitely of the Fort Walton Period. Piney Point (WI-5).—Alaqua Bayou is a small basin in the north- 16 218 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I13 east of Choctawhatchee Bay, and Piney Point is a sandy, grass-grown peninsula, dotted with pine trees, which extends out into the mouth of Alaqua Bayou from the east (map 15). There are two extensive shell-midden areas near the point. The first and largest is north of the point. The second lies 300 meters away from the point to the southeast. Both front on the water’s edge. A number of old excava- tions in the northernmost midden show oyster and clam shells and sherds. Depth of the refuse is probably about 1 meter. The 1940 collections from the two middens were combined and are given below: Fort Walton Complex Fort Walton Series: Fort. Walton Anctsed: 22.).. sods has cen csc ane ee eee ii Lake Jackson Plata oa05 sscesac te cscs s oes caeetee ee eee 3 Pensacola Series: Pensteola: incised) oscieliictespanies anesacshiet eee 2 Pensacola: Plain 0s. ./42¢.2s0,«acss weenie acne eee 6 Weeden Island Complex Wakulla. Check. Stanrped ss... Carrabelle: Punctated | cccwes sce aco ssn acete eee Figs. 66, 72 Tucker Ridee-pinched! jis 0. aie sc.su-resaicve cleric /eteleerer terete Fig. 65 Switt ‘Creek Complicated Stamped. ....°: Sav. e.-cate erate Figs. 60, 73 (Late Variety.) In addition to the specimens illustrated by Moore, three additional vessels from the same mound were examined in the Peabody Museum, Harvard University. These are all Weeden Island Period types, two being Weeden Island Plain (Nos. 56738, 56725) and one Carrabelle Punctated (No. 56726). Pippen’s Lake (Wl-12)—This sand mound was excavated by Moore (1918, pp. 531-532). It is situated near a small lake not far from the Choctawhatchee Bay shore just across the line from Oka- loosa County (map 15). In 1940 we were unable to relocate the site 4% 43a J. W. Griffin and R. P. Bullen inform me (April 1949) that the Pippen’s Lake mound is on the Okaloosa side of the line, Moore to the contrary not- withstanding. WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—WILLEY 223 The mound was circular, 2.5 feet high and 35 feet in diameter. Burials may once have been made in it, but no trace of them could be found in Moore’s digging. He found, however, a mass ceremonial- pottery deposit and several scattered pieces of sheet mica and hema- tite fragments. The pottery was intentionally perforated. Moore mentions no check or complicated stamped wares. He illustrates (1918) three vessels which are classified below: Weeden Island Complex Weeden Island Series: Weeden: Ustand) Mnictsed tics cctereonisastateiniein elerecs. gs pc's sa tn atetatate ars aime eee ore d ta aswel oe Fig. 108 Weeden Island Complex Weeden Island Series: Weeden: Island) Incised sc cccier cls /oreiesve ve are ces Figs. 102, 112 (?) WVecoenvisiand Plaine’ sisi san ceceee-cacecu mn caveee Fig. 104 Swift) Creek Complicated. Stamped ss cjsiais scicisus ons s soielon Fig. 120 (Late Variety.) The R. S. Peabody Foundation has additional material from the Moore excavations at Chipola Cut-off. This lot includes one Lake Jackson Plain vessel (No. 39267), one Fort Walton Incised vessel (No. 39313), and one St. Petersburg Incised jar (No. 39053). Non- ceramic specimens from the same site are a shell spoon, a Busycon 250 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I13 celt, a Fasciolaria chisel, and a single-grooved columella pendant or plummet. This mixture of both Fort Walton and Weeden Island specimens in the same mound is almost without parallel in Moore’s excavations. His excavation data do not, unfortunately, give sufficient detail to be of much help in explaining this situation. Two possibilities sug- gest themselves. One is that of intrusion, the assumption being that Fort Walton Period peoples utilized a Weeden Island Period burial mound for a cemetery. If this is so, the intrusive burials must have been made at considerable depths, as some of the skeletons accom- panied by brass ornaments of European origin were found in sub- mound pits. Furthermore, the intrusive diggings must have been very extensive in order to have placed a large cache of Fort Walton Period pottery in the mound in addition to the many graves. The second possibility is that a Fort Walton community had retained a number of Weeden Island pottery vessels in a mound of their con- struction. One objection to a continuity of this sort is that Wakulla Check Stamped, the marker type of the Weeden Island II Period, is absent from the mound, implying that the Weeden Island component belongs to the earlier or Weeden Island I Period. It would be more reasonable to expect continuity between Weeden Island II and Fort Walton than from Weeden Island I to Fort Walton. Both hypotheses are possible, but neither is provable from the avail- able data. The only conclusion we can be certain of is that the mound represents both Weeden Island and Fort Walton Periods. St. Josephs Bay (Gu-6).—This site is not located but only listed on map 16. In 1893-94 C. H. B. Floyd sent a collection of vessels, some stone celts, and shell artifacts to the United States National Museum (Nos. 155318-155329) from a mound which he located as about 25 miles from Apalachicola on St. Josephs Bay. At that time it was identified as “Franklin County,” but today a St. Josephs Bay location would be in Gulf County. There is the possibility that this might be the Gotier Hammock (Gu-2) mound which Moore found so thoroughly dug over on his 1902 expedition. The collection is, however, a fine one and a distinct period lot. It is also illustrated and well known from the literature (Holmes, 1903, pl. LX XVIII). Late Variety Swift Creek, Weeden Island Plain, and Weeden Island Incised are all represented in the photograph. The presence of the heavy rim reinforcement which ex- tends down over the vessel walls, as seen in the vessel in the upper left corner of the Holmes illustration, is reminiscent of Oklawaha WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—WILLEY 257 Plain, a Weeden Island I horizon marker from north-central and northeast Florida. This similar treatment is also noted on the in- cised vase or globed jar, lower row, third from right (Holmes illus- tration). These features and the Late Swift Creek specimens place the collection as Weeden Island I. GADSDEN COUNTY Aspalaga (Gd-1).—The Aspalaga site is on high ground back from the river landing of that name in the western corner of Gadsden County (map 16). It was excavated by Moore (1903, pp. 481-488) but not visited by the 1940 survey. The site consisted of a shell midden and three sand mounds. Two of these mounds, both low and inconspicuous, were found, according to Moore, to be domiciliary in character. That is, no burials or special caches of pottery were found within them. The third mound, which is the basis of Moore’s report, was 6 to g feet in height and about go feet in diameter. In outline, the mound was slightly oblong rather than round. Fifty-four burials were found in the mound, and these were single skull, bundle, and flexed types. Charcoal was in association with several. Most pottery was in a mass cache, but some specimens were with individual burials. Nonceramic artifacts were also with and apart from specific burials. Lance points, projectiles, celts, hammer- stones, sheet-mica fragments, a mica spear-point form, hematite ore, a truncated stone disk, shell cups, shell tools, and shell beads make up this group. Pottery is listed as plain, incised, red painted, and complicated stamped. Tetrapodal supports and notched rims are other common features which are given special mention. The illustrated pottery (Moore, 1903) is classified below: Weeden Island Complex Weeden Island Plain....... Figs. 146, 150, and probably 152 and 153 Santa Rosa-Swift Creek Complex Switt Creek Complicated Stamped: 5.044....-c6eecceeee ce Fig. 151 (Early Variety.) Geystalw River. Ineised\az.d..:. o-crs'-ecurea sles aes Figs. 147, 148 The additional panel of sherds shown in Moore, 1903, figure 154, consists of four Swift Creek Complicated Stamped (top, and left side), one of which (lower, left) is definitely of the Early Variety ; one St. Andrews Complicated Stamped (bottom, center) ; one Weeden Island sherd (bottom, right) ; and one sherd of Fort Walton Incised (center, right). 258 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 113 The R. S. Peabody collections from Aspalaga show a single vessel which is four-lobed and of the type Basin Bayou Incised (No. 39131). The ceramic evidence from this site points to a Santa Rosa-Swift Creek dating overlapping into Weeden Island I. Other sites in Gadsden County—Moore (1903, pp. 491-492) also investigated a group of mounds near Chattahoochee Landing on the Apalachicola River. There were seven mounds in all. Some of them appear to have been substructure or temple-type mounds. Others, which were low, rounded mounds, were excavated, and as nothing was found in them they were considered as domiciliary. HOUSTON COUNTY, ALA. McLaney place (Hn-1).—This site is located in the extreme west- ern tip of Houston County, Ala., on the Choctawhatchee River. It was visited and excavated by Moore during his 1918 survey (Moore, 1918, pp. 528-530) but was not relocated in 1940. The site consists of a burial mound on a bluff above the river. The mound measured about 5 feet in height and over 30 feet in diameter. Five burials were discovered but are not described as to form. They lay upon or near mound base and apart from a mortuary cache of pottery. A chipped-stone celt or ax was the only artifact other than pottery. Only plain and red painted wares are mentioned. A gourd-shaped bowl of the type Weeden Island Zoned Red is illustrated (Moore, 1918, pl. SOM ee Ie) The mound is placed as Weeden Island. Fullmore’s Upper Landing (Hn-2).—This mound is in the north- eastern corner of Houston County, Ala., and was excavated during Moore’s 1907 excursion up the Chattahoochee and Flint Rivers. The site was not visited in 1940. Moore (1907, pp. 438-444) describes the mound as about 3 feet high and as located on a slope. A few burials were in the mound, but these were too decomposed to deter- mine burial form. Apparently masses of stones had been used for burial coverings. Plain, incised, and red painted pottery are the only types listed. Vessels were “killed” and had been placed in a mortuary deposit in the mound. Illustrated (Moore, 1907) pottery is classified below: Weeden Island Complex Wreedenslslandilincisedtracai.cititiineiele clatter Figs. 18, 20, 22, 24 Miscellaneous Unclassified... asda weidemeiinaes 6 auica Nee ee ee Fig. 17 (This specimen is related to the Weeden Island Series.) This site is of the Weeden Island Period. WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—WILLEY 259 DECATUR COUNTY, GA. Hare’s Landing (Dr-1).—This mound is located a mile or so inland from the Chattahoochee River. It is described by Moore (1907, pp. 429-437) as being 5 feet high and 48 feet in diameter. The site was not visited in the 1940 survey. Forty-three bundle and flexed burials were found in the mound. Both masses of phosphate rock and char- coal were found near many of the skeletons. Possibly the rocks had been used to construct a crude sort of cover for the burials. Pottery was in a major mortuary deposit and also in caches distributed through the mound. Other artifacts were found at random in the sandy-clay mound fill. Celts, hammerstones, lump galena, sheet-mica fragments, and conch-shell cups are all listed. Moore mentions plain, incised, check stamped, complicated stamped, and red painted pottery. Vessels were either perforated or were made with holes as special mortuary ware. Moore (1907, pp. 429 ff.) illus- trates the following pottery which is classified below: Weeden Island Complex Weeden Island Series: WVieedenmislandeincisedsseisacprecioaeceiiantee cna Figs. 3, 5,7 Wieedens Islands Punctatedisr.,..ciz;cv ee 4 Sp Andrews Complicated: Stampeders -ielesteleteiettieiere I (Some of these complicated stamped sherds may belong to the Late Variety and as such are more properly grouped with Weeden Island than with the Santa Rosa-Swift Creek Complex.) [Dbebatie hake Sd With oeaneee Se OD OME RIOD COCICAOHION AD OO COODRCORE 6 Deptford Complex Wepttorde Wineanes@heck |Stamped\sy.jje-)-1iice-eieiee leit steieieiavcie) «re 5 Destiord Bold) Checks Stamiped wajscialars Seeie tele 's2 blew wind olew.ce 17 IEP EOL: SIMMS OLAMIPe ss. gies ao weave len shale « «onl A eeialeleic/orw)s'e 10 270 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I13 Miscellaneous Plain Reds soe os cya Saas ds eae ares rsie See eyo ee eret ene 4 SmoothwiPlainticrs sosvsicie ce sietatemreesinicie cle-wiarevotelerevomver eee eee 28 Residual J Platt ri aso.s yee cersetin a cidio.w sare ucle fesierchistose ete 253 Indeterminate ‘Stamped 2.7. oc ccs cals cis ocielcrclelonie sreretereletenite 27 Unelassified! ccxccieieicars sis cis totter teow « 0 aie sie 6 cle oe rele eie ee ene 10 Total sherds.... 734 Deptford, Santa Rosa-Swift Creek, and Weeden Island I and II Periods are all represented at this site, with Weeden Island pre- dominating. The Tucker burial mound is described by Moore (1902) as being 80 by 86 feet in basal diameters and about 9 feet high. The slope of the east side of the mound was more gentle than the other sides and appears to have been a ramp approach or graded way to the top of the mound. Evidences of the borrow pits could be seen nearby. Burial forms encountered were bundle, single skull, and primary flexed. Most of the 79 interments were grouped together on one side of the mound. Occasional flexed burials were found in pits in mound base. Two instances of oyster shells piled over burials and one of charcoal accompanying a burial were noted. Pottery was found largely in a mass cache deposit and also in smaller caches. Nonceramic artifacts were sometimes with burials, sometimes not. These included a flint scraper, polished celts, galena beads, sheet mica, copper fragments, shell cups, shell beads, and shell tools. A few skulls showed frontooccipital deformation. Tucker mound pottery was made with mortuary “kill” holes and also, in some cases, had been perforated after firing. Plain, incised, complicated stamped, check stamped, red painted, and pinched wares are all mentioned. The illustrated (Moore, 1902) pottery is classified as below: Weeden Island Complex Weeden Island Series: Weeden Island Incised............ Figs. 215, 221, 225, 227 (?) Weeden “sland 9Plaines ccc «eset: Figs. 214, 216, 217, 220, 226 Tucker “Ridge-pinched °.../506%2 tices cs ae sconce eememe Fig. 219 Switt: ‘Creek Complicated” Stampedtja. laine aed cram terse cate ctr Voie ooo eime eal as eo oe I Weeden Island Complex Weeden Island Series: Wreeden" Island (Umcised\. oma gaa ao cco s ce dece wean esses 3 Weeden. stand Platina. tcvciats clears jer sisco faite aise eeie wlelsida 9 Warrapelle vPaneratedl ers idatro sin cis ne Minewiciee esicmetntes Roeeee 6 Rete lnslncs Sed ® scr Seaeuct crates tejmroie seis eta wienarctavanrertiaeee ale marc erente 2 Dvceixtslicn “Ohecle” Stamapends as wants sc sare ae cwieres cate Tae senna 13 Switt Greek ‘Complicated Stanipeds.sicc 22255. cies cc awe veces 8 Wriesteb lords Ord-t Are o.sjc/0's atevatays a avepeiai@eie tater evaierarc cues ese I Deptford Complex DI EAtIOLd SUNIL S EAIIPEC 5 o.0 <, svets-c crs cogmnreruielele ge stereo mm sleeve I Miscellaneous S MIO Obey antral ce tysreys teva e: sytareh ctersyororerovePe ekavevener acracarstararaicretavetoierei sisters 4 SLUR ROC sie fete, yess ciee)»niclala oid yw akocnl ore cette ele ween oasis I ROSA iii chraks ctocve & a vie ale sini 0 a eT CEE eis ORS Wleme o 79 MCAS SIM eee Bee poke ore Navas So sxcrerel oye ToT OR ERO Cressi ee eioheic toss 6 Total sherds.... 135 This collection is Weeden Island I and II. 280 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I13 Following the railroad one-quarter of a mile west of the first area there is a second refuse area and mound lying on the west side of the modern cemetery. The mound is a flat-topped pyramid, 2.50 meters in height, approximately 15 meters square at the summit, and 30 meters square at the base. Its measurements correspond roughly with those given by Moore (1902, p. 228) for Pierce mound C. The north side of the mound has been sliced away by the railroad cut, exposing a sand and midden fill covered over with a shell mantle (pl. 11, bottom). A shell-scattered village area of 3 or 4 acres lies to the south of the mound, and beyond this, on the south, as well as the north and west, are dense swamplands. The collection from this section, both the mound top and surround- ing midden, is classified below: Fort Walton Complex Fort? Walton Incised 27 .cdkic. co ae cisce me setoee cteie: Seer eee II Pensacola Incisedy syjess.sac: noe es eeisiee a nevera od ceite onto eee Fe Safety: Elarbors Incisedtrsie sone terse see c recta pias eee I Weeden Island Complex Weeden Island Series: Weeden Tsland Plains dc nyye selec Aerie cco cee eee I Gartabelle: incised: 2.12, 5toni Aaae sceiauacetes koe eee I Wakulla Check Stamped... ocsc00 setmn es one dees occas 15 Miscellaneous Smooth “Plaine iy ssjeso3 Share aisiee yo os Seele odes CERT Ee 2 Residualy Plain nyse iisccseseasue eros ousteh eraser le eS eis ete Ore oe eee 52 Wriclassihied| o1.< yo sicrasevsscucigcs sie ete ed eles Ot eee eee 2 Total sherds.... 92 This collection unit is divided between the Weeden Island II and Fort Walton Periods. On a small rise of ground just west of the second area there is another mound *! and shell-refuse area. The distance betwen the two locations is some 20 meters, and the two are separated by a sluggish stream and swamp-filled ravine. The mound is small, steep, 3 meters high, and is situated very near the railroad cut. The shape of the mound is somewhat indeterminate, but on the south side there is a sloping projection which may be a ramp. Sherds and shells are found on a patch of high ground extending some 75 meters south of the 51 It is difficult to identify this mound with those listed by Moore at the Pierce site. The most likely is mound B (Moore, 1902, p. 228), although this mound was considerably higher than 3 meters (approximately Io feet). Possibly sub- sequent digging and erosion have reduced its size. WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—-WILLEY 281 mound. A collection of sherds from this village area, as well as the mound summit, is given below: Fort Walton Complex OREM VValtonLNCISEC aac esac cs ties seit e a aclosie ee ecie oisiers eed atere sie rs I Weeden Island Complex NVickaiiam @heck#Stamped- 1s joe scarce cee ciens ae oe see sense 13 Sywitte Greeks Complicated = Starmpedereccriccie sete) eisiteleetrers oretaterr= 3 Wives: llerscky (GormSibid atlas cnaccdcocoguaunouacochsonucnuace I Miscellaneous IRGSiG ibe ll OEN I eee eae as Os OOOH ABO OI io coe OCOD IAD Ca 16 MO naclassitied: Moise wiciere: Sa Steere erator eee ee: aeons revatepahcrayalele'< xe eve 4 Total sherds... . 38 This third collection is largely Weeden Island Period. Considered as a whole, the three Pierce surface collections, which represent a sampling from only a relatively small area of the total shell refuse in the vicinity, indicate occupation for the Weeden Island I and II and the Fort Walton Periods. Earlier Santa Rosa-Swift Creek occupation is not well represented. Of the Pierce mounds excavated by Moore (1902, pp. 217-228), only the first, Pierce mound A, yielded results of any consequence. This mound must have been rectanguloid and flat-topped, as the measurements given are basal diameters, 96 by 76 feet, and summit diameters of the same axes, 40 by 34 feet. Height of the mound was 8 feet. No mention was made of a ramp approach; apparently none existed. The mound was composed of yellow sand except near the base and in occasional irregular lenses in the body; here oyster shells were common. On or near mound base were many evidences of fire. Of the 99 burials the greater part were primary and flexed al- though there were a few examples of extended burials and of secondary types.°? The latter were single skulls, miscellaneous bone piles, and at least one mass burial of a secondary nature. Evidences of fires having been placed over or near burials were common although no true cremations were found. Pottery was found in small caches in various parts of the mound, usually near the base where most of the burials had also been placed. Some instances of pottery with specific burials were noted. Other arti- facts were both at random in the body of the mound and with indi- vidual burials. The nonceramic artifacts included chipped projectiles and chisels, stone celts, stone plummets, pearls (fresh-water), a cop- per tube, copper ear ornaments with hammered-silver plating, shell 52 Moore notes that this number represented a great many more individuals. 20 282 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I13 beads, shell drinking cups, gougelike tools of shell columellae, a large bone gorget (identified as bison bone), and canine teeth. Pottery vessels were “killed” by perforation. Such pottery traits or types as plain, incised, complicated stamped, and red painted wares are mentioned or illustrated. Miniature vessels, tetrapodal supports, notched or scalloped rims, and cord-marking were also noted by Moore. A pottery pipe of the monitor type was recovered. Illustrated pottery is classified below: Santa Rosa-Swift Creek Complex Crystal River Series: Crystal. River. Zoned «Red: 2543 ied See Figs. 158, 163 Pierce Zoomed Red oeisix. 24 ws aioe aloe ceva elaine eee Fig. 155 Santa, Rosa) Stamped!.cjocuk octuscd ene ss sera o.ceee eee Fig. 162 Swiit-Creek (Complicated Stamped-.. i... . 42.05 see eee Fig. 165 Miscellaneous RONCIASSINE Sco ite, cele ee ee oe eet ee ee eee Figs. 156, 160, 164 In the R. S. Peabody Foundation collections there are two specimens from the Pierce mound A. One is Pierce Zoned Red (No. 39301) ; the other is a monitor clay pipe (No. 39182). This last may, or may not, be the one to which Moore refers. The mound is dated as Santa Rosa-Swift Creek. A number of interesting ceramic elements occur, particularly the types of the Crystal River Series and the unusual plain-ware forms such as a multiple-orifice vessel and a “grubworm” effigy. Of the other mounds of the Pierce group, there is little to aid in cultural identification. Mound C, probably the mound of the second 1940 collection, was trenched and found to be constructed over a basal shell layer. Flexed burials and sherds of the check and compli- cated stamped types were encountered. Mound D proved, according to Moore, to be domiciliary in that it was composed chiefly of refuse. Mound E he also defined as domiciliary, but in this last case the mound was of clean sand rather than refuse. Jackson mound (Fr-15).—Moore (1902, pp. 229-234) found this mound 2.5 miles west-northwest of Apalachicola (map 16). He de- scribes it as 9 feet high with a basal diameter of 72 by 66 feet. No other comments are made as to mound shape. The data on the 26 burials are none too sound, as skeletal material was badly decayed ; however, one cremation was definitely recorded. Other bone remains were centrally located in the mound. Bundle and single skull types of secondary burials were met with, but other identifications were uncertain. Pottery was, for the most part, concentrated on one edge of the mound in a special cache. Other artifacts, such as pipes, were WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—-WILLEY 283 found with burials. Projectile points, hematite ore, celts, hones, smoothing stones, hammerstones, galena, stone beads, plummets or pendants of stone, a plummet of quartz crystal, and a sheet-copper fragment were also present. The pottery pipes from the Jackson mound come with a burial which was only 18 inches from the surface. This is a notable pro- venience, as all other burials were found on or near mound base. Also, the bones of this particular burial are reported to have been in a much better state of preservation than the others in the mound. Because of these facts Moore was of the opinion that the burial and the pipes might be later than the mound proper and its contents (see also the Huckleberry Landing site, Fr-12). The pipes in ques- tion are of the elbow variety (Moore, 1902, figs. 166, 167) made for the insertion of a stem or mouthpiece of wood or reed. One has a slightly flaring bowl and crimped or scalloped rim suggestive of the vessel rim treatment of the Santa Rosa-Swift Creek pottery type Franklin Plain. The edge of the bowl is also ticked with fine incised notches in a manner similar to a stone pipe from Crystal River, Ci-1 (Moore, 1903, fig. 38). Another pipe, found in the marginal fill of the Jackson mound, was made of steatite and is described as being of the “common rectangular block pattern” (Moore, 1902, p. 232). Pottery vessels from the Jackson Mound were “killed” by perfora- tion. Plain, incised, and complicated stamped are the only major ware categories mentioned. Illustrated (Moore, 1902) material is classi- fied below: Weeden Island Complex Switt Creek ‘Complicated Stamped? <2. ..0..ccsncecenece Fig. 168 (Late Variety.) Santa Rosa-Swift Creek Complex Aliheator Bayo ‘Stamped..c..\.tc- seca axe eee oe ais cw cite ls Fig. 170 Switt Creeks Complicated’ Stampeders. ecicicle ic sliela slelcielels Fig. 169 (Early Variety.) A single vessel from the Jackson Mound in the R. S. Peabody Foun- dation collections is a plain piece, possibly Weeden Island Plain (No. 39108). The Alligator Stamped type decoration on the multiple-orifice vessel at this site is interesting and further substantiates the dating of Pierce mound A where a multiple-orifice vessel was found in a context believed to be Santa Rosa-Swift Creek. The Jackson mound may be verging toward Weeden Island I, as suggested by the very definite Late Variety Swift Creek bowl; however, Santa Rosa-Swift Creek elements are dominant. 284 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I13 Other sites in Franklin County.—In the 1940 survey we visited a large shell midden on East Point peninsula, across the mouth of the bay from the city of Apalachicola. There are two middens at this site, both semilunar in form and composed of oyster shells (pl. 2, top). They are about 25 meters from the water’s edge on a flat, sandy point. Both are as much as 3 meters in height. Only a single plain, sand- tempered sherd was found at the site after a diligent search. There is little doubt that these shell piles are artificial and of human origin. Their size, their undisturbed condition (as of 1940), and the scarcity of artifactual material combine to make the site extremely attractive as a place for future work pointed toward the problem of the earlier horizons. Additional sites excavated by Moore in the Apalachicola vicinity but not included in the present analyses for lack of data are the Cool Spring mound, the mound near Apalachicola, the Cemetery mound, and the Singer mound (Moore, 1902, pp. 216-217 and p. 229). The mound near Apalachicola seems to have been a sterile sand structure and was considered domiciliary, but the others were burial mounds, probably of the Weeden Island or Santa Rosa-Swift Creek Periods. A possible exception to this last is the Cool Spring mound which might have been a Fort Walton Period temple mound used as a burial place. Farther east, near Carrabelle, Moore (1902, p. 257) mentions a mound on the west bank of the Carrabelle River about 1.5 miles inland. This was a burial mound but contained little. In the Florida State Museum at Gainesville there is a collection taken from a mound near Apalachicola. The collection was presented to the museum by H. L. Grady. It consists of Fort Walton Period vessels, an elbow pipe of the Georgia Lamar Period type, shell pins, a Busycon, shell celts, and stone celts. Possibly this mound was one of the above which were excavated by Moore (1902). LEON COUNTY Lake Jackson (Le-1).—See excavation account of Lake Jackson, Leon County (Le-1), under the section, “Excavations on the North- west Coast: 1940.” Lake Lafayette (Le-2).—Lake Lafayette lies less than 10 miles southeast of Lake Jackson.®? About I mile northwest of the northwest tip of Lake Lafayette there is a flat-topped pyramidal mound of clay surrounded by fields which bear evidence of having been an old village site (map 16). The mound is oriented north-northeast by south- 58 See Boyd, 1930, p. 272, for mention of this site. WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—-WILLEY 285 southwest and is 36 meters on a side at the base. Estimated height is 4.50 meters. The only visible excavation is a hole made in the top. In 1940 a sherd collection was gathered from the surrounding village area. This collection is classified below: Fort Walton Complex Fort Walton Series: One my Vialton Uneisedepinc ceutetescre tie st sasterettal © aemhcne etal 4 Aeaice Jackcson Me lanihs Jem idace tance scleys ecibais cernplge hepmercte 6 Rensacolals ELlaiitay yssversrorsctesestanstere rst cei ate es octane al ois pate tarenole I BEDE ReGen SCM DEM idea ners eles acted atat ase ein aided ateuctalale 2 iene Complicatem’ Stamped, oid .tre ite t cit eos sis acne crn ae « 7 Weeden Island Complex Garranellembrmnctatediusepreccwrcesrrse cere erce se seictoveinic cit ists se aie I NWiaikanllam @heclkos Stampeders ci raise oie cre or terairercreie cistar cls eels a's I Miscellaneous RRESicl tiellembolaitian ys crac te crckstarecrere eiaveters ens ieiaicrece ater eherete tere, oie ic 6 sieeve 128 (Most of these were body fragments from Fort Walton Series vessels. ) lhaglencineareehie Susnalacl neoqneoascbossoeaoooonuecHteapecoudnc 9 Mp eratine lassie! scjahis x hat aineisce cashes acelaysgsrelatotayabe spol) mcala tarot of 4 Total sherds.... 163 This village site and mound must date from the Fort Walton Period. Rollins site (Le-3).—At a centrally located point on the eastern shore of Lake Jackson there is a small peninsula that protrudes very noticeably into the lake.°* The Rollins mound and site are located on this peninsula (map 16). The mound is situated on the top of the high bluffs that overlook the lake valley. It is a flat-topped pyramidal structure which has no longer retained its full rectangular form. The base measures 27 meters north-south and 25 meters east-west. Esti- mated height is approximately 2 meters. There is a large, irregular excavation in the top. In the adjoining cultivated fields 5 Fort Walton Incised and 13 plain (probable Fort Walton Series) sherds were found in the red loamy soil. The site dates from the Fort Walton Period. San Luis Mission site (Le-4)—The old San Luis Spanish Mis- sion site is located at the James Messer Place, 24 miles west of Tallahassee ** (map 16). Along the crest of an eroded hill slope, in back of the Messer house, a number of aboriginal sherds and some Spanish olive-jar fragments were recovered in the small washes and 54 See Boyd, 1930, p. 272, for mention of this site. 55 Boyd, 1930, pp. 264 ff. 286 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I13 ravines. More pottery fragments were also picked up in the roadway to the west of the house. Three of the aboriginal sherds were of a faint complicated stamped type, suggestive of the Lamar style; the remainder were all plain. In addition to the 1940 survey material there are a few sherds in the United States National Museum (No. 148125) from this same site. One of these is a folded and crimped rim of the Lake Jackson or Lamar type; one is of the type Aucilla Incised; and three are unclassified plain. The San Luis site is well documented in the historic sources. Ex- cavations in the site, by Dr. M. F. Boyd, have revealed Spanish Colonial features. The limited aboriginal ceramic evidence places the site as post-Fort Walton. This late contact period in northwest Florida has been referred to as the Leon-Jefferson Complex (H. G. Smith, 1948). Lake Iamonia (Le-5).—The exact location of this site is not known, so that its indication on map 16 is given only as the lake. The collection, which is in the Peabody Museum, Harvard Uni- versity (Nos. 12-11/81200-81202), is apparently from a site unit, however. It is classified and described below: Fort Walion Complex Fort Walton Series: Bort. Walton’ Encised':.. ccs sc: oe eles ss crave a sle.ce enero 17 Lake sjJackson Plain:.:..cicc. itksiele veel sees «eee 12 “Chunky: sherds @eakejacksons 2] ain) eerie rceereeteeie rere I Miscellaneous Indeterminate ‘Stamped ce. 2. dcien tise vie cic w cin cioeel cee eee I Baked-clay ball with simple stamped decoration, diameter 4cm.. I Total sherds and artifacts.... 32 The collection is clearly Fort Walton Period. WAKULLA COUNTY Marsh Island (Wa-1)—Marsh Island, on Ocklockonee Bay, was not visited on the 1940 survey (map 16). Moore (1902, pp. 274-281) describes the mound as oblong with rounded corners, being 98 by 68 feet at the base with a height of 7 feet. It is not clear as to whether or not the top was a flattened platform. About one-half of the burials encountered were of the bundle type, another one-quarter were single skulls, and a final quarter were primary flexed burials. These burials were found in two groups within the mound. Seven burials, some containing more than one individual, came from the upper 2 or 3 feet, as near as can be WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—-WILLEY 287 judged. The remainder were found on, near, or below mound base. This distinction is of real importance as it represents a cul- tural stratification. By position in the mound, and by accompany- ing artifacts, Moore demonstrates the seven upper burials to be later and intrusive. In view of this distinction, the two com- ponents from Marsh Island are discussed separately. The Marsh Island Intrusive graves (Moore, 1902, pp. 275- 277, Burial Nos. 61, 70, 85, 92, 104, 105, and g1) contained, in five cases, more than one individual. Interment was secondary, of either massed skulls and long bones or simply skulls. Another burial had only a single skull, and the remaining burial was an urn interment of an infant. One of the group secondary burials showed partial calcination of the bones, implying either actual cremation or partial burning of the bone in situ. Shell pins, shell beads, and, in the case of the urn burial, pottery vessels were the only completely native artifacts found with the upper level or Intrusive burials. European artifacts, or objects fashioned by the Indians from European goods, were more numerous. These in- cluded scissors, iron tools or weapons, sleigh bells, tubular brass beads, glass beads, and brass bracelets. The two pottery vessels with the urn burial, one as a container and one as a cover, were not “killed” by perforation. These vessels (Moore, 1go02, fig. 241) are of the type Marsh Island Incised. Loose in the mound was a single discoidal stone with two con- cave surfaces. The cultural associations of this object in the mound are not known ; however, as the discoidal stone is so commonly found on the Middle Mississippian horizon throughout the south- eastern states it is likely that it is a part of the Intrusive component at Marsh Island. Most of the Intrusive grave skulls showed frontooccipital deformation. There is no doubt that the Intrusive component at this site is late, extending well into the historic contact period. The Marsh Island Incised type differs from typical Fort Walton Period types, but I am inclined to believe it about contemporaneous ; thus, secondary burials and the urn burial equate more closely with the Fort Walton Period than they do with the somewhat later Leon- Jefferson Period of the missionized Indians. The lower component of the Marsh Island mound, presumably the component to be identified with the builders of the mound, is characterized by masses of shell placed over some of the approxi- mately 100 skeletons, by undeformed crania, by the large mor- 288 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. II3 tuary cache of “killed” pottery buried on one side of the mound, by scattered pottery caches, and by nonceramic artifacts both with and apart from individual burials. Celts, smoothing stones, hones, plummets, hematite and plumbago ore, and sheet-mica fragments were present. Plain, incised, complicated stamped, and red painted pottery categories are listed. Illustrated (Moore, 1902) ware is classified below: Weeden Island Complex Weeden Island Series: Weeden” Island! Mncisedeie. tee enitme cle: nee Fig. 242 WieedeniMsland® Plante tess eb ee enone Figs. 243, 244, 246 Switt Greek Complicated Stamped. : 2isas ees -seicieeeree Fig. 248 Miscellaneous Wielassified: Bisse. Ee tos oe ae eoae oO eee Fig. 247 Three vessels in the R. S. Peabody Foundation undoubtedly be- long with the lower component. These are a Swift Creek Com- plicated Stamped, Late Variety (No. 39922) and two Weeden Island Plain (Nos. 39143, 39170). The lower mound component is Weeden Island I Period. Surf (Wa-2).—Surf is a small settlement on the north shore of Ocklockonee Bay, 2 miles west of the Ocklockonee Bay bridge (map 16). There is a low shell midden in the midst of the set- tlement at a distance of about 25 meters from the bay. In all, the shell covers an area approximately 100 meters square. The refuse is a mixture of black midden and clam shells which has been thoroughly dug over and disturbed. A small collection was made here in 1940: Weeden Island Complex Weeden Island Plaine) ce. .tccaaies «siisdat e6idans eee eee 3 Wakulla ‘Check Stamped’. < o:o2.<:50 ce nse eee ane eee 20 Miscellaneous Residual {Plainy jetties s se es ss eee es eee Oe eee 27 Unelassified:, 1.54 c crete ter ene 10 Other *unclassitied: sinet bee Figs. 306, 307, 308 Miscellaneous Wrelassitied) sslscass. 2x Se:5, sieve ofeiare-s lovee aisvovelelevorieucuh ate ecole ene eee Fig. 312 (Probably Weeden Island Series but decoration is not typical.) There is a single specimen from this mound in the R. S. Peabody Foundation. It is Weeden Island Plain (No. 39273). The mound is obviously Weeden Island and, probably, Weeden Island II. Refuge headquarters (Wa-13).—This midden takes its name from the present headquarters of the National Wildlife Migratory Bird Refuge which is located on the site (map 16). The general location is some 3 miles southeast of the mouth of the St. Marks River and WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—WILLEY 297 in the same section where Moore excavated the burial mound in 1902 (site Wa-12). The midden is on a hillock which is the western end of a series of sand ridges rising about 5 meters above the surrounding swampy regions. All the ridges show trifling evidences of occupation, but the principal concentration is within a 5-acre area around the headquarters buildings. This concentration is marked by dark or- ganic stain, broken conch shells, and sherds. Depth of the deposit is nowhere disclosed by excavations ; however, it is probably less than I meter deep. The 1940 collection is tabulated below: Weeden Island Complex Weeden Island Series: Wieedenelslandiinciseds:.csgecitecmeteaseis saves Gicisve soko mis 10 WVecdenels) anid Plantvtsrchote ceteris erent sues ccern ciatelonesaysis:siere Il Carrabelle. Piunctatedss,ccreieicv-erioco oe en acter eicte sisidie care cue I utc MOhMeC AS iain peti. cm ants ve ewer coe a eee ce oaleed eae t ae 80 Swit Creek Complicated Stampedye.. 2.02. 400 6600 sheckon ed I Miscellaneous TERIA 12G alg tness. 6 Mice a heat el cee er I Minin Fen Oo ten alles Cer Pe er ee I Sto Ot ila byl Att Wararriar,. cteyatet ora. setsastevelore wisicleuets ste ytoteeaevs at werethelale.s I Gs Telttetl tates Nees eves oper Caran tap aye sate o's ced csoies susie, wrecon ake cuelsols sietions 100 Indeterminate “Stampede srs ccs cs croieiores ers leracaieloisicvs sietetev ore eereve 5 Total sherds.... 210 The site falls into the Weeden Island II Period. Refuge tower (Wa-14).—This midden is named for the lookout tower on the same National Wildlife Refuge range as the previous site. It is situated about one-half mile south-southwest of the head- quarters site (Wa-13) (map 16). The location is a sand ridge, cov- ered with black midden and soil, some 200 by I0o meters in extent. Depth of the midden could not be determined in 1940 but is probably under 1 meter. In addition to sherds there are numerous flint chips and cores on the ridge. This site was revisited in 1946 and 1947 by John W. Griffin, Hale G. Smith, and Dr. John M. Goggin and has been reported upon, briefly, by them. (See references given for site Wa-I5, below.) The 1940 sherd collection is classified below: Fort Walton Complex ices ll Cold MICISEM 5 (50's s:d Spevoclard arasaiw oa arse yaaa sta niles ova aes I Weeden Island Complex Riparian Ghecku Stamped «o-..2< Xs os sera os pele oe 10 Swatt (Creek: (Complicated Stamped’... sca. sles e see nee 6 Ruskin Dentate Stamped 5)... .0:c:s:s suis wane > see cee ee 9 Papys’ Bayou Punctated<: isc5. ss. om cate nes | oe Cer I Biscayne. Plait oi0.i5 2 shies aid ora enon ew nlle sce ratte eee ee 3 Glades Complex Glades Plain. 2 bs). dee ss silo cis ee oe cetientisee cern t Cees 2 Miscellaneous Residual’ “Platts. sc. sycs.ejcrsct 4.5 otsve sisvece eres wiert wines Sree ee 14 Unclassified ‘se. \.lcacelts cise tacos cllvae csiee renee eRe 3 Total sherds.... 55 On the basis of this collection the mound dates in the Weeden Island Period. WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—WILLEY 3II Palmetto Island (Lv-7).—This collection was made by Decatur Pittman in the 1880’s. It was classified in the Florida State Mu- seum at Gainesville (No. 542) by Dr. John M. Goggin in July 1947. The classification sheets were subsequently checked over and revised in consultation with Goggin. Safety Harbor Complex PMC ASRIUMCISCOE c[oitere iets nna eae sd Seb catieciaee area Moree I Weeden Island Complex Weeden Island Series: Nvecdenelsland Ineisedsrs: cnad teas -casaiccs s Salaun vee cee 143 Mreedenmislands Bunctated? sens -- ieee. store sto oc acta 62 Weeden Island >Zoned, Redixd tacind oso s oee oe ae e's bles 139 Wreedeny tsignd Plame cei. ccaccce ote Wowetee awoneeuce es 1,009 aera pelieiMiileisedim aes dete ton sce ces he oes Sabo Sera ons 384 CarrabellePunctatedy . ssen ence scemonte. oe Se aie thaeiote is 477 Mndianwbassmlincisedicrseys cist crners re eieione wracersteioeee cee ale 56 PeGHw INEISEKee nei ccc ad nelskeuee es mane aoe ek oueaane 104 ehicken elid se—piticliedabaerinw ercreteromiare oi orats ioravnre, sieves fe ole ae 25 Hare Hammock Surface-indented.............2.c.000 8 Weeden Island Punctated with incised serpent head..... I WY alcttlilaie GheclaeS tarp ed acre lecerct siete aroraiore steiavs Qos | 5 ' 6279 re gs ' P we, E K "Kk 2 if 5, e i Ug 4D ~TAMPA 1 ear, of I qe) LSBOROWGHA N 0 LLAS=\ ! WN) i, | 5 depile \ LONG KEY wi 7, 2 Cite MAM lo /0 20 bo oe BS Ww MILES Map 19.—Site map of Hernando, Pasco, Pinellas, and Hillsborough Counties. Sites indicated by numbers. HERNANDO COUNTY Bayport (He-1).—This is a mound about 1 mile north of the town of Bayport (Moore, 1903, pp. 415-424) (map 19). It was between 3 and 4 feet high and had a basal diameter of approximately 80 feet, although it was oblong rather than circular. About 40 burials were found, but these may represent a slightly higher number of skeletons. Bundles or bunched burials were the most common and 64 John K. Small (1927) located mounds and shell middens in the vicinity of Crystal River; others have also been reported from Black Point, Coney Creek, Chair Island, Boggy Bay, Salt River, and near the mouth of the Homosassa River. 326 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I13 were discovered in all parts of the mound; a few instances of crema- tion or partial cremation were also met with. Shell tools, pottery, hematite ore, and a celt were associated with burials. A large vessel fragment with a loop handle had been placed over one burial. Pro- jectile points, celts, and copper fragments were found loose in the mound fill. Pottery, besides that placed directly with burials, was in small caches. Most of it was broken or “killed” by perforation. The trait of prefired “killing,” or the making of holes in the vessels before they were baked, is also noted. Plain, incised, punctated, red painted, check stamped, and complicated stamped categories are listed. Moore’s (1903) illustrated pottery is classified below: Safety Harbor Complex Safety arbor incised. ..0-- cree Figs. 66 (3d row, 2d sherd), 71 Weeden Island Complex Weeden Island Series: & Weeden Island Incised....Figs. 66 (top, left ; top, 2d sherd), 74 Weeden Island Punctated....Figs. 66 (top, 3d sherd ; top, right; center, left; 2d row, 3d sherd; 2d row, right; 3d row, Ist sherd; bottom, left; bottom, 2d sherd), 67, 68, 69, 70 Wreaclsn ligknadl Ibi, csaabausouccbandosneoe Figs. 76, 77, 78, 79 Garfabelle Imeised’ sce asnse ce eeeees Fig. 66 (3d row, right) Kieith nicised vs cei sects stactern oe Olasiens connaee Fig. 66 (bottom, right) Swift Creek Complicated Stamped...... Fig. 66 (2d row, Ist sherd) Miscellaneous Unclassified’ incised: 224 55 5s ac ce le Sa cid seic oeiee eee eee Fig. 72 This collection is augmented by a piece in the R. S. Peabody Founda- tion. This is Pasco Check Stamped (No. 38946). The mound was obviously built in Weeden Island times and was probably used through both Weeden Island I and II Periods. The few Safety Harbor sherds, and the description of the loop-handled bowl as a burial cover indicate the mound’s reuse as a Safety Harbor burial place. Other sites in Hernando County.—Moore’s other two excavations in this county cannot be satisfactorily placed as to period. Indian Bend, near the Chassahowitzka River (Moore, 1903, pp. 414-415), was a burial mound with both secondary and primary burials. Check stamped pottery was found in the mound as was a sherd of the type St. Petersburg Incised (Moore, 1903, fig. 65). This type is usually found in Weeden Island Period contexts but seems to be a very late Weeden Island II type. The Indian Creek mound is 5 miles south 65 Some of these may be Papys Bayou Series sherds. WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—WILLEY 327 of Bayport. It was a burial mound (Moore, 1903, p. 424), but con- tained no artifacts. Recent important discoveries in Hernando County were brought to my attention during the month of August 1948 by Dr. A. J. Waring, Jr. Dr. Waring excavated a midden site on a marsh island at the mouth of the Chassahowitzka River. The excavation (a 5- by 10-foot test pit) revealed an upper stratum composed mainly of oyster shells and salt-water mollusks. This upper stratum contained potsherds, principally of the Weeden Island Period. At a depth of 2.5 feet Waring encountered a lower midden stratum of fresh-water gastropods, fine fish bone, and ash. This lower stratum was solidified into a hard, rocklike mass. Flint artifacts were found imbedded in the stratum, but there was no pottery. To quote Waring directly (letter of August 19, 1948): The flint forms consisted of scrapers as well as very interesting massive forms. These last are large crude hand axes, very much like the coup-de-poing form. They measure 10 to 12 inches in length. Other flint items are large hand chop- pers made of big, wedge-like flint flakes with edges battened round from use, and fragments of large ovate blades and knives. Projectile points were found, but these are not described. Waring also took a longitudinally grooved stone from the lower stratum. This is a form which occurs in the Glades area on a late (Glades III) horizon, but, as far as I am aware, it is unknown in the later periods of the Gulf Coast. From the description it would appear that this Chassahowitzka River site is the first prepottery station of the Gulf area to be reported. Waring continues: The finding of nothing but fresh-water molluscs (in the lower stratum) interested me considerably. The river is sufficiently salt at the site right now for oysters to grow, and river snails are about 3 miles inland. All this suggests a perceptible rise in ocean level since the time of the archaic site. The replacement of fresh-water shells by salt- or brackish-water species strikes a familiar note as this is the sequence along the St. Johns and Atlantic Coast where the Archaic prepottery cultures were also associated with fresh-water mollusks. This change in shell- fish has there been interpreted as a result of a prehistoric rise in ocean level at the close of the Archaic. Waring’s findings and analysis would appear to parallel and verify the east coast phenomena. However, even more recent excavations at the same site cast doubt upon the preceramic interpretation. R. P. Bullen, of the Florida Park 328 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I13 Service, in a letter dated May 23, 1949, writes concerning the same Chassahowitzka site: You will remember that Waring reported an upper stratum of oyster shells containing pottery and a lower deposit of consolidated snail shells. Apparently he found no sherds in this lower zone but we did. We also found, in a trench about 50-60 feet from his, the large, coarse, scraper- like blades closely associated with pottery. In fact pottery was found at sub- stantial depths below the concentration of these large tools. Here the lower zone consisted of black dirt and oyster shells but no snail shells (incidentally there were a few oysters in the snail shell breccia of the other trench). Obviously a thorough analysis of all materials and data from the Chassahowitzka River site, by both Waring and Bullen, is in order. Perhaps this will clarify the question of the possible preceramic layer, the position of the large flint blades or choppers, and the succession of — fresh-water and marine shells. For the present we can only point to this newly discovered site as important if somewhat confusing. PASCO COUNTY Wekiwachee River (Pa-1).—Moore locates this mound as 2 miles south-southeast of the mouth of the river (Moore, 1903, pp. 425-426) (map 19). The mound was of oblong shape, being 86 by 64 feet and about 4 feet in height. The original summit shape is unknown, as the top had been trampled flat by cattle. In the middle of the mound was a virtual layer of human bones made up of secondary burials. Numerous bundle burials were found at other places in the mound as were a few flexed burials. A total of 145 burials, including a pocket of calcined human bone, make up the total, and this probably repre- sents almost 200 persons. Shell drinking cups and stone celts were found with burials, and hammerstones, lance points, and a grooved- stone pendant were loose in the mound. A small deposit of sherds was at one side of the mound, and others were scattered about at random. A “killed” vessel accompanied the great mass burial in the center of the mound. Besides perforation “killing” of vessels, one example of prefired “killing” of a pot was recorded. Check stamped, complicated stamped, and punctated ware are specifically mentioned. An illustrated panel (Moore, 1903) is classified below: Weeden Island Complex Weeden Island Incised.............. Fig. 80 (top and bottom, left) Ruskine Winear aeunctated sayrrreeteseses ce eels Fig. 80 (bottom, right) Miscellaneous Unetassified ‘punctated™ s.c..cearces see cee eee Fig. 80 (top, right) WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—-WILLEY 329 A vessel of Dunns Creek or Biscayne Red is in the R. S. Peabody Foundation (No. 38933). The mound is Weeden Island, probably Weeden Island II. Pithlochascootie River (Pa-2).—S.T. Walker (1880a, pp. 393-394) located and excavated two mounds near the mouth of this river on its south bank (map 19). One of these was a large flat-topped temple mound, made of alternate layers of sand and shell, oblong in shape, and measuring 168 by 55 feet and 5 feet in height.°° Excavations in this mound revealed nothing. The second mound that Walker dug was 100 yards distant from the first. It was made of sand; an irregu- lar oval shape; and had either a ramp approach or a projecting wing on one side. Walker records its size as 175 by 50 feet at its maximum diameters and 7 feet high.*? Walker found burials, arrow points, pottery, and an iron spike in this mound. He makes a dubious claim of having found the skeletons arranged radially in the mound with heads toward a common center. Moore found 62 burials in the same mound, but one of his burial units contained the remains of 57 individuals while others were com- posed of more than one person. Perhaps 150 individuals had been buried in the mound. Secondary burials were more common than primary. The single skull, bundle, and mass bone pile were all repre- sented. There were also a few extended and flexed burials. With a big mass secondary burial were three deposits of calcined bones. As a rule the primary burials were found on mound base, while the secondary types of treatment were in the upper part of the mound. Pottery was found with individual burials and in small scattered caches. Nonceramic artifacts were both scattered through the mound or with specific burials. Lance and drill points, quartz crystals, stone pendants, celts, pebble hammers, shell tools, shell cups, shell pendants, and bone tools were in the mound. Pottery was “killed” by perfora- tion, and incised, check stamped, plain, complicated stamped, and painted categories are listed. The illustrated material (Moore, 1903, pp. 426-433) is classified below: Englewood Complex PAgiewood of Sarasota Incised. (la... . I BiSCayitie | lagi 5 sie :c-w aie weisie aha oe wiele ays, v1 a8e we eiepalere oelede iinet I Glades Complex Bie bea hs ween oe sha-gncySas fang avotaiare toro ioke tase aloha reais sim akaretene oreenryers 23 Miscellaneous Simoeun IB a Gun doeauvoopocepcoadeusay cp dongeunabadqatoe I Wrclassined plain (may be Pinellas)- 2: ~~... ..0css--c elas = 2 lWrclassiied «cord-markedimcrmcisieitelielsise oticisilelsVelleletl-ir)l-r=) 1-1-1 I Total sherds.... 44 The site indicates Weeden Island II, Englewood, and Safety Har- bor Period occupations. The presence of Glades Plain is consistent with the excavated sites in this region (see “Excavations on the West Coast : 1923-1936,” this report) in that this type appears coevally with the later Gulf Coast periods. Boca Ciega Island (Pi-6).—This collection is from one of the islands forming Boca Ciega Bay (map 19). It was collected by Mrs. Pearl Cole, of Washington, D. C., and presented to the United States National Museum (No. 36066). Presumably it is from a midden site. It is classified below: Safety Harbor Complex ARENA EL AL DOE LTICISCRs 2's ojo) ae %ein der’ sichn asecis sata, opeheleidib,e 8s e's) ere I enisacolay te leantaiaesstetorscscreiste eerie eee isso Reo atte oteue tera foraiets 2 Weeden Island Complex Meer + RAM PE=PINICHEM v.ctdhcaneretacriene « ROR ate eae eee as I Wralctullae Gheclk>Stampedtacwrernicciaae eterno aerate es a atiesierusts 8 Biscayne (or St. Johns) Series: Biscaynes @hecka Stampedheaereiaceiiseceiaceimceteniebrectcts 18 Biscayne: Cord=tmiatkeds,incesee tere e eerie ba /cwied Mem oaininlern ales, s I Glades Complex Gladesmiblainitentetevcs ecw t aaa eschews caterer mom etee tietsiereree 8 Bellen Glade bel atin erates «slave tals steicterstisrvapermrn sheichone sie aesieke eres teres I Total sherds.... 40 The collection dates from Weeden Island II into Safety Harbor times. Bayview (Pi-7).—There is some confusion about the location of this mound (map 19). Walker (1880a, p. 410) places it as about 1 mile north of Bayview Post Office, on the south bank of Alligator Creek. According to his map (1880a, p. 412) this would place it somewhere near Safety Harbor, although Walker (same map) seems 334 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I13 to have Safety Harbor too far north. From the description of the Bayview mound it does not seem likely that it is the Safety Harbor or Phillippi Hammock mound. He describes it as circular, 46 feet in diameter and less than 3 feet high. He found human burials in three distinct strata. Nothing is said about burial form other than it “‘differs slightly” from what he had previously indicated as a bundle-type burial. In the bottom layer no artifacts were found with burials. In the middle and top layers he found glass beads, brass and copper ornaments, scissors, looking- glass ornaments, crockery, and other trinkets of European pro- venience. The collection of pottery from the site, which is now in the United States National Museum (Nos. 35315-35326), numbers only four aboriginal sherds. These all belong to the Safety Harbor Complex. Seven Oaks (Pi-S).—This is a sand burial mound located about one-half mile west of Seven Oaks (map 19). There is a collection of material from the mound in the Florida State Museum at Gaines- ville which has been classified by Dr. J. M. Goggin. Safety Harbor Complex Safety Harbor Series: Satety Harbor’ Incised.. .: .¢s022.4 62 sas os 0 oe ee 39 Pinellas@Ineisedy (2 A} eAces ote videtotercieea cine Me Oe DEE 26 Pinellas Plaine. cc4iccid cide ti eacbiocis ce a Re 19 eon: ‘Check -Stampedésjs ainsi. sa cio ves Sona delenit Eee 31 Wamar-likecomplicatedmstamped sc -macmirs cee aise eee 12 (May be Jefferson ware.) Weeden Island Complex Weeden Island Incised. (2) <2 ...0% »0 5» dec eamenieee ea see I Wakulla Check ‘Stampediadiincacc cicecreisiscio ce reese cree eee 8 Thomas ‘Simple Stamped. .i\6sccceet stacces see ails eee 6 St. Johns Series: St.. Johns. Check. Stamped|.:...\.i. eden oie. ee eee 65 Sts Wlohnse "Plaine icc ccc tion onic cieeelsrorsoe tee ee 10 Pasco: PP laity Mod oti terclu dea erate les essieus cies Gneleiare rove aie eee enna I Glades Complex Belle: :Gladei Blain seis escterers ccs oie coo crditie siclerore lens ae eee I Miscellaneous Residual) Plain.» 29 iors ove ois! sre ve sieicieveiaie sites. « toierevere Saree eee 42 Unclassified i iss fraisings el oe webbed Gai swlens kalo e eee eee 10 Total sherds.... 271 In addition to the aboriginal pottery, the collection contained Spanish ware, sheet and coin silver beads, glass beads, a gold-plated clay bead, and an amber bead. WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—WILLEY 335 The diagnostic pottery from this mound argues for a Safety Har- bor Period dating, and the early European trade material confirms this. The Wakulla Check Stamped, Thomas Simple Stamped, and St. Johns types are not sufficiently diagnostic of Weeden Island to predicate an earlier stage for the mound. Both Wakulla and the St. Johns types are known to continue over into the Safety Harbor Period. Other sites in Pinellas County—There were a great many mound and midden sites along both shores of Pinellas Peninsula. Both Walker (1880a, b) and Moore (1900, 1903) saw, excavated, and de- scribed several of these. Circular burial mounds, flat-topped temple mounds, midden heaps, and complex shell works are represented. The burial or temple mounds were made of sand or sand and shell, and they are carefully distinguished from the shell-midden sites. Burial and temple mounds which they list are at Anclote River,®* the Myers site, Dunedin, Saxe’s mound or the “Mound near Clearwater,’ Hog Island, Long Key, Four Mile Bayou, Pine Key, Point Pinellos, and Bethel’s Camp. A flat-topped mound with a complex of shell embank- ments attached to it is described from Maximo Point. Great shell mid- dens were noted at Dwight’s Orange Grove near Clearwater (prob- ably Pi-5), Indian Pass Church, Four Mile Bayou, Bear Creek, Boca Ciega Bay (possibly Pi-6), the vicinity of Point Pinellos, Big Bayou, Cox property, and Booker Creek. These midden sites appear to be the village areas which are related to the temple and burial mounds. In some cases, however, the lower levels of these middens may date back to an early Glades culture horizon (Perico Island Period) which antedates the later Gulf Coast cultures. The mound and embankment complex at Maximo Point is very similar in configuration to some of the Glades area sites of the Glades III Period. HILLSBOROUGH COUNTY Thomas (H1-1).—This site is described under the section on “Ex- cavations on the West Coast: 1923-1936.” Cockroach (Hi-2).—This site is described under the section on “Excavations on the West Coast: 1923-1936.” Picknick (Hi-3).—Sometimes referred to as the Thatcher mound, this site is on the south prong of the Alafia River near the town of Picknick (see map 19 for approximate location). The site was 68 A small collection of Wakulla and Biscayne Check Stamped sherds comes from one of these sites known as “Spanish Wells.” The collection is in the U.S. National Museum (No. 149358). 336 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I13 excavated by J. Clarence Simpson during the Florida State Archeo- logical Survey of 1937 (Anon., 1937). It is described as a mound 60 by 70 feet at the base and 4 feet high. Burials were taken from the lowest mound levels. These were loosely flexed. A pottery vessel of the Safety Harbor Period accompanied one burial. Glass beads of European provenience and small “bird” points of flint were screened from the sand of the upper level of the mound. Simpson describes narrow-necked pottery jars, use of effigy rim lugs or adornos, a frog-effigy bowl, and the trait of vessel “killing” by perforation. Copper-covered ear ornaments of wood were found with a skeleton near mound base. A collection from the mound is now in the Florida State Museum at Gainesville. Dr. John M. Goggin has provided me with the fol- lowing classification : Safety Harbor Complex Safety Harbor Series: Safety Harbor. Incised:)..%..5.<...2e«cs00 sa noon eee 23 Pinellas ‘Tneised) o.oo. 04 vod toe es eae ote ee 43 Pinellas, Plata 6. Sone oo8 seks roan ade Ui Fort Walton frog-efhiey vessel.’ ..c)....i0)<.s +s0e es one eee I Weeden Island Complex Weeden Island Series: Weeden’ Island: Tncised. i.425)0 se cle ellen cle ee 2 Weeden Island \Punctated ) 2.0.0 ncaa eee eee 2 Weeden [sland | Plaini. .:..05 oi «.cshs aes scene eee 2 ittlesManatee:Zoned Stampeduaec) aeance ec center I St. Johns (or Biscayne) Series: St.) Johns: Check) Stamped. '3...0¢0.62)..04 505026 eee 2 St. Johns: Plats ce sade och cs eelculs eiele.ce lo Uslealeele ee ee I Glades Complex Belle Glade Plaines ozads saga secie soe ed 6 \sicim ete ale eeslerel a ee 9 Miscellaneous Residual “Plann ese outesrecivs crores ace a there trios ore everett eee 9 Unclassified complicated stamped?’ .\...c)..0. 220+. ose seen I Red bottle form (probably Safety Harbor)...............06. I Total sherds.... 104 The collection is divided between Safety Harbor and Weeden Island II Complexes. The probabilities that the mound, or its lower levels, were built in pre-Safety Harbor times are strengthened by the fact that a burial with copper-covered ear plugs was found near mound base. The European contact seems superficial. WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—-WILLEY 337 Jones (Hi-4).—This mound, on the east bank of Pemberton Creek near Thonotosassa, was also investigated by Simpson (Anon., 19392) (map 19). It was a sand mound 75 feet in diameter by 3 feet in height. It was partially surrounded by a horseshoe-shaped ridge of sand which opened on the east. Most of the 174 burials were in or on the old basal or premound layer. The common burial form was semiflexure. Some artifacts were in direct association with the dead. Evidences of house floors are reported at 12 inches below the surface of the mound; however, there were no post molds in association. Most of the mound pottery was found in caches where it had been placed after being destroyed. Red ocher, “killed” shell cups, stone and shell pendants, including some carved animal forms, stone celts, and shell tools were found in addition to the pottery. A Florida State Museum collection from the mound has been clas- sified by Goggin: Safety Harbor Complex Safety Harbor Series: Salcevatachor. MiGised sin wis sgreamvacaiies sapreiestelaa’doies serine 3 Te Masa MCI SCS oyarsrasa,crcveiscourei st oreieueteieie iors Sle cuewisielaisieiale sleietece 6 Weeden Island Complex Wieedenel'sland.iPunctatediiva-tr. cietajcsteleione © o suetesscie sie.cioictereteys hers 17 Walaa Check: vS tarp ed sore) er-nejeiareictela) oteveles», oisssiererelsielseieleisicre'# sei I St. Johns (or Biscayne) Series: Sonus Check, stampede ets an. clonic sicieta ssa es tees I SE ounsyE lai... rere cae eee eee eae a hate wees es 4 Miscellaneous RCSIGUalMeE Laltin Sarecterca cece ciate sinerane sicterolecverteistchetecersloreterorecers 8 Total sherds.... 40 The mound dates from the Weeden Island II and Safety Harbor Periods. Snavely (Hi-5).—This burial mound, near Thonotosassa (see map 19), was excavated by Simpson during the Florida 1937 survey (Anon., 1937). Goggin has examined collections from the site and dates the material as of the Safety Harbor Period. Buck Island (Hi-6).—This mound is located near the junction of Cypress Creek and Hillsborough River, 12 miles northeast of Tampa (map 19). It was excavated by Simpson, and a collection 338 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I13 in the Florida State Museum at Gainesville has been classified by Goggin: Safety Harbor Complex Safety Harbor Series: Safety’ Harbor Inctsed: «.... 0. sale sects sot ee eae II Pinellas Tncised: 42 \yadeicas ss ata nee dae detks «Sos eee 9 Pinellas: Platienccecce cies ascew ae cabo sores. ee 13 Weeden Island Complex Weeden Island Series: Weeden Island “‘Punctated!s. o2.c..cccicca. «2. oe se 17 Catrabelle Punctateds oo 203040 fA cee oo de I fLampa “Complicated: Stamped... zaveiscics c/s ecd/sae se ele eee I Wakulla: ‘Check :Stamped’s..4.. dace’ «6 asics «a cic op bas ee I Combination Carrabelle Incised and Wakulla Check Stamped.. 1 St. Johns (Biscayne) ‘Plains. 3.7.00 ts ees se + a aheee ee 2 Papys Bayou: Punctatedos.in Wes fis ilies ig oe oe ee 5 Glades Complex Belle\Glade. ‘Platine acs csc ceca ok cei oe o's) areas ae I Miscellaneous Residuals Plats ios /a3,0s siais ie. ote t103se 304 8 ae ete eee 21 Total sherds.... 83 This mound is also divided between a Weeden Island II and a Safety Harbor dating. Rocky Point (Hi-7).—The site is mentioned as a big shell heap 5 or 6 miles west of Tampa (Shepard, 1886) (map 19). Earlier, Stearns had given it a brief note (Stearns, 1870). There is a collec- tion in the United States National Museum (Nos. 3246632-3246663) which is classified below : Safety Harbor Complex Safety Harbor Series: Pitiellas: Trictsed:.... (sis. s:< ssasisad a aane's aero sa doe ee I Pinellas’ Plain: cached eats tase csek ee oe eee 6 Weeden Island Complex Papys Bayou Series: Papys: Bayou Punctated siow .cvtents sic5.« hore veto sie I Papys, Bayou (Plain. 2 hci 3d cr<.naja0 0 s/nscle-s/cisa o's oe I Hillsborough: Shell) ‘Stamped si. << s..5.c10 ss oc ew ce oe cae sien eaae I Biscayne (or St. Johns) Series: Biscayne Checky Stamped io/cie.< aioe «.cvenas, att cielo cee I Biscayne, Plait |, 6; sis/sfs:< sire, nin.0 spoihiels Ske si ess eee 2 West Florida, Cord=marked)s ..c0:< o.0,.0s «c's store c ee sige eee 2 (Probably Late Variety.) WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST——-WILLEY 339 Miscellaneous Wmnestone-tempeked! CoLrd-mahiceds rs ste cis cietecle scree cieveie eres isle ale I (Related to Pasco Series.) (ORO ESE th 5 ha tari Cidia BCI TOnIO ICICI IR eI obo 105 cls I Total sherds.... 17 Both Weeden Island II and Safety Harbor Periods are represented. Other sites in Hillsborough County.—There are numerous other sites, both mounds and middens, in Hillsborough County. One of the first to attract attention was the old Fort Brooker mound near Tampa. This was investigated and described by Lt. A. W. Vogdes in the 1870’s, but the best account and appraisal is given in Walker (1880a, pp. 411-413). This was probably a platform or temple mound, and it is described as being constructed of alternate sand and shell layers. At least one burial was found in addition to some check stamped pottery. Another large and well-known site is the Bullfrog mound on Bull- frog Creek, a tributary of the Alafia River. Shepard (1886) and Walker (1880b, pp. 421-422) describe it, and Moore (1900, pp. 357- 358) gives it brief mention. Long destroyed, the site appears to have been a complex of high shell mounds, shell ridges, and shell refuse piles. A similar site, at least in outward appearance, is Mill Point on the Alafia River (Moore, 1900, pp. 356-357). The Alafia River is also known for archeological objects taken from the phos- phate beds in its immediate vicinity. Among the more recent mound excavations that have been carried out by J. C. Simpson are those at the Spender, Cagnini, Branch, and Lykes mounds (Anon., 1937). Another mound in the vicinity of Tampa was excavated by J. J. Hall. This site, judging by a collection examined by Goggin in the University of Michigan Museum, was probably of the Safety Harbor Period. Two excellent collections, but both lacking location data, probably came from Hillsborough County. They are listed only as “Tampa Bay.” One is at Peabody Museum, Harvard University, and was ob- tained by F. W. Putnam in 1899. It is pure Safety Harbor but with a large increment of Lamar-like complicated stamped. The other was donated to the United States National Museum (No. 35373) by J. W. Milner. It is divided between types of the Safety Harbor and Weeden Island Complexes. 340 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I13 MANATEE COUNTY Parrish mounds 1-5 (Ma-tr, 2, 3, 4, 5).—These sites are described under “Excavations on the West Coast: 1923-1936.” Perico Island (Ma-6).—This site is described under “Excavations on the West Coast: 1923-1936.” Shaws Point (Ma-7).—This is one of the best-known sites on the Gulf Coast (map 20). Walker (1880b, pp. 416-422) described it as Map 20.—Site map of Manatee, Sarasota, Charlotte, and DeSoto Counties. Sites indicated by numbers. a series of shell tumuli stretching along the water front for over 150 yards with some of the hillocks being as much as 20 feet in height. The location has also been considered the landing place of the DeSoto expedition. In spite of its general notoriety, little is actually known of the archeology of the site. It was Walker’s opinion that the site represented a great refuse heap or series of such heaps. Animal and fish bones, evidences of old fires, and potsherds were observed on the surface as well as in occasional cross sections. There is a collection of pottery and artifacts from this site in the United States National Museum (Nos. 317078-317432, 329771- WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—WILLEY 341 329774, 341237-341244) which was presented by Charles T. Earle. It is given below: Safety Harbor Complex Safety Harbor Series: DAeLy) Hlathon PUCISEU ses Monet fades eka asec ees I Pineilas INCised. Gk tiatacs ve. ciad Suse yobs Ske Nea S See eee 10 LEAS RIO 2 hi cee ne fey moat rt aE ae ee Lm ee Oe, SU 4 Meare @heckss Stamped Macq sisios cts 35/00. daln ss blo'ek aaske ons Cara aie 5 1 P/S2 Seca) Stites EET WM tye 7 ON Rina ange en Pas ie Aint ety IO MwAnled hab ric-impnressed: te.ce claus «cisress aesais s Sa eeies Se ioe B (Shell-tempered. ) Weeden Island Complex Weeden Island Series: NVieGen US) anid STACISE.,cajereoiss 0 Seda avec vad ars hoe sels I Weeden (land. Panckated: 2 so. sie occ ucse iv ole Sclaic s dare wre 2 Kieedem Ustanidiy- Plaines sscccc.cu Stree Shas a araiels, secs sos Sooke sad 3 WraltillanCheclke:Stampedjccn a concrete on enc oeee ate eas 8 Complicated Stamped Series: Switt Creek Complicated Stamped... ic0.c00. 08 ee cess 3 San (City, Complicated ‘Stamped osc dec.) scicecewis asec ass I Puiishorouch, Shell Stamped... /ccelen os ecene ig scenes cece os 4 EAU SAV OU ME LINCEALCU fora cea avateia. «rd o stancietictne natekae vies Saisie 2 leittewvtanatee,Zoneds Stampedsss cic come c ce oockecce tien cceer I Biscayne (or St. Johns) Series: Pascayne. Clreck=Stamined g ais.e:c) x caickein 3 4a5 Sounr< bie weciele ora ayole 14 ECU ICME DAMEN yikes hadas eivis'e Way teidine ie stevewiolataes Ge table wcle 3 SENG VALES RCO eee c ot oe fers eta hae aa na oats clean tate mace ens 3 Simple Stamped; Biscayne pasts s .sctisrers a6 ames eo thats oot 5 Santa Rosa-Swift Creek Complex Pulivaton bayou stamped saris eee oe aoe lect ones I Facing ange EAN” $8 tanec eialcrars ats bis We weer Gerace sie Bieler siete. ociea er I Deptford Complex Weprorcds Bold GheckStampede-easreaececeecticteine se eee I Glades Complex (GLE ES AT ELET as ei ea re A e O el P ee ea 34 EEN ME PANT ge oe ch ts sarc oe egy wes oeeiene Sree S slave nl eeale save males arenes 3 Belem Glades laine. sire ca cro ners cae eis oe ere eee 3 Miscellaneous Sr OO tineeby] aise eaeteyey cre veucve sieve, ecg vor eveysr ah ciets oy cepatatororeteete eravecehstale te I Beessteliaet ln PATE cccte wash ecs aiesapaiais.s Soca seuetan sue Meniecege da eee Gaaene oh 2 MRE ASSURE er eiete cian svete Midden sites of the period are evenly distributed along this coast. Fewer cemeteries are known, but these, too, are found in all parts of the region. Burial mounds are rare. Temple mounds are known throughout the region but are most common near Tallahassee. They also occur inland on the Apalachicola River. Settlement pattern.—The coastal villages of the Fort Walton Period are much the same as those of preceding periods. Most of the sites are located along coasts, bays, or waterways. They are characterized by shell-midden refuse, and vary in extent from 25 to 100 meters in diameter. Most of the larger Fort Walton midden sites are mixed 75 This distribution of Fort Walton sites was, of course, one of the reasons for establishing the northwest coast region as apart from the other subdivisions of the Gulf Coast area. 454 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I13 sites. Fort Walton Period middens seem to have less depth than those of the preceding periods. This was observed by excavations in mixed sites where Fort Walton Period refuse was found to be very superficial. The inland villages of the Fort Walton Period are larger than most of those found on the coast and contain only scant amounts of shell refuse. It is not known if Fort Walton sites are more numerous in the interior than sites of other periods; this impression that they might be is undoubtedly strengthened by the fact that many of the inland Fort Walton sites are marked by large mounds, and these have attracted more attention than would be accorded small middens or small mounds and are, consequently, better known. Several of the inland Fort Walton sites have large flat-topped temple mounds and rather extensive surrounding village areas. One such site in the vicinity of Tallahassee (Lake Jackson (Le-1)) is a group of six temple mounds. Temple mounds also occur at the larger Fort Walton sites of the coast, but these coastal mounds do not compare in size with the big mounds of the Tallahassee section or the upper Apalachicola River. A review of the Fort Walton temple mounds reveals the following with regard to shape and size. Of the eight definite mounds, six are rectangular and two circular. Two mounds, questionable as to a Fort Walton Period dating (Pierce mound C (Fr-14) and Nichols (Wa-3)), are also rectangular. The rectangular mounds vary in height from 4 to 30 feet. Basal measurements are as much as 195 by 145 feet in one case and 150 by 150 feet in another. All the rec- tangular mounds tapered upward to a rectangular flat top or platform which was usually about one-half as large as the basal area. Shape outline was both square and oblong, and sloping earth ramps were observed leading to the tops of some but not all of the mounds. The two circular mounds are 3 and 7 feet high, respectively. These are mounds which had been excavated (Bear Point (Ba-1r) and McBee’s Mound (W1-4)) previous to any reported observations. It is possible that their outlines had been modified by such excavations. Two burial mounds of the period are reported and, in addition, there are two examples of Fort Walton Period reuse of Weeden Island Period mounds. Four Fort Walton cemeteries are known. Economy.—On the coast marine foods continued to play some part in the Fort Walton diet, although this seems to have been a diminishing one. Interior sites, such as Lake Jackson (Le-1), show very little shell refuse, the debris on these sites being mostly decayed organic matter. Animal and fish bones still occur in middens, but the impres- WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—-WILLEY 455 sion is that they, like shell, compose a smaller proportion of the rubbish than in former periods. These observations concur with what we would expect for the period. The Fort Walton culture is essentially Mississippian in type and equates with the late Middle Mississippian time horizon in the Southeast. As other Mississippian cultures pos- sessed a well-developed agricultural economy at this time, there is little doubt that agriculture was known to the Fort Walton peoples, although we lack the definite finds to prove it. The suggested shift of the bulk of the population at this period, from the coast to the more fertile interior, fits in with the assumption that native agriculture became more important and widely practiced during the Fort Walton Period than previously. Organization of society—There are archeological implications for change in the political and social organization of the Fort Walton Period as opposed to the Weeden Island. In the first place, the ratio between the number of temple mounds and the number of midden sites during Fort Walton shows fewer mounds than comparable ratios between Weeden Island or Santa Rosa-Swift Creek burial mounds and middens. This reduction in the number of ceremonial sites or centers, with relation to the numbers of middens or villages, indicates a trend toward politico-religious cohesion in the Fort Walton Period. It is doubtful that this trend resulted in the entire northwest coast region becoming a single political unit in the Fort Walton Period, but it suggests that there were only several rather than a great multi- tude of autonomous units. There is support for this in the ethnohis- torical accounts of the sixteenth century which describe what seem to be political assemblages or federations that include several com- munities, a considerable area, and a capital town. Secondly, the temple mounds themselves imply a different social orientation than the burial mound. The latter is solely a place of burial while the former is a means of lending prestige and impressive- ness to a political or religious building and its occupants. Burials are sometimes found in the temple mounds, but these are clearly later additions incidental to its principal function. The shift from the Weeden Island burial mound to the Fort Walton temple mound suggests a decline in the importance of the death cult and a con- comitant rise in the significance of tribal leaders. Together with the presumed larger political units of the Fort Walton Period, this would mean that secular powers and functions of the leaders had been in- creased over previous periods. The nature and distribution of grave goods conforms to this in- terpretation. Specially made mortuary ceramics disappear and, in 456 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 113 general, grave goods are less abundant and elaborate than they were for either Santa Rosa-Swift Creek or the Weeden Island Periods. How- ever, grave artifacts, especially pottery, are more often found with individuals than placed as group offerings, and there is a greater varia- tion in the amount and quality of objects placed with individual burials than during the Weeden Island Periods. Disposal of the dead.—The burial customs of the Fort Walton Culture, as compared with Santa Rosa-Swift Creek or Weeden Island, are neither as distinctive nor as standardized. Fort Walton dead were buried in cemeteries, in the floors of temple mounds, and in burial mounds. Among the latter there are instances where earlier mounds of the Weeden Island Periods were used as repositories for Fort Walton burials. The records of burials in three temple mounds, Bear Point (Ba-r), Fort Walton (Ok-6), and Jolly Bay (W1-15), all state that the graves were intrusive into the tops and flanks of the mounds. At Bear Point and Jolly Bay burials were of the secondary types. The burials found in the big Fort Walton (Ok-6) mound were both secondary and pri- mary extended. The other Fort Walton Period temple mounds have not been thoroughly explored for burials. The burials in one of the Fort Walton burial mounds were primary extended, and in the other they were all secondary. In the two Weeden Island mounds, into which Fort Walton Period burials had been intruded, the Fort Walton burials were secondary in one case (Marsh Island (Wa-1)) and both secondary and primary flexed in the other (Chipola Cut-off (Gu-5)). The cemeteries contained both massed and scattered second- ary burials in two instances, Hogtown Bayou (WI-9) and Point Washington (WI1-16), and extended and flexed primaries in the other, Wildlife Refuge (Wa-15). The Hogtown Bayou and Point Washington cemeteries were characterized by low rises of sand, not over a foot in height and a few feet in diameter. These small hum- mocks marked areas of mass burials. Burials in the temple mounds ranged from 27 to 66 in number, although I think it likely that some of the big mounds actually yielded more than were tabulated. Ap- proximately 100 individuals were found in each of the two cemeteries with the massed secondary burials. Large destroyed caches of pottery vessels were found in both the Fort Walton Period temple mounds and in the cemeteries. These were also found in the burial mounds, although in the case of the Chipola Cut-off (Gu-5) mixed-period mound it is not clear from Moore’s ac- count (1903, pp. 445-466) whether these caches consisted of Fort Walton as well as Weeden Island Period vessels. As mentioned, WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—-WILLEY 457 pottery deposition for the Fort Walton Period often differed from Weeden Island in that vessels were frequently found with individual skeletons. In fact, a distinctive Fort Walton trait was the inversion of a large pottery vessel over the skull of a dead individual. This was noted both in temple-mound and cemetery burials of single skulls. There is also one instance of urn burial at Marsh Island (Wa-1). Nonceramic artifacts were commonly placed with individual skeletons and were also found unassociated in cemeteries and mounds. Cranial deformation was noted for Marsh Island (Wa-1) and Point Washington (WI1-16). The sum total of impressions concerning Fort Walton disposal of the dead is that the period was a time of change in mortuary customs. The old burial-mound idea competes with the newer modes of inter- ment in temple-mound floors or in cemeteries. The fact that burials were placed in the earlier Weeden Island mounds suggests some con- tinuity of the burial-mound practice, but the custom had waned to such an extent that the desire was lacking to build a mound for the purpose. Burial in the temple mounds may, to a degree, represent a continuity of the mound-burial idea although in a different form. Treatment of the bodies of the dead was much the same as in the earlier periods. Charnel houses and bone cleaning must have remained in vogue. Primary burial, more often extended than flexed, also was practiced. Old ideas concerning mortuary goods were retained to the extent that vessels were still intentionally “killed” or destroyed and sometimes mass caches of pottery were buried near the dead as common offerings. New elements are seen, however, in the vessels placed specially with individuals, and the greater emphasis on indi- vidual grave artifact arrangements. Ceramic arts——Fort Walton pottery is technically competent and abundant, but it lacks the esthetic excellence that characterized Weeden Island. As a ware it is somewhat coarser than the previous periods with either a heavy grit or crushed shell being used as temper. The appearance of shell temper is one of the diagnostic traits of the period. Surface decoration, except for some check stamping, is exe- cuted mainly in heavy incision and punctation. Red pigment is occa- sionally used, and engraving appears so infrequently that it is con- sidered a foreign element. Modeled ornamentation is fairly common and takes the form of zoomorphic and anthropomorphic rim adornos. These adornos represent the Fort Walton ceramic art at its best. There is a tendency for vessels to be larger than in the preceding periods, and the casuela bowl and the bottle are the most distinctive and the most common of the new forms. As opposed to the earlier 31 458 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I13 periods, there is a strong tendency for conformacy in both design and vessel shape. The exotic effigy forms of the Santa Rosa-Swift Creek and Weeden Island Periods are missing. Design elements and arrangement are more standardized and predictable than before. Except for the quite realistic adornos there is little detectable life naturalism in the decoration. Designs are both curvilinear and rectilinear and highly conventionalized. The drawing lacks the grace, lightness of touch, and flowing quality of the Weeden Island style. Fort Walton ceramics represent a break with the older Gulf Coast traditions in that new influences of a Middle Mississippian origin have been strong factors in the formation of the style. Some resem- blances to the incised and punctated types of Weeden Island are suggested, although it is not clear if this is the result of a local Gulf Coast continuity or some more indirect relationship. An obvious local carry-over is the type Wakulla Check Stamped which lasts on into the Fort Walton Period as a minority type and probably serves as the basis for the development of the later type, Leon Check Stamped. Moundville Engraved, an obvious inland Alabama type, occurs on the Gulf Coast during the Fort Walton Period, probably as trade. The Fort Walton pottery types in the order in which they are described : Fort Walton Series: Lake Jackson Plain Fort Walton Incised Point Washington Incised Pensacola Series: Pensacola Plain Pensacola Incised Pensacola Three-line Incised. Pensacola Red Marsh Island Incised Moundville Engraved Type Descriptions Type name-—LAKE JACKSON PLAIN. Definition as a type: From the Florida northwest coast, present paper. Brief synoptic description in Willey and Woodbury (1942). Based upon rim and appendage sherds. Associated body sherds were grouped in the Residual Plain category, although the plain ware of the Fort Walton Period can be sorted out from Weeden Island and Santa Rosa-Swift Creek unidentified sherds with about 80 percent accuracy. Ware characteristics: Method of manufacture: Coil fractures observed. Temper: Sand and medium-size grit particles and what appears to be crushed clay. In general, tempering material coarser than in previous periods. WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—WILLEY 459 Paste texture and color: Paste is sometimes fine and compact but more often it is coarse, lumpy and contorted. Paste cores are usually gray; fired surfaces are often whitish buff, buff, or reddish buff. Surface texture, color, and finish: Surfaces are smoothed but never achieve a polish. They are distinguished by temper particles, usually pieces of quartz, which extrude through onto the surface. In the case of the predominantly clay-tempered specimens hard clay particles give the surface a coarse texture. Surfaces are commonly mottled, and vary in color, depending upon the degree of firing. (See pl. 44.) Hardness: 3 to 4. Thickness: 6 to 9 mm., with average about 8. (\ oot. Db . ‘ . : , e ' ' Fic. 56.—Fort Walton Period rim forms. Lake Jackson Plain. (Interiors to right.) Form: Total vessel: Vessels tend to be larger than in earlier Gulf Coast periods. The casuela bowl, collared globular bowl, globular bow! with flared orifice, open bowl and complete frog-effigy bowl are all recorded. Rim: Casuela rims are inturned. Collared ollas have straight, inslanting or outflared rims. Rims often thickened near or at margin. Some rims are unmodified. Exterior rim folds, when they occur, are thin and flat but wide. They are often pinched or fluted. Sometimes a strap of clay was added to the exterior wall just below the rim, and this strip is pinched or fluted. Small applique nodes often added along rim exterior in place of fluting. Fingernail or long punctations, placed in row beneath rim, or on rim exterior, are another variation. (See fig. 56.) 460 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 113 Lip: Sometimes pointed or round-pointed, sometimes flat or squared. Most characteristic feature is a row of close-spaced notches which are always placed diagonally on the exterior edge of lip. Base: Rounded. Appendages: Small vertically oriented lugs placed on the rim exterior. These may or may not have a projection above the lip. Both vertically placed loop and thick strap handles occur. The large ones are 6 to 8 cm. long and over 2 cm. wide. Small loop handles are 3 to 4 cm. long and about I cm. wide. Some handles have single, double, or triple nodes. Geographical range of type: Most common in the northwest from Pensacola to St. Marks and inland as far as the Tallahassee section and well up the Apalachicola River system. Rarely found in central coast or Manatee regions where it is replaced by Pinellas Plain. Chronological position of type: Fort Walton Period. Continues into Leon- Jefferson Period. Relationships of type: Related to late Mississippian pottery of south and central Alabama and Georgia. Lamar Series types have similar rim treatment and shapes. Related to Pinellas Series of Safety Harbor Period and to types of the Leon-Jefferson Period. Bibliography: Willey and Woodbury (1942, pp. 245-246). Type name-—FORT WALTON INCISED. Definition as a type: From the northwest coast of Florida, this paper. Briefly described by Willey and Woodbury (1942, p. 244). Has been revised since that time to include only “substyles 1 and 2” as they were defined in that paper. Ware characteristics: (See Lake Jackson Plain.) Decoration: Technique: Lines and punctations incised into soft surface of vessel. Lines are deep, wide, and usually rectangular in cross section. Large round dot or square punctations most common. Hollow-reed punctations sometimes used. Design: Elements are volutes, interlocked scrolls, running scrolls, circles, trifoil figures, crescentic forms, S-shaped and reverse-S figures, recti- linear stepped figures, pendant loops, and triangles. Elements are usually repeated around vessel in a connected design pattern. Dot punctations used as filler for both backgrounds and for design proper. Incised lines sometimes used as filler. (See fig. 57, a-d; pl. 43; pl. 45, f, g.) Distribution: Around upper part of bowl, as a rule, and on vessel exterior. Interior decoration occurs on upper surfaces of rim appendages or pro- jections to large, open bowls. Form: Total vessel: Shallow bowls with lateral expansions, casuela bowls, collared globular bowls, short-collared jars, beaker-bowls, bottles, gourd-effigy forms, flattened-globular bowls with effigies affixed. Rim: Usually inslanting or incurving but depends on vessel form. Most rims thickened except for a perceptible thinning at the lip edge. Long, Fic. 57.—Fort Walton Period vessels. a-d, Fort Walton Incised; e, Pensacola Three-line Incised; f, Point Washington Incised. (Redrawn from Moore. Year, figure number, and site: a, 1901: 24, Fort Walton; b, 1903: 105, Chipola Cut- Off; c, 1901:17, Fort Walton; d, 1901: 22, Fort Walton; e, 1901: 81, Point Washington; f, 1901: 96, Point Washington. Vessels a, b, d, f are 4+ actual size; c, e, 4 actual size.) 461 462 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 113 thin folds are common. These are usually underlined with an incised line. (See fig. 58.) Lip: Rounded or round-pointed. Rather close-spaced notches are placed diagonally on the exterior margin of the lip. Base: Rounded. Appendages: Lateral or horizontal rim projections. Bird head-and-tail effigies placed on opposing sides of rim. Small vertical lugs, usually four to a vessel, placed just below the lip on exterior. These lugs may be flush with lip or may project above it. Fic. 58.—Fort Walton Period rim forms. Fort Walton Incised. (Interiors to right.) Geographical range of type: Northwest coast of Florida with inland extension for at least 100 miles. Distribution unrecorded along Alabama coast and interior Alabama but probably occurs in these areas. Chronological position of type: Marker type for Fort Walton Period. Relationships of type: General similarity to many late Middle Mississippian incised types of both grit- and shell-tempered wares. Rather close to Lamar Bold Incised of Georgia. Undoubtedly is prototype and con- temporary of Pinellas Incised of Safety Harbor Period of the central Gulf Coast. Bibliography: Willey and Woodbury (1942, p. 244). WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—WILLEY 463 Type name—POINT WASHINGTON INCISED. Definition as a type: Northwest coast of Florida, this paper. Previously in- cluded in a very brief description of Fort Walton Incised (Willey and Woodbury, 1942, p. 244: see “substyle 3”). Ware characteristics: (See Lake Jackson Plain.) Decoration: Technique: By incision in soft vessel surface. Use of series of two, three, or four lines to carry out all designs (fig. 57, f). Design: Loop figures in isolation, complicated scroll patterns, running scrolls, ovals intersected with cross bars, diamonds and V-shaped figures, and combinations of curvilinear and rectilinear elements into composite patterns are the principal design motifs. (Pl. 45, a-e.) Distribution: Most often confined to upper portion of vessel exteriors but there are examples of designs extending to and over the base. Form: Total vessel: Shallow bowl or dish, flattened-globular bowl, casuela bowl (most common), collared globular bowl, short-collared jars, bottles (common), double bowl, jar with cambered rim, simple bowl with effigies affixed, gourd-effigy form. Rim: (See Fort Walton Incised.) Lip: (See Fort Walton Incised.) Base: Usually rounded. Appendages: Animal or bird head-and-tail effigies on bowls. Geographical range of type: Northwest Florida coast. Seems more common to the western end of this range. Probably extends into coastal Alabama and interior Alabama. Chronological position of type: Fort Walton Period. Relationships of type: Closest similarities of design are seen in Fatherland Incised, a Natchezan type of southern Mississippi (Quimby, 1942) which utilizes the parallel three- or four-line element in the formation of various scroll and other curvilinear designs. Also has close affinities with Fort Walton Incised in ware, shape, and decoration. Bibliography: Willey and Woodbury (1042, p. 244: “substyle 3” under Fort Walton Incised) ; Moore (1901, Point Washington cemetery site, pp. 472-496) ; Quimby (1942, pp. 263-264). Type name—PENSACOLA PLAIN. Definition as a type: Northwest Florida, this paper. Ware characteristics: Method of manufacture: Some fractures suggest coiling method, although this is not as definite as in the other types or series. Temper: Crushed live shell. There is also a little sand and grit. Paste texture and color: Sometimes compact; sometimes laminated and contorted. Paste core usually is gray and surfaces usually buff or red- buff. In some cases pottery was fired gray-black throughout. Surface texture, color, and finish: Surfaces were probably smoothed and polished before erosion. Most specimens are pitted as a result of the temper particles leaching out. Several of the black or gray-black sherds have retained a polished, unpitted surface. Color varies according to firing. 464 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I13 Hardness: 2.5 to 3.5. Thickness: Ranges from 5 to 11 mm., with average closer to lower figure. Form: Total vessel: Forms undoubtedly comparable to those of the Fort Walton Series. Rims: Usually unmodified except for an occasional heavy, round exterior fold. Lip: From flat to round-pointed. Base: Probably rounded. Appendages: Small vertical loop handles and ornamental nodes beneath the rim. Geographical range of type: Most common in extreme western end of north- west Florida, but it is found in small quantities as far east and south as Tampa Bay. Probably very common in south Alabama. )\\ Fic. 59.—Fort Walton Period rim forms. Pensacola Incised. (Interiors to right. ) Chronological position of type: The Fort Walton Period. Also found as a minority type in Safety Harbor sites. Relationships of type: Related to shell-tempered ware of late Middle Mississip- pian horizon throughout the Southeast. Has more specific relationships to types in south and central Alabama. Resembles Lake Jackson Plain in vessel forms. Type name-—PENSACOLA INCISED. Definition as a type: This paper, northwest Florida. Description: Ware same as Pensacola Plain. Vessel forms comparable to those described for Fort Walton Incised. (See fig. 59.) Decoration about the same as Fort Walton Incised except there seems to be a little more emphasis on incision with less punctation. (See fig. 60.) Some designs suggestive of highly stylized “death’s-head” motif. Geographical range and chronological position: Found mainly in western north- west Florida but extends to the east and south as a minority type. Belongs to the Fort Walton Period. Fic. 60.—Fort Walton Period vessels. All Pensacola Incised. (Redrawn from Moore. Year, figure number, and site: a, 1901: 21, Fort Walton; b, 1901: 2, Bear Point, Alabama; c, 1901:5, Bear Point, Alabama; d, 1901: 37, Fort Walton. Vessels a-c are } actual size; d, 4.) 465 466 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 113 Type name —PENSACOLA THREE-LINE INCISED. Definition as a type: This paper, northwest Florida. Description: Parallels the type Point Washington Incised except that the ware is shell-tempered (fig. 57, ¢). Geographical range and chronological position: Western end of northwest Florida and probably south and central Alabama. May extend to south Mississippi to merge with distribution of Natchezan type, Fatherland Incised. Belongs to Fort Walton Period in Florida. Type name-—PENSACOLA RED. Definition as a type: This paper, northwest Florida. Description: A Pensacola Series shell-tempered ware which has been slipped with a medium to dark red pigment. Not common but very distinctive owing to temper and color. Type name-—MARSH ISLAND INCISED. Definition as a type: This paper, from northwest Florida. Description: Very limited data. Observed in Moore (1902, fig. 241) and in a few surface collections. Ware characteristics probably parallel those of Lake Jackson Plain. Decoration by medium-bold line incision. Design in band around vessel below rim. Consists of opposed triangles which are filled with parallel incised lines. Also chevron arrangements made up of diagonally placed series of parallel incised lines below rim, and parallel vertical lines below rim. Open-bowl form. May be related to the earlier type, Carrabelle Incised. Type name-—MOUNDVILLE ENGRAVED. Definition as a type: Has been informally recognized by Southeastern workers for several years but not yet published as a type. Very brief sketch offered here is based upon a vessel in the Moore collection from Jolly Bay (Moore, 1901, figs. 53, 55). Ware characteristics: A black polished pottery tempered with finely crushed shell. Decoration: Employs both incision and engraving. Major outlines of designs usually done in broad-line incision. Background work done in fine- line engraved hachure or cross hachure. Designs usually consist of life forms such as eagles, men with eagle masks, human skulls, etc. These design elements are a part of the Southern Cult (see Waring and Holder, 1945). A collared globular bowl is shown by Moore. This specimen appears in a Fort Walton Period context which is ap- proximately coeval with, or only slightly later than, other similar examples of cult art in Alabama and Georgia. Other arts and technologies——As in Weeden Island, ceramics were the outstanding craft products of the Fort Walton Period or, at least, the outstanding examples still available to the archeologist. The virtual absence of stamped pottery may signify a decline in wood carving, al- though this is by no means certain. Chipped-stone work seems less well developed than in the earlier periods. The large, fine blades found in on WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—-WILLEY 467 Santa Rosa-Swift Creek and Weeden Island burial mounds are not as common as formerly. Some projectile points and knives do occur in middens and as burial goods. Medium-size triangular-bladed stemmed and barbed points and large triangular points with tapered stems were found in midden-site excavation along with chipped-stone scrapers. The small triangular points, so common to other late Mississipian cul- tures in the southeast, may be a feature of the Fort Walton Complex, but our data are insufficient to make a definite statement of this. Stone celts, hones, pebble hammers, discoidal stones, and beads are the re- ported ground-stone products. The celts of the period are of two forms. The medium to large pointed-poll celt (pl. 42, 7), also found in Weeden Island, is present, as is a smaller, thinner, rectangulate celt (pl. 42, &) which is more diagnostic of the Fort Walton Period. The only ores or minerals found in Fort Walton burial sites are hematite and limonite. Shell cups are rare, but shell tools, beads, and long spike-form ear pins are frequent. Shell plummet-type pendants and shell gorgets are occasionally found. Bone awls, bone fishhooks, and bone beads are recorded from one or two sites. In general, except for shell ear pins and shell beads, the shell and bone industries have declined from earlier standards. Native metalwork is exceedingly rare. At one site, two copper spear-form objects were found, but it is not certain if these were made from aboriginal copper. Curiously, no pipes of either stone or pottery were reported from any of the sites reviewed. At two sites toadstool-shaped objects of pottery were recovered. These may have been pottery trowels or, possibly, as Moore suggested, bottle stoppers. Objects made by the Indians from European importations include sheet-silver, copper, and brass ornaments and tablets. European manu- factures are iron tools, nails, and weapons, silver buttons, coins, glass beads, and a glass ring. These materials indicate either casual or trade contacts, and, with the exception of metalwork, they probably had little effect upon native crafts either as a deterrent or as a stimulant. In general, Fort Walton manufactures reflect little change in stone- work other than a probable slight decrease in stone-chipping skills. Stone tools and artifacts are otherwise about the same as in Weeden Island except for the appearance of discoidals which probably heralds the introduction of the “chunkee” game. Shell ornaments are more common than in Weeden Island and the shell ear pin is a new type for the Gulf Coast area. Otherwise, work in shell is even less popular than before. The shell drinking cup as an element of burial furniture is 468 _ SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I13 rare. Virtually all native metalwork is gone. Trade contacts in copper or copper artifacts, vigorous in Santa Rosa-Swift Creek but faint in Weeden Island, seemed to have now disappeared altogether. Galena and mica, two trade minerals from the north, have also dis- appeared. The smoking custom has either disappeared, weakened, or changed its form of appliances. European metals and metal objects were reworked by the Indians during the period, and this may pos- sibly account for diminished trade with and interest in sources of northern native copper. European manufactured items were, ap- parently, rare and highly prized as tools, weapons, and ornaments. Speculations on population and period duration—The guess-com- putation formula which we employed in estimating population totals for the Weeden Island and Santa Rosa-Swift Creek Periods is not equally applicable to the Fort Walton Period. In Fort Walton times the Indian settlement pattern had changed. Temple-mound sites were the ceremonial and political centers and the population nuclei. For Santa Rosa-Swift Creek and Weeden Island periods we used the number of burial mounds as the basis for our population computa- tion; for Fort Walton this is not feasible as it is certain that we have nowhere near a representative record of the number of burial sites for the period. Cemeteries, being unmarked by mounds, are difficult to locate, and the four of which we have note probably represent but a very small proportion of the total, while burial mounds had vir- tually disappeared. For Fort Walton our best possibilities for reason- able estimates lie in a consideration of the middens and temple mounds. There are a total of 23, pure and mixed, Fort Walton habitation sites. These are simple middens without mounds. This number com- pares favorably with the 50 Weeden Island Period middens which we recorded, and with the Santa Rosa-Swift Creek midden total of 19. Both Santa Rosa-Swift Creek and Weeden Island counts per- tained to the entire Gulf Coast area. The Fort Walton count, as op- posed to this, pertains only to the northwest coast region. Hence, a total of 23 middens for the Fort Walton Period would indicate a population density in northwest Florida, during this period, con- siderably greater than Santa Rosa-Swift Creek and at least as great as that of the Weeden Island Period. This probability is further strengthened if we add the temple-mound sites to the simple midden total. Most of the temple mounds are surrounded with big middens. Although these temple-mound centers may have been supported in part by outlying villages, they were also sustained by their own im- mediate inhabitants who, to judge from the amount of refuse around the mounds, were fairly numerous. Thus we are, I believe, on firm WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—WILLEY 469 footing when we say that the population of the Fort Walton Period in northwest Florida was the equivalent of Weeden Island for the same area. Weeden Island population for the whole Gulf Coast has been estimated at from 15,000 to 10,000 persons per period (Weeden Island I and Weeden Island II). For the northwest coast alone we might halve this, or 7,500 to 5,000. There is, however, another important consideration in this prob- lem of Fort Walton Period population estimates. Our surveys and site reviews have been largely confined to the immediate coast and the major rivers. In the Fort Walton Period we note, though, that some of the biggest sites are located inland on lakes or small streams. This implies proportionately greater population concentrations in the interior than on the coast, and tends to raise the population estimates for the Fort Walton culture area as a whole. Mooney and Kroeber (see Kroeber, 1939, p. 138) have estimated a total population of 12,000 persons for the combined Apalachee, Apalachicola, Sawokli, Chatot, and Pensacola tribes. These were the tribal groups occupy- ing the Fort Walton culture area in the sixteenth and seventeenth centuries. The Mooney-Kroeber figure is considerably in excess of the 7,500 to 5,000 computed here; however, by giving additional weight to the big interior sites, our figures could reasonably be boosted to the 12,000 estimate. We have not in this discussion considered historic or ethnographic evidence. The Mooney-Kroeber estimates are based upon ethno- historic rather than archeological data, and it is interesting to note that the Narvaez and De Soto chronicles of the sixteenth century both indicate that the large population centers were inland rather than on the Gulf shore. For a consideration of the sixteenth-century population of northwest Florida (presumably identical with the Fort Walton culture and period) from the standpoint of history and ethnography see pages 533-535. The Fort Walton Period is estimated as being of relatively short duration, from A. D. 1500 to 1650. The terminal date is established upon the basis of the entrance and influence of the Spanish missions. After 1650 the Indian culture of the region was profoundly changed and enters the final or Leon-Jefferson Period. We feel quite sure that the Apalachee Indians encountered by De Soto on his journey possessed the Fort Walton rather than the earlier Weeden Island culture. The trade goods of the Fort Walton Period graves are those of sixteenth-century Europeans. It will be remembered that a Span- ish-Mexican coin dating between 1521 and 1550 was found in a Fort Walton cemetery at Bear Point (Ba-1), Ala. The date and the 470 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. II3 locality suggest that the Indians may have obtained the coin from the De Luna expedition of 1559. As to how far back of the Spanish entradas we should place the beginning of the Fort Walton Period is more perplexing. Fort Walton is a Middle Mississippian-influenced culture. Furthermore, it is late Middle Mississippian, having close links with such cultures as the Georgia Lamar. In view of this, I hesitate to push the beginning date back of the year A. D. 1500. It should be mentioned that the relative brevity of the period is indi- cated by depth in refuse deposits. Whereas Santa Rosa-Swift Creek and Weeden Island deposits were 1 to 1.40 meters in depth at the Fort Walton (Ok-6) midden site, Fort Walton Period types were confined to the upper 10 to 30 cm. This superficiality of Fort Walton Period refuse was also noted at other midden sites. THE ENGLEWOOD PERIOD Period definition—The diagnostic pottery types of the Engle- wood Period are Englewood Incised and Sarasota Incised. Asso- ciated types, such as Englewood Plain and Lemon Bay Incised, have little or no diagnostic value at our present stage of knowledge. Bis- cayne Series pottery and Glades Plain, also found in association, are not marker types for the period. In general, the period is little known and the evidence scant. The sites—The present review records only three sites at which sufficient materials of the Englewood Complex were found to enable us to designate them as stations of the period. All, including the type site, are mixed sites. Mixed sites Middens: Clearwater (Pi-5) Osprey (So-2) Burial mounds: Englewood (So-1) The Osprey and Englewood sites are in the Manatee region and the Clearwater site is near the southern margin of the central coast region. Occasional Englewood types are found at other Manatee and central Gulf Coast sites.7° To my knowledge, none are known from the northwest coast. 76 Moore illustrates very little Englewood pottery ; however, the vessel shown on page 430, figure 83 (Moore, 1903) is either Englewood Incised or Sarasota Incised. This specimen came from the Pithlochascootie River mound (Pa-2). The site dates as Weeden Island II. WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—WILLEY 471 Settlement pattern—Shell-midden villages with nearby burial mounds along the Gulf Coast and bay shores. Probably very com- parable to Weeden Island patterns. Economy.—Marine foods played an important part in the sub- sistence. By analogy with Weeden Island and Safety Harbor, it is presumed that agriculture was practiced. Organization of society—Probably similar to Weeden Island with autonomous villages, or small group of villages, each maintaining their own ceremonial mound for burial purposes. Disposal of the dead—The only known burial site is the sand burial mound at the type site. The burial mound at the Englewood site was one of two. It was 110 feet in diameter and 13 feet high. Conical in form, the tumulus was composed of sand, and its interior structure was revealed as a smaller interior mound over which the larger outer mound, or mound layer, had been heaped. Under the inner or primary mound was a shallow burial pit. This had been covered with sand mixed with red ocher before the construction of the primary mound. The absence of any sod line or weathered sur- face on the primary mound implies that little time had elapsed after its completion before the secondary mound was put down. Over 100 burials in the burial pit underlying the primary mound were bundle or single skull interments. Scattered at various depths in the body of the primary and secondary mounds were over 100 others. Most of these were bundle burials, although a few primary flexed burials were among them. The bulk of the pottery from the Englewood mound was in a great mortuary deposit on the mound base. This pottery had either been broken or intentionally perforated. Ceramic arts—Englewood pottery seems to offer a partial tran- sition between Weeden Island II and Safety Harbor. The Engle- wood mound (So-1) is a case in point for such a transition, con- taining, as it does, ceramics of all three complexes. Nevertheless, Englewood is sufficiently distinctive to be considered as a separate pottery group. On the whole, it shows less imagination and variety than Weeden Island and is less well done both as to form and decora- tion. As compared to Safety Harbor it is better made, fired, shaped, and decorated although there is probably less range of variation in decoration than in Safety Harbor. In addition to the characteristically Englewood types, others are noted, and some of these, undoubtedly, were also made by the Engle- wood potters. The Biscayne Series types are the most important of these. Glades Series types may also have been made locally as a result of Glades area influence. 472 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. II3 The Englewood pottery types in the order in which they are de- scribed : Englewood Series: Englewood Incised Englewood Plain Sarasota Incised Lemon Bay Incised Type Descriptions Type name-—ENGLEWOOD INCISED. Definition as a type: This paper, from the Englewood site, Sarasota County. Ware characteristics: Method of manufacture: Apparently coiled. Temper: Fine sand. Paste texture and color: Fairly even and granular. Buff, gray, and black. Surface texture, color, and finish: Usually buff-colored. Hardness: 3 or harder. Thickness: Averages 6 to 7 mm. Decoration: Technique: Medium-deep incisions and punctations. Design: Usually rectilinear designs but with simple curvilinear elements sometimes combined. Interlocking rectilinear elements in which incised bands are alternately filled with teardrop-shaped punctations. Con- tinuous crisscross or diamond elements in which band zones are left plain and background filled with teardrop or dot punctations. Zigzag incised bands or connected chevrons in which background is filled with punctations and chevrons are left plain. Rectangular panels with diagonal incised bands and backgrounds filled with punctations. Con- tinuous curvilinear plain bands arranged S-fashion with punctated backgrounds. Vertical and diagonal bands alternately filled with punctations. (See fig. 61; pl. 46, a, b; pl. 47.) Distribution: Covers either upper shoulder or most of exterior walls. Plain band at rim usually set off from decoration by one or more horizontal incised lines. Form: Total vessel: Cylindrical beakers, short-collared jars, simple jars, deep bowls or beaker-bowls, open bowls. Rim: Direct and unmodified. May be outslanted or incurved. (See fig. 62.) Lip: Flat or flat-round. One example of ticking on outer edge. Base: Flat-circular most common. Appendages: None known. Geographical range of type: Known only from Sarasota County and im- mediately surrounding regions. Chronological position of type: Probably occupies a position between Weeden Island II and Safety Harbor Periods, to some extent overlapping with both. Relationships of type: Has resemblances to both Weeden Island Incised and Safety Harbor Incised. Also vague design resemblances to early types. Bibliography: Willey (1948a). WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—WILLEY 473 Fic. 61.—Englewood Period vessel. Englewood or Sarasota Incised. (Redrawn from Moore, 1903, fig. 83. Pithlochascootie. About 4 actual size.) Fic. 62.—Englewood Period rim forms. Englewood Incised. (Interiors to right.) 32 474 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 113 Type name-—ENGLEWOOD PLAIN. Definition as a type: From Englewood site, this paper. Ware characteristics: (See Englewood Incised.) Form: Boat-shaped bowls with small vertical rim projections at ends of long axis, simple bowls with slightly incurved rims, simple open bowls, and ollas or jars with short outflared collars. Rims are direct and unmodi- fied or have slight marginal thickening. (Pl. 46, d, e.) Range, period, and relationships: Parallels Englewood Incised. Bibliography: Willey (1948a). Type name -—SARASOTA INCISED. Definition as a type: From Englewood site, Manatee region, and Tampa Bay, this paper. Ware characteristics: This is a Biscayne or St. Johns type paste, soft and temperless. Core is gray-black and surfaces are buff. It is well smoothed, both in and out. Thickness averages 5 to 6 mm. Decoration: Technique: Medium to fine line incision and triangular-shaped or teardrop punctations made in soft vessel surface. Design: Bands arranged diagonally to vertical axis of vessel and filled with punctations (fig. 61), interlocking rectilinear bands and chevron arrangements alternating plain and punctate-filled, triangles pendant from rim filled with punctations, and series of triangles point-to-point or nested and filled with punctations are the principal motives. (Pl. 46, c; pl. 48, g, h.) e Distribution: Vessel exterior. Tendency to restrict to upper part of vessel. Form: Flattened-globular bowls, open bowls with slightly incurved rims, and pot forms. Rims unmodified. Lips flat to round-pointed. Geographical distribution of type: Manatee region. Also north to Tampa Bay and beyond as a minority type. Occasionally has turned up in St. Johns area. 5 Chronological position of type: Weeden Island II, Englewood, and Safety Harbor Periods. May have a little greater time span, extending both earlier and later than Englewood Incised. Relationships of type: Related to Biscayne-St. Johns and similar wares. Re- lated to Englewood Incised. May have some relationship with St. Johns incised and punctated types. Bibliography: Willey (1948a). Type name—LEMON BAY INCISED. Definition as a type: Manatee region, this paper. Ware characteristics: Sand-tempered ware averaging 3.5 to 4 on hardness scale. Well smoothed and often polished. Surface color from reddish buff to gray; core is sometimes gray-black. Decoration: Technique: Thin to medium wide and deep incised lines that have been made in vessel after clay had dried but before firing. Occasionally, lines look as though they are engraved, as they display fine fractured edges. Triangles and rectangles of paste are also cut out in champlévé fashion. (Pl. 48, c-f.) WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—-WILLEY 475 Design: Composed of series of parallel lines about 5 to Io mm. apart. May be arranged vertically, diagonally, or horizontally to rim. Ele- ments essentially rectilinear, although some curvilinear segments are noted on some sherds. Data for over-all design patterns are lacking. Form: Beakers. No rim modifications. Lip rounded. Geographical range of type: Rare type. Found in Manatee region in small quantities. Occasional sherds occur on coast farther north. Chronological position of type: Uncertain. Englewood Period and, possibly, Weeden Island II. Relationships of type: Uncertain. Possibly to Weeden Island types such as St. Petersburg Incised. Other arts and technologies——Data are very few. In the Engle- wood mound were a few perforated conch-shell cups, a single Busycon contrarium hoe, some flint scrap, and a few pieces of red ocher. Speculations on population and period duration——Data are too limited to make any population estimates for this period. Apparently, sites of the period are not numerous, although this impression might be due to insufficient field survey. As a period, I have placed Engle- wood intermediate between Weeden Island II and Safety Harbor in the Manatee region. No attempt has been made to indicate the duration of the period, except that it was brief, probably lasting for a few decades after A. D. 1500. It is also possible that it continued for a time coeval with Safety Harbor. THE SAFETY HARBOR PERIOD Period definition—The key types for the period are Safety Harbor Incised and Pinellas Incised. The Pinellas Series shows close rela- tionship to Fort Walton Incised and Lake Jackson Plain of the Fort Walton Period, and there are also connections with the Leon-Jefferson Period of the northwest. Occasional Pensacola Series sherds occur in Safety Harbor contexts. The sites—There are a total of 25 sites of the period. Sixteen of these are pure sites; 9 are mixed. Pure sites Temple mounds (and villages) Safety Harbor (Pi-2) Parrish mound 2 (Ma-2) Burial mounds: Safety Harbor (Pi-2) Johns Pass (Pi-4) Bayview (Pi-7) Seven Oaks (Pi-8) Snavely (Hi-5) Parrish mound rt (Ma-r) 476 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I13 Parrish mound 2 (Ma-2) Parrish mound 3 (Ma-3) Whittaker site (So-4) True site (So-5) Gasparilla Sound (Ch-2) Hickory Bluff (Ch-5) Arcadia site (De-1) Cemeteries : Buzzard’s Island (Ci-2) Mixed sites Middens: Boca Ciega Island (Pi-6) Rocky Point (Hi-7) Shaws Point (Ma-7) Burial mounds: Bayport (He-1) Thomas (Hi-1) Picknick (Hi-3) Jones (Hi-4) Buck Island (Hi-6) Cemeteries (and middens) : Pool Hammock (So-3) Safety Harbor sites are found largely in the Manatee region but also extend north into the central coastal region. The areal demarca- tion between Fort Walton and Safety Harbor is not determined, but, presumably, it must fall somewhere near the Aucilla River boundary which separates the central from the northwest coasts. Buzzard’s Island, the northernmost Safety Harbor site listed by this review, is atypical in that the marker type Safety Harbor Incised is not found and in that the burials are in a cemetery and not a mound. The diag- nostic pottery at the site is, however, Pinellas Incised. The disproportionately large number of burial mounds as com- pared with middens which the review shows is, undoubtedly, the re- sult of sampling rather than a reflection of a cultural situation. Syste- matic midden-heap survey has not been conducted along the central or Manatee coasts. Settlement pattern—Safety Harbor village sites are middens, com- posed largely of shell, which are situated along the coast, bays, and rivers. The ecological picture is much the same as that observed for Weeden Island. There is little information on the size of Safety Harbor Period middens. Some of the midden sites around Tampa Bay and to the south are extensive although many of these were also occupied in other periods and the extent of the Safety Harbor occupa- tion has not been determined. At Safety Harbor proper (Pi-2), a site which is reportedly of only the one period, the area of occupation is WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—-WILLEY 477 several hundred yards in range. This site is probably larger than most of those of the Safety Harbor Period, however. Burial mounds are found near some of the middens. In other in- stances burial mounds have been described as being isolated from any evidences of habitation, but it is likely that a search would have re- vealed nearby middens. Usually the burial mounds are located singly, although in a few cases two or more mounds were found close to- gether. These additional mounds may have been other burial tumuli or they may have been domiciliary mounds. The frequency and impor- tance of house or temple mounds during the Safety Harbor Period are not fully known. That they did exist, however, is certain. The best example is the combined burial and temple mound at Parrish mound 2 (Ma-2). Another, which is quite probably Safety Harbor Period, but has not been explored, is the big flat-topped mound at Safety Harbor proper. The temple or house mound is not, judging from our reviewed sample of Safety Harbor sites, as numerous in Safety Harbor as in Fort Walton. For the present, the Safety Harbor settlement-pattern data can best be summarized by saying that most communities were small. A village, or a little group of villages, maintained its own burial mound much as in the old Santa Rosa-Swift Creek or Weeden Island Periods. The temple-mound idea was known, and there is one large village site (Safety Harbor (Pi-2)) which has such a mound in addi- tion to a burial mound. This site seems to have been much bigger than other villages of the period, being more comparable to a Fort Walton Period site. At another site (Parrish mound 2 (Ma-2)) is a mound which seems to have functioned both as a burial tumulus and as a temple substructure. Economy.—Marine foods continued to be a major dietary item in this period. There is little doubt that the period was also agricultural. Organization of society.—The trend toward political and religious cohesion estimated for the Fort Walton Period could also be said to extend to Safety Harbor but to a less marked degree. The large sites with temple mounds were known, but the pattern was not as com- mon as it was farther to the north. Disposal of the dead.—Most Safety Harbor burials were made in burial mounds. Such mounds were built by the Safety Harbor people for this purpose; also mounds of the earlier Weeden Island Period were utilized a second time by the Safety Harbor Indians. In form and structure, the burial mounds of the period were much like those of Weeden Island, being circular in outline, conical or dome- shaped in form, and ranging from 2 to 10 feet in height. Mound di- 478 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 123 ameter varied from about 35 to 80 feet. Two mounds, Jones (Hi-4) and Parrish mound 3 (Ma-3) were partially surrounded with crescent- shaped sand embankments. Besides the mound burials, two cemetery sites are also reported. There is also one instance of burials in a mound which also served as a temple substructure (Parrish mound 2 (Ma-z2) ). Secondary burial was the most common form of treatment. These were bundle, single skull, and mass burials. Some of the burial mounds contained only secondary burials; others had mixtures of secondary and primary burials, the latter being both extended and flexed types. In two mounds only primary burials are described. In most mounds, cremation, either complete or partial, was very rare; however, in the Parrish mound 2 (Ma-2), the combined temple and burial mound, masses of semicremated human bones were found in a submound pit as well as in the body of the mound. Two in situ cremations of primary burials were discovered in this same site. At the only cemetery site from which burials are described, Buzzard’s Island (Ci-2), massed and scattered secondaries, and primary flexed and extended types are all listed. The numbers of burials in the mounds and cemeteries ranged from as few as 15 to over 200. Funerary artifacts consisted of pottery, nonceramic aboriginal artifacts, and articles of European provenience. Pottery vessels found in the mounds were usually “killed” by intentional perforation. In some cases they were arranged in special caches or deposits on mound base; in other instances they were placed with or inverted over indi- vidual skeletal remains. In most mounds, other artifacts, of shell or stone or of European manufacture, had been placed with individuals. Cranial deformation does not seem to have been practiced during this period. In general, we can say that Safety Harbor burial practices conform with Weeden Island more closely than do those of Fort Walton. Mound burial was customary with cemetery burial being rare. The best-known cemetery of the period is well to the north in Citrus County, approaching the Fort Walton region. Secondary treatment of the dead and the charnel house or bone-cleaning complex continued, and some primary inhumation was also followed. In the placement of a pottery vessel with or over the individual dead, Safety Harbor seems to be pursuing the same trend noted in Fort Walton. Ceramic arts—Safety Harbor pottery is generally poorly made, fired, and decorated. Shapes tend to be badly formed and designs vaguely conceived and executed with carelessness. Although there WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—-WILLEY 479 is quite a range of excellence or lack of excellence in Safety Harbor types, the best are usually below Weeden Island, Fort Walton, or Englewood standards and the worst are absurdly handled. The total feeling is one of break-down in the ceramic art, carrying with it the implications of an impoverishment of the cultural forces and traditions that served as an incentive and guide to the aboriginal pottery maker. Two streams of influence are clearly seen. The one, represented in the type Safety Harbor Incised, follows out of old Weeden Island vessel form and decorative concepts ; the other, seen best in the Pinel- las types, is an offshoot of the Fort Walton style traditions and the Middle Mississippian impetus behind them. A minority representation of Lamar, or Lamar-like, Complicated Stamped pottery shows con- nections with the late Georgia period. Biscayne types, particularly Biscayne Check Stamped, are also found in Safety Harbor Period sites and probably represent a con- tinuation from Weeden Island II through the Englewood Period. Glades types, which may also have been made locally, are associated with the Safety Harbor Complex; but, like the Biscayne ware, are not diagnostic markers for the period. The Safety Harbor types in the order in which they are described : Safety Harbor Series: Safety Harbor Incised Pinellas Plain Pinellas Incised Lamar Complicated Stamped Type Descriptions Type name-—SAFETY HARBOR INCISED. Definition as a type: From Tampa Bay and the Manatee region, this paper. Stirling (1936) gave this name to a general ware group, covering the entire Safety Harbor Series. Willey and Woodbury (1942, pp. 244- 245) presented a synoptic description in which both Safety Harbor Incised and what is now Pinellas Incised were covered. Ware characteristics: Method of manufacture: Probably coiled. Temper: Ranges from fine sand through medium-coarse sand to a clayey, possibly temperless paste. Paste texture and color: Depending upon aplastic, paste ranges from granular and compact to coarse, contorted and laminated. The latter is crumbly and friable. Although lacking in temper particles there is no resemblance to the smooth, soft, even Biscayne ware. Paste color is usually brown. Surface texture, color, and finish: Surface is rough and often crackles easily. Lumps of clay may extrude. Color is gray to brown to buff. Poorly smoothed on both surfaces with tooling marks in evidence. Hardness: 2.5 to 3. Thickness: 5 to 8 mm. e Fic. 63.—Safety Harbor Period vessels. All Safety Harbor Incised. (Drawn from photographs and sketches. Vessel c, from True Site; others from Arcadia site. Actual heights as follows: a, about 19 cm.; b, about 26 cm.; c, ?; d, about 26 cm.; e, ?; f, about 14 cm.) 480 WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—-WILLEY 481 Decoration: Technique: Incised line and dot punctations made in soft or wet surface of vessel before firing. Design: Crude and poorly drawn rectilinear and curvilinear designs which are essentially geometric. Volutes pendant from rim, concentric diamond elements (fig. 63, e), parallel lines in curvilinear formations, encircling wavy or zigzag bands (fig. 63, b, f), X-shaped elements with scroll ends (fig. 63, c), continuous intertwined bands, and filfot- cross elements (fig. 63, a). Punctations used as filler for designs or backgrounds. Punctations are also used to follow or outline incised line designs (fig. 63, c). Indentation or fluting and low-relief modeling are sometimes seen as decorative features (fig. 63, e, f). Portrayal of naturalistic designs is occasionally attempted but is always highly stylized. A serpent design, suggested by a highly conventionalized Fic. 64.—Incised and punctated design from a Safety Harbor Period vessel. (From Parrish mound 3. Compare with pl. 54A.) snake’s rattle (fig. 63, c), is observed on one vessel, and serpent or feather symbolism is seen on another (fig. 64). (See pl. 49; pl. 52, c-h; pl. 54A, a, b, e.) Distribution: Vessel exteriors. Usually neck and upper half of body. Some- times most of body. Form: Total vessel: Bowls with slightly incurved rims, flattened-globular bowls, deep bowls with curious recurved rim and flat base, beaker-bowls, short- collared jars, long-collared jars, bottles. Rim: Unmodified or slightly thickened. Lip: Quite often flat, sometimes rounded. Base: Usually rounded, but some are flat and circular. Appendages: None. Geographical range of type: Manatee region, Tampa Bay, and for an undis- closed distance north of Tampa Bay. Chronological position of type: Safety Harbor Period. 482 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. II3 Relationships of type: Descendant of Weeden Island Incised and, probably, Englewood Incised. Also has some relationship to Fort Walton in such features as bottle forms. Bibliography: Stirling (1936) ; Willey and Woodbury (1942) ; Willey (1948a). Type name—PINELLAS PLAIN. Definition as a type: Central Gulf Coast and Manatee regions, this paper. Ware characteristics: (See Safety Harbor Incised.) As a rule, however, tends to be harder, more compact than Safety Harbor Incised. Form: Total vessel: Large open bowls with slightly incurved rims, casuela bowls (fig. 65, a, b; fig. 66, d), collared globular ollas, and pot forms (fig. 67). Rim: Sometimes folded on exterior. Use of nodes and pinched punctations on or below rim (fig. 67). (See pl. 50, a-d, f; pl. 52, b.) Lip: Characteristically has deep to slight indentations, crimping, or tick- ing on exterior edge. Base: Probably rounded. Appendages: Small, vertically placed loop handles with nodes at top, crude efigy (?) handles with nodes, large, ovate horizontal rim projections. Geographical range of type: Manatee and central coast regions. May extend farther north than Safety Harbor Incised. Chronological position of type: Safety Harbor Period. Relationships of type: Closest relative is Lake Jackson Plain. Type name—PINELLAS INCISED. Definition as a type: From central coast and Manatee regions, this paper. Ware characteristics: (See Safety Harbor Incised.) Also tends to be harder, more compact than Safety Harbor Incised. Decoration: Technique: Medium to broad line incision and small to large dot puncta- tions made in soft or wet paste of vessel before firing. Design: Much more limited in range than Safety Harbor Incised. Single line arcade encircling vessel below rim with a row of heavy, often rectangular, punctations above is a common motif. Rectilinear- curvilinear guilloche or meander with background sometimes filled by punctations (fig. 65, b, c; fig. 66, b) (this motif is quite similar to Fort Walton Incised), parallel incised lines and volutes, and running scrolls or wavy bands (fig. 66, a, c) composed of two or three lines are other designs. In general, execution is slovenly. (Pl. 51; pl. 52, a; pl. 53, a.) Distribution: Almost always confined to upper exterior portion of vessel. Form: Total vessel: Big collared ollas or bowls, simple open bowls with slightly incurved or straight rims, and casuela bowls. Rim: Unmodified or slightly thickened. Lip: Indentations or pinchings on exterior rim edge. Base: Rounded. Appendages: Vertical loop handles with nodes at top. Geographical range of type: Central Gulf Coast and Manatee regions. Chronological position of type: Safety Harbor Period. Relationships of type: Closest relative is Fort Walton Incised. WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—WILLEY 483 Cc Fic. 65.—Safety Harbor Period rims and vessel forms. a, Pinellas Plain; b, c, Pinellas Incised. VOL. I13 SMITHSONIAN MISCELLANEOUS COLLECTIONS 484 “uleyd Se[[oulg ‘p : pastouy Seoul 9-D *SUIOF [OSSoA pue SUIT Ppollad Joqiey AjazeS—99 “OI Pp WD NM F7KIS 2 2 ug WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—-WILLEY 485 Type name-—LAMAR COMPLICATED STAMPED. Definition as a type: From central Georgia by Kelly (10938, pp. 46-47). De- scribed in detail by Jennings and Fairbanks (1939) and briefly by Willey (10939, p. 144). A similar type has been described from the Georgia Coast by Caldwell and Waring (1939: type Irene Filfot Fic. 67——Safety Harbor Period rim and vessel reconstruction, Pinellas Plain. Stamped). Brief description in this report is based upon these earlier descriptions and upon a few vessels and sherds from Tampa Bay and the Manatee region. Ware characteristics: A coiled, sand- and grit-tempered ware. Varying in hardness from 3 to 5. Thickness of body about 8 mm., with rims and bases thicker. Decoration: Impressions of carved-paddle or stamping unit. Designs are com- plicated and composed of curvilinear and rectilinear elements. The 486 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. II3 Florida specimens from the Parrish Mounds (Ma-1 and Ma-3) show a circle-and-dot and a complicated connected-rectangle design. (PI. 53, c; pl. 54A, c.) Form: Pot forms slightly constricted below orifice but with outflaring rims. Bases rounded. Rim is folded and then pinched or crimped. (Applies to Florida specimens under discussion.) Geographical range of type: Lamar Complicated Stamped is found throughout most of Georgia and into adjacent Alabama and South Carolina. Simi- lar types are seen in North Carolina and eastern Tennessee. In Florida it occurs as far south as Tampa Bay, but these occurrences may be only the result of trade. Eastern extension in Florida not well known but probably occurs there only as trade if at all. Chronological position of type: In Georgia it is the marker type for the Lamar Period in central Georgia, a period believed to be about contemporaneous with Fort Walton. On the Georgia coast Lamar Complicated Stamped or closely related types continue later than this. Florida Gulf Coast occurrences are in the Safety Harbor Period. Relationships of type: Most likely a Georgia import and not a Floridian de- velopment out of Swift Creek Complicated Stamped. In Georgia the type probably developed out of the earlier complicated stamped types such as Swift Creek and Napier. The Lamar influence in Florida may be responsible for the very late Jefferson stamped types of the Leon- Jefferson Period. Bibliography: Kelly (1938); Willey (1939) ; Jennings and Fairbanks (1939) ; Caldwell and Waring (1939). Other arts and technologies—The Safety Harbor peoples had no craft skills in which they excelled those of the other culture periods of the Gulf Coast. As stated, their ceramics were decidedly inferior to Weeden Island. Their techniques in stone chipping are about on a par with those of Weeden Island although they seem to have pro- duced fewer large, fine blades. Smaller projectile points of several varieties were made in quantity, however. Large and medium-size stemmed points vary from triangular to ovate-triangular in form and may or may not have shoulder barbs. Small, stemless triangular points, the common late period point in many parts of the Southeast, are frequent. Larger stemless points are also present as are a variety of knives, scrapers, and drills. (See pl. 54B; pl. 55; pl. 56; pl. 57, a-u; pl. 59, d, e.) Stone plummets are found in the mounds but not in great numbers (pl. 59, c). Stone celts, on the other hand, are either absent or extremely rare. Minerals are scarce, red ocher and lump kaolin being the only ones recorded. Shell tools and weapons are more common than in previous periods except for Perico Island. Busycon picks and hammers and conch cups and dippers were found in most of the Safety Harbor mounds. Plummet-type pendants and beads (pl. 59, g, h) were made of shell. WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—WILLEY 487 The scarcity of complicated stamped pottery suggests a possible de- cline in woodworkers’ arts, but in one of the Safety Harbor mounds a piece of charred, carved wood was found which, from the nature of its design, indicates the retention of carving skills (pl. 59, 7-1). No aboriginal smoking pipes are known from the burial mounds or mid- dens of the period, but one elbow-type pipe was found in the Buzzard’s Island (Ci-2) cemetery in Citrus County. Native metalwork with indigeneous metals is not recorded, but European trade metals had been reworked to form various ornaments (pl. 58, g). Imported European objects taken from wrecked Spanish ships, or traded south from early settlements, were common to many of the mounds. These include actual manufactures such as iron axes, swords, knives, guns, olive jars, glass seed beads, larger glass beads (pl. 58, ¢), glass pendants (pl. 58, f), copper and brass ornaments (pl. 59, b), and clay pipes (pl. 57, v-~). Bone or tortoise-shell combs (pl. 58, a, 0; pl. 59, a) are probably also of Spanish manufacture. Sheet silver rolled into tubular beads (pl. 58, d), sheet gold fashioned into cones (pl. 58, c), and miscellaneous sheet-metal ornaments are native re- workings of imported materials or objects. The abundance of Safety Harbor artifacts of chipped stone may partly be explained by the natural sources of flint which the Indians had near at hand in the vicinity of Tampa Bay. It may also partly be explained by a greater continued dependence upon hunting as a means of livelihood than was the case in the northwest with the Fort Walton culture. However, neither of these explanations are fully satisfactory as the previous Weeden Island Periods in the Tampa Bay section do not produce an abundance of chipped-flint work. The absence of ground-stone axes may simply be due to the absence of suitable stone this far south. In the northwest the In- dians were much closer to the sources of supply in the southern Ap- palachians. Stone celts were replaced in the south by the Busycon hammer and pick tool. This shell pick or hammer is a characteristic Glades area artifact and was common in the Manatee region during the Perico Island Period. Although not commonly used by the Weeden Island peoples it seems to have been taken up in the later Safety Harbor Period. European trade goods appear to have had a greater effect upon the Safety Harbor Indians than upon those of the Fort Walton Period. Foreign metals and glass beads replaced native craft ornaments to a very large extent, and the old smoking complex was continued almost entirely with European pipes. Speculations on population and period duration—Estimates on Safety Harbor Period population have been made much as those 488 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I13 for Santa Rosa-Swift Creek or Weeden Island. Our site review shows 18 burial mounds built or used by the Safety Harbor people. Multiplying this by 4 gives us a total of 72 as a possible maximum number of mounds. Computing three villages to one burial mound brings us to a total of 216 village communities. This multiplied by 30 persons per community totals about 6,000. This figure refers principally to the Tampa Bay district and the Manatee region, as very few Safety Harbor Period sites are known north of Pinellas County. The dating and duration of the Safety Harbor Period is conditioned by the dates assigned to the Fort Walton and Leon-Jefferson cultures. Archeological and ethnohistorical data indicate a range of from about A. D. 1500 to 1725. THE LEON-JEFFERSON PERIOD Period definition.—This period is just emerging as a result of very recent field work (see H. G. Smith, 1948). The diagnostic pottery types are Mission Red Filmed, Aucilla Incised, and a series of compli- cated stamped types referred to as Jefferson ware. Minority types indicate connections to other late periods in the Southeast. These are Ocmulgee Fields Incised, Lamar Complicated Stamped, and Alachua Cob-marked. There is continuity with the Fort Walton Period as seen in the presence of the types Fort Walton Incised and Leon Check Stamped. European trade items are also a diagnostic of the period. Spanish olive-jar fragments, glass, and various metal objects are common. The sites of the Leon-Jefferson Period were Spanish-Indian settle- ments operated by the Franciscan monks between 1650 and 1725.77 The sites—Three pure sites of the complex are known from the present review. Pure sites Middens : San Luis Mission (Le-4) Pine Tuft (Je-1) Scott Miller place (Je-2) All these sites are in northwest Florida in Leon and Jefferson Counties. 77H. G. Smith (1948) notes that in 1633 the Franciscans began to establish a chain of missions from St. Augustine to the Apalachicola River. They were in full operation by at least 1650. In 1704 the Apalachee mission Indians were virtually destroyed by Col. James Moore of South Carolina. It is likely that within 20 years after this date all the Apalachee were gone and the missions abandoned. WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—-WILLEY 489 Setilement pattern.—Settlements of this period were built around the nucleus of a Spanish mission, trading post, or fort. There are no outstanding aboriginal monuments such as mounds. The living com- munities of which we have knowledge are all located inland in fertile farm country. They are hilltop villages from which the natives worked the crops grown in the vicinity of each site. The average size of individual sites has not been determined fully, but the ones that I have observed appear to have included an area of perhaps 200 meters in diameter. Concerning one site, the Scott Miller place (Je-2), Smith (1948, p. 316) writes: The excavated mission had buildings constructed by the wattle and daub tech- nique with hewn and split log framework and roof construction. The wood- work was secured by wrought iron nails. The mission complex was made up of two buildings and a borrow-refuse pit. The larger of the two structures was composed of two rooms with a surrounding patio wall. This complex measured 58 x 39 feet. The smaller building was probably an aboriginal abode, a stable, or some other minor building. This building measured 19 x 16 feet and did not have heavy construction or very high wattle and daub walls. Presumably most of this building was thatched. Economy.—Agriculture was probably practiced in the old way but under the eyes of the priests and other Europeans. Organization of society—The mission church undoubtedly was the focal point of the new society of this period. In a way, this provided a continuity, to which the Indians were undoubtedly accustomed, of secular and sacred control being held in the hands of the same authori- ties. The old war patterns were probably extinguished or suppressed. The easy defeat of these mission Indians at the hands of the Creeks under Moore may, in part, be attributed to this change of habits. Disposal of the dead.—No data are available. One would surmise, however, that inhumation as primary extended burials in Christian cemeteries was the custom. Ceramic arts—Pottery making continued in this period, although the vigor and skill of Weeden Island, Fort Walton, or even Safety Harbor was lost. Technical standards of manufacture and firing seem to have suffered less than those of form or decoration. This is, of course, to be expected, as the latter were features more intimately tied up with aboriginal tribal lore and religion. Nevertheless, the decorative techniques and vessel forms that persist in Leon-Jefferson have roots in earlier periods and can be traced back to these. Incised pottery shows some similarity to Fort Walton types, with Fort Walton Incised continuing in small percentages. Plain pottery also shows a 33 490 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I13 relationship to the Fort Walton Period, and check stamped and com- plicated stamped wares are known. One interesting aspect of the ceramics of the period is in the effects of European pottery styles. This is seen, for example, in the ring- base plate form. The Leon-Jefferson pottery types have been described briefly by Smith (1948), and the present descriptions, with a few noted excep- tions, are based upon his work. The principal Leon-Jefferson types in the order in which they are described are: Mission Red Filmed Miller Plain Aucilla Incised Leon Check Stamped Jefferson ware In addition there are a number of types which appear to be trade influences, either from Georgia or from the east. These are, in order of their description: Lamar Bold Incised Ocmulgee Fields Incised Alachua Series: 78 Alachua Plain Alachua Cob-marked Prairie Cord-marked The type Fort Walton Incised which occurs in this complex has been described with the types of the Fort Walton Period. Smith (1948, p. 316) also lists a Gritty Plain ware which he does not describe. Presumably this is a type comparable to Residual Plain. Type Descriptions Type name—MISSION RED FILMED. Definition as a type: From Jefferson County, by H. G. Smith (1948). Description: Apparently a hard, grit-tempered ware with smooth, highly polished surfaces. Plate forms have interior decoration in red-painted zones, and cups and small globular jars are completely red-slipped. These red surfaces are fired and highly burnished. The plate with an annular ring base is the most common form. At the present, the type is known for Leon and Jefferson Counties and dates from the Leon-Jefferson Period. The type is related to other historic period types such as Kasita Red Filmed (Fairbanks, 1940) and red filmed types of the St. Augustine Period (Smith, 1948). Bibliography: Smith (1948); Fairbanks (1940). 78 Smith (1948, p. 316) lists only Alachua Cob-marked as appearing in Leon- Jefferson sites. Goggin (1948a) considers the other two Alachua types to be approximately coeval, at least in part. WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—-WILLEY 491 Type name—MILLER PLAIN. Definition as a type: From Jefferson County, by H. G. Smith (1948). Description: A sand- and grit-tempered ware with a compact paste and hard surfaces. Vessel forms are similar to those of the Ocmulgee Fields Complex of Georgia (Kelly, 1938, pp. 56-57; Jennings and Fairbanks, 1939; Fairbanks, 1940). The most common shape is a shallow bowl with incurved, straight, or flaring rim; flat or rounded lip; and a rounded, flat or annular base. Other forms are the plate, bottle, casuela bowl, lugged shallow bowl, and European-influenced pitcher shape. The type is known for Leon and Jefferson Counties and dates from the Leon-Jefferson Period. Its affiliations are with the related types of the complex and with the plain wares of the Ocmulgee Fields Complex. Bibliography: Smith (1948) ; Kelly (1938, pp. 56-57) ; Jennings and Fairbanks (1939; type Ocmulgee Fields Incised) ; Fairbanks (1940; various types of Ocmulgee Fields Complex). Type name —AUCILLA INCISED. Definition as a type: From Jefferson County, by H. G. Smith (1948). Description: Ware is apparently much the same as Miller Plain. The decora- tion is composed of two to five parallel incised lines, which may or may not dip to form a loop design encircling rim of vessel; chevron designs and various other rectilinear and curvilinear designs, which may or may not have the background filled by punctates, also occur. Punc- tates are usually placed below the lip. Shallow-bowl and casuela forms are represented. Rims are usually incurved with lip flat or rounded. Lip lugs noted. (See fig. 68, a-d.) Leon and Jefferson Counties of northwest Florida. Type dates in the Leon-Jefferson Period. Has close resemblances to Ocmulgee Fields Incised (Jennings and Fairbanks, 1939); and to Lamar Bold Incised (Jennings and Fairbanks, 1939); and to Pinellas Incised. Bibliography: Smith (1948) ; Jennings and Fairbanks (1939). Type name—LEON CHECK STAMPED. Definition as a type: Northwest Florida, this paper. Ware characteristics: Sand and coarse-grit temper. Black paste core and buff surfaces. Thickness ranges from 8 to 10 mm. Decoration: Technique: Impression of checked or grid-bar stamping medium into soft surface of vessel (fig. 68 k). Design: Individual checks average about 1 cm. in diameter. The lands are sometimes low and faint, sometimes wide and heavy. Checks are diamond-shaped on some specimens. Some specimens appear to have been slightly smoothed after stamping which tends to blur and obscure the lands. Sometimes fingernail punctations occur on collar. (PI. 60, a-c.) Distribution: Vessel exterior. Extent unknown. Form: Outflared rims, round-pointed lips. Geographical range of type: Through northwest Florida but centering in Leon and Jefferson Counties. Occasionally found in central Gulf Coast region. 492 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I13 Chronological position of type: Leon-Jefferson Period but also found in Fort Walton and Safety Harbor contexts as a minority type. Relationships of type: Probably a development out of Wakulla Check Stamped. Bibliography: Mentioned by Smith (1948). is : , D \ B as ALY FLL WW G HU Vf a Iw Ret +! j ia ) ISS \ (ae WPUy) + SS rel DLE” SASS Wr (ORS ILE PE I LL ee LIP Fic. 68.—Leon-Jefferson Period and associated sherd types. a-d, Aucilla Incised; e, Fort Walton Incised; f, Ocmulgee Fields Incised; g-7, l-n, Jefferson ware; k, Leon Check Stamped; o, Alachua Cob-marked. From northwest Florida. (After Smith, 1948, pl. 32.) Type name-—JEFFERSON WARE. Definition: Not a type in the usual sense but a series of types as yet undif- ferentiated into separate type divisions. Defined by H. G. Smith (1948) from Jefferson County. Ware characteristics: Paste compact with laminations occurring along vessel wall in some cases. Temper is grit and sand as a rule but occasionally coarse quartz or crushed sherds. Surfaces are smoothed. Core is gray, brown, or black. Hardness ranges from 3.5 to 5. WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—-WILLEY 493 Decoration: There are one plain and four complicated stamped types within the ware group. These stamped types include the following motifs: (1) concentric rectilinear figures; (2) concentric circles with raised- dot centers; (3) a pattern of triangles and circles; and (4) herringbone figures. The plain type is characterized by a pinched or punctated rim which has sometimes been folded over before pinching treatment. (See fig. 68, g-j, l-n; pl. 60, d-f.) General relationships: The complicated stamped types have strong similarities to the Lamar Complicated Stamped type of Georgia (Jennings and Fairbanks, 1939). There are also similarities to late Swift Creek Complicated Stamped and to the other complicated stamped types of the Swift Creek and Weeden Island Periods. The reappearance of a rather degenerate complicated stamped series in the Leon-Jefferson Period is something of a puzzle in terms of sequence continuity. For the most part, complicated stamped dies out in Gulf Florida, especially northwest Florida, during the Weeden Island II Period and is virtually absent in the Fort Walton Period. Complicated stamped pottery in the Leon-Jefferson Period would not, in view of this break in the se- quence continuity, appear to be a local development out of Florida Late Swift Creek. It seems much more likely, in the light of present evidence, to have been a secondary introduction of the complicated stamping idea into Florida from Georgia at the end or close of the Fort Walton Period and the beginning of the Leon-Jefferson Period. The plain ware of the Jefferson Series, represented by the pinched or punctated rims, has its closest parallels with Lake Jackson Plain of the Fort Walton Period, out of which it probably developed, and with Pinellas Plain of the Safety Harbor Period, which may have exerted an influence upon it during their parallel growth. Bibliography: Smith (1948). Type name—LAMAR BOLD INCISED. Definition as a type: From central Georgia, by Kelly (1938, pp. 47-48) ; and by Jennings and Fairbanks (1939). Description: A grit-tempered ware decorated with broad incised lines, hollow- reed punctations, and, rarely, dot punctations. Curvilinear designs (scrolls) usually combined with rectilinear elements, particularly hori- zontal lines between scrolls. Some rectilinear elements stand alone. A row of hollow-reed punctates is usually placed at base of the incised decoration which is a border around upper portion of the vessel. The forms of casuela bowls, on which the incised designs occur, are often covered on the base with Lamar-type complicated stamping. Geographical distribution of type: Georgia, eastern Alabama, and parts of the Carolinas. In northwest Florida only as occasional pieces. Chronological position of type: The Lamar Period in Georgia, particularly the latter part of the period. In northwest Florida it occurs in the Leon- Jefferson Period. Relationships of type: Is related to Fort Walton Incised. May be ancestral, in part, to Ocmulgee Fields Incised and Aucilla Incised. Bibliography: Kelly (1938, pp. 47-48); Jennings and Fairbanks (1939) ; Willey (1939). 494 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I13 Type name—OCMULGEE FIELDS INCISED. Definition as a type: By Kelly (1938, p. 56), from central Georgia, and by Jennings and Fairbanks (1939). Description: A grit-tempered ware in which, very rarely, crushed shell is used. Is orange-buff to light brown or brown on surfaces with darker paste core. Smoothed surfaces show tooling marks, and there is a possibility of a light clay wash. Decorated with narrow incised lines which appear to have been partially smoothed over. Designs are scrolls, guilloches, combined scrolls and straight lines, chevron elements, and horizontal lines parallel to lip. Decoration confined to rim area of casuela forms or the upper and interior surfaces of flaring rims on open bowls. Rims insloped or incurved on casuelas, outflared on open bowls. Lips rounded or flat-round and generally thickened on exterior edge by slight protruberance. (See fig. 68, f.) Geographical range of type: Jennings and Fairbanks (1939) report type for most of Georgia. Its Florida diffusion seems rather limited, being mainly in north or northwest near Georgia boundaries. Chronological position of type: Dates from Ocmulgee Fields Period at Macon (ca. 1650-1750 and later). In Leon and Jefferson Counties, Fla., it is a minority type of the Leon-Jefferson Period. This is approximately coeval with the Georgia dating. Relationships of type: Related to Aucilla Incised which is contemporaneous. Probably a development out of Lamar Bold Incised. Bibliography: Kelly (1938, p. 58); Jennings and Fairbanks (1939). Type name —ALACHDUA PLAIN. Definition as a type: By Goggin from vicinity of Gainesville in central Florida (Goggin, 1948a). This description follows his. Ware characteristics: Coiled ware. Tempered with medium-grain quartz sand. Fairly uniform paste texture but with some lamination. Gray, buff, or dark brown color range. Poorly smoothed surface. Form: Simple bowls with unmodified rims. Geographical range of type: Around Gainesville. Chronological position of type: May occur as early as Weeden Island and Fort Walton Periods, extending into Leon-Jefferson. Type name -—ALACHUA COB-MARKED. Description: After Goggin (1948a). An Alachua Series ware. Surface is poorly smoothed and marked with corncob. Pattern of corncob stamp- ing may be rows of parallel marks (in pairs) up to I.5 cm. apart, or they may be so close together as to be overlapping, or the marking may be haphazard. Treatment applied to all parts of vessel exterior. Found in vicinity of Gainesville and west to Jefferson County. A very late type, contemporaneous with Leon-Jefferson Period, but may occur earlier. (See fig. 68, 0.) Type name-—PRAIRIE CORD-MARKED. Description: After Goggin (1948a). Ware same as Alachua Plain. Cord impressions usually formed by rows of medium to small cords impressed with a cord-wrapped paddle (?). They may be parallel, crisscrossed, WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST-——-WILLEY 495 or haphazardly applied. Apparently applied over all of vessel exterior. Simple bowl forms and globular bowls. Rims are unmodified. Found in vicinity of Gainesville. Probably coeval with Alachua Cob-marked and also may be somewhat earlier. Other arts and technologies——Native crafts of the missionized In- dian culture of the Leon-Jefferson Period, in addition to pottery making, are reflected in the following types of artifacts which Smith (1948, p. 316) reports from the Scott Miller site (Je-3) : Small and large triangular projectile points and large notched points; chipped- stone scrapers with blunt scraping edge; discoidal stones, which he says are not “chunkee”’ stones ; grinding or smoothing stones ; hones ; corn pounders; and mauls. This rather small inventory reveals that some of the former pursuits, such as hunting and agriculture, were carried on in the ways of the past. Significantly, ceremonial objects of stone, shell, wood, mica, or native metals are absent. Smith lists such European ornamental or ceremonial items as a metal ring and a crucifix. Other Spanish artifacts, but of a utility nature, were olive jars, glazed pottery, glass bottles, a spur rowel, a flintlock striker, guns, an ax, and various household items. (See pl. 60, g-1, for com- parable items. ) Speculations on population and period duration—Estimates for the Leon-Jefferson period are much less speculative than those of the other culture periods. The archeological manifestations are well identified with the Spanish missions established among the Apalachee Indians. For the mid-seventeenth century, when the missions were at their peak, Swanton has estimated a total population of 5,000 In- dians (Swanton, 1946, p. 91). By 1704, the year in which most of the missions were destroyed, this figure had dropped to 2,000. During the intervening period the Indians were clustered about nine mission stations (Swanton, 1922, p. 110). Period duration for the Leon-Jefferson culture is relatively brief and accurate. We know that the missions were introduced into the Apalachee country in 1633 and largely destroyed in 1704. Allowing a few years for the transition from the essentially aboriginal Fort Walton Period culture to the European-Indian Leon-Jefferson cul- ture, and a few more for the final subsidence of Leon-Jefferson, we have set the beginning date at 1650 and the terminal date at 1725. APPENDIX A: CLASSIFICATION OF FLORIDA GULF COAST VESSEL FORMS Vessels are described here under a series of descriptive headings and organized under various categories. There are three main cate- gories: (1) basic forms; (2) composite forms; and (3) effigy forms. The basic forms include the bowl, jar, beaker, beaker-bowl, bottle, and pot. Under each of these are a number of subdivisions. Com- posite forms are those which, in various ways, combine the basic forms. These are defined as compartment vessels, multiple-orifice vessels, composite-silhouette vessels, and pedestal vessels. Under each are subdivisions. The effigy forms duplicate the basic and com- posite forms in most cases, but they are treated separately in order to emphasize their effigy quality. They include effigy-affixation, semi- effigy, complete-effigy, and lobed forms. Each is, in turn, subdivided. These are the descriptive types, for vessel forms, used throughout the report. They apply to the ceramics of all periods. The classifica- tion is based principally upon the C. B. Moore collections. Basic Forms Bowls: Vessels whose maximum diameter is greater than their width but which have appreciable side walls. Orifice may be open or constricted. Bot- toms are usually rounded but may be slightly flattened at the center of base. Simple bowl—A medium-deep bowl with maximum diameter at or near orifice. Walls are vertical or slightly incurved. Bottoms rounded or flat-round. Size range Io to 22 cm. in diameter with average at 14 cm. Height approximately 3/5 of diameter. Common. (Fig. 69, a.) Globular bowl with flared orifice—A deep bowl with globular body and outslanted or outflared upper walls. There is considerable variation in the upper wall segment. Bottoms are rounded or flat-rounded. Size range 12 to 21 cm. in diameter with average about 16 cm. Height approximately 4 to 2 of diameter. Moderately common. (Fig. 69, b.) Shallow bowl or dish—A very shallow bowl with outslanted or slightly outcurved sides. Bottoms are flat-rounded. Size range 33 to 16.5 cm. in diameter. Height approximately 4 or less of diameter. Rare. (Fig. 69, c.) Shallow bowl with lateral expansions—A shallow bowl with maximum diameter from tip to tip of lateral expansions around orifice. Midwalls curve outward. Lateral expansions or projections may turn slightly upward. Bottom flat-rounded. Size range 17 to 35 cm., with average 25 cm. Height 4 of diameter or less. Common. (Fig. 60, d.) Flattened-globular bowl—A medium-deep to deep bowl with maximum diameter at about midpoint of vessel and with inturned sides and 496 _ = .- —. ; Fic. 69.—Florida Gulf Coast vessel form classification. Bowls. a, simple; b, globular bowl with flared orifice; c, shallow bowl or dish; d, shallow bowl with lateral expansion; e, f, flattened-globular bowl; g, casuela bowl; h, squared flattened-globular bowl; 7, collared globular bowl; 7, boat-shaped bowl. 497 498 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I13 constricted orifice. Walls curve inward with varying degrees of abrupt- ness. Some bowls of this class have sharp-angled shoulder. Orifice modified, if at all, only by rim thickening or folding, not by collar or neck. Bottom usually rounded or flat-rounded, occasionally flat at center of base. Size range 9 to 35 cm. in diameter, with average about 18 cm. (Only a few specimens are over 25 cm. in diameter). Height varies from 4 to almost the equivalent of the diameter. Tends to grade into simple jar form. Common. (Fig. 69, e, f.) Squared flattened-globular bowl—Very similar to flattened-globular bowl form except for the distinguishing characteristic of squared base and body. The portion of the bowl above the point of maximum diameter is rarely as definitely squared as the base or lower part of the body although the corners may be marked. Bottom is flat. Size range 13 to 20 cm. in diameter, with average 16 cm. Height from 4 to ? of diameter. Moderately common. (Fig. 60, h.) Casuela bowl—A medium-deep bowl with maximum diameter somewhat above the midpoint. Upper walls are inturned less markedly than in the case of the flattened-globular bowl. In some cases there is a slight inset or indentation of the upper walls above the shoulder. The orifice is slightly constricted, but is, proportionately, much larger than in the case of the flattened-globular bowl. Bottom rounded or flat-rounded. Size range 13 to 48 cm. in diameter with average at 30 cm. Height 4 to 2 of diameter. Common. (Fig. 69, g.) Collared globular bowl—A deep globular bowl with incurving walls which constrict at the base of a short neck or collar. The collar may be out- turned, inslanted, or straight. The maximum vessel diameter is at midpoint on body. Orifice varies from almost the equivalent of maxi- mum diameter to 4 maximum diameter. Bottom rounded, flat-rounded, or occasionally flattened at center of base. Size range 9 to 26 cm. in diameter, with average of 18 cm. Height from 4 to equivalent of diameter. Moderately common. (Fig. 69, 1.) Boat-shaped bowl—Medium-deep or shallow bowl which is oval or ovate- rectangular in form. Vessel walls usually outslanted or straight. Bases are usually flat-rounded. Both diameters considerably greater than height. Moderately common. (Fig. 69, 7.) Jars: Vessels whose total height is greater than maximum diameter and whose walls curve inward to a constricted orifice or to a constriction at the base of a collar or neck. They may or may not have a collar. Bottoms are rounded, flat-rounded, or flat. Simple jar—A jar without a collar. Maximum diameter may be at, above, or below the midpoint of the vessel. Walls incurve or inslant to a constricted orifice. There is considerable variation in the proportional size of the orifice. Bottoms are round, flat-rounded, or flat. Size range, 7 to 22 cm. in diameter, with average 13 cm. Height greater than diameter by ratio of 6-5 to 2-1. Common. (Fig. 70, a.) Short-collared jar—A jar with a short collar. Collar may be straight or outflaring so that the diameter of the orifice equals that of the maximum diameter of the body. Maximum diameter of body is usually about at midpoint. Collars are 4 to 4 of the total vessel height. Bottoms are round, flat-round, and flat. Size range 11 to 16 cm. in diameter, with BF tei Og G A aa) a RE eter | 20 CM. ay mM a) Oo Fic. 70.—Florida Gulf Coast vessel form classification, Jars, bottles, com- posite forms, effigy forms. a, simple jar; b, short-collared jar; c, squared short- collared jar; d, long-collared jar; e, pedestal vessel; f, simple jar with effigy affixed; g, h, simple jar with lobes; i, short-collared jar with lobes ; 7, pedestal jar with lobes; &, bottle; J, multiple-compartment tray; m, double bowl; 1, double beaker; 0, double bowl with effigy affixed. 499 500 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I13 average 13 cm. Height greater than diameter by ratio of 4-3 to 5-4. Common. (Fig. 70, 6.) Squared short-collared jar—A jar with body-collar proportions similar to the short-collared jar. Body of jar squared. Inturn at shoulder sharp with upper walls converging into base of collar. Walls below shoulder usually slope slightly inward to base. Maximum diameter is usually a little above midpoint of body of jar. Bottoms flat with squarish base. Size range 8 to 16 cm. in diameter with average of about 11 cm. Height greater than diameter by ratio of 5-4 to 5-3. Moderately common. (Fig. 70, c.) Long-collared jar—Jar with globular or flattened-globular body and walls incurving to base of collar. Maximum diameter of body at midpoint. Collar is long and straight or slightly outslanting. Occasionally diameter at orifice is equal to maximum diameter of body. Length of neck 4 to + of total vessel height. Bottoms are round and flat-rounded. Size range 5 to 22 cm. in diameter, with average about 12 to 15 cm. Height greater than diameter by 3-2 ratio. Common. (Fig. 70, d.) Beakers: Vessels whose total height is considerably greater than their maximum diameter. Walls are straight, or very nearly so, and are vertical, out- slanting, or a trifle inslanted. Bottoms are flattened with a definite angle between walls and base. Cylindrical beaker—A beaker that is cylindrical or round. Diameter varies but little from base to orifice. Size range 7 to 15 cm. in diameter, with average about 9 cm. Height greater than diameter by ratio of 2-1 to 1.5-I. Moderately common. (Fig. 71, 1.) Squared beaker.—Similar to round or cylindrical beaker except that vessel is square or rectangular at the base. Squareness with pronounced corners may, or may not, extend to rim. Walls are straight or outslant- ing. Size range 10 to 17 cm. in diameter, with average 13 to 14 cm. Height greater than diameter by ratio of 2-1 to 1.5-1. Rare. (Fig. 71, h.) Beaker-bowls: Vessels whose maximum diameter is greater than total height and which have straight walls and flat bases with a definite angle be- tween walls and base. Cylindrical beaker-bowl.—A beaker-bowl with a round or cylindrical body. Walls may be vertical or outslanting. Size range 12 to 14 cm. in diameter. Diameter greater than height by ratio of about 7-5. Rare. (Fig. 71, 1.) Squared beaker-bowl—A beaker-bowl with a squared body. Walls vertical or outslanting. Size range 9 to 12 cm. in diameter. Diameter greater than height by ratio of about 6-4. Rare. (Fig. 71, k.) Bottles: Vessels with an elongated collar or neck and an orifice which is markedly smaller than the maximum diameter of the vessel body. The body of the vessel is usually globular. Bottoms flat or flat-round. Size range 9 to II cm. in diameter, with average 10 cm. Total height approximately equivalent to diameter. Common. (Fig. 70, k.) Pots: Vessels whose total height is greater than maximum diameter with conoidal or rounded bases and with straight walls which may converge or outflare very slightly at the orifice. Pot with slightly converged orifice—Near the mouth the walls of the pot contract very slightly, but not enough to give the effect of a shoulder. ngs ME Ge at m™ Y Fic. 71—Florida Gulf Coast vessel form classification. Effigy forms, lobed forms, beakers, beaker-bowls, pots. a, simple bowl with effigies affixed; b, d, gourd forms; c, flattened-globular bowl with effigies affixed; e, f, 9, flattened- globular bowl with lobes; h, squared beaker; i, cylindrical beaker; 7, squared beaker with effigy affixed; k, squared beaker-bowl; 1, cylindrical beaker-bowl ; m, pot with slightly converged orifice; , pot with slightly flared orifice. 501 502 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I13 Size range 12 to 20 cm. in diameter, with average about 16 cm. Height is greater than maximum diameter by ratio of from 2-1 to 8-7. Mod- erately common. (Fig. 71, m.) Pot with slightly flared orifice——Near the mouth the walls of the pot may contract slightly and then reflare or they may slant outward from base to orifice with a final outflare. Size range 9 to 25 cm. in diameter with an average at about 16 cm. Height is greater than maximum diameter by ratio of 3-2 to 6-5. Moderately common. (Fig. 71, n.) CoMPOoSsITE ForMS Compartment vessels: Vessels having two or more compartments or container sections. Multiple-compartment tray.—Relatively flat trays having three or more container sections or conjoined bowls. The sections or bowl units are medium-deep with hemispherical interiors. They are separated from each other by thick compartment walls. The bowl units measure from 6 to 14 cm. in diameter with the average nearer the lower figure. A cruciform arrangement of five compartments is common with the central bowl often a little larger than the other four and sometimes elevated above them. Divisions of three and four compartments may be in a T-shaped arrangement or in a straight line. Compartments are rectangular, triangular, and crescentic as well as circular. Over-all diameters of trays range from 15 to 27 cm. Moderately common. (Fig. 70, 1.) Double bowl—Double or conjoined hemispherical bowls. Bowls average about 8 to 9 cm. in diameter. Over-all length 15 to 27 cm. Rare. (Fig. 70, m.) Double beaker—Conjoined cylindrical beakers. Each unit about 4.5 cm. in diameter and 6.5 cm. in height. Rare. (Fig. 70, n.) Conjoined pot—Two pot forms joined together and partially separated by a partition which is open at the bottom. The vessel has small tetra- pod feet. It is about 10 cm. high and 18 cm. long. Rare. (Fig. 72, k.) Multiple-orifice vessels: Vessels having two or more openings. Triple- and quadruple-orifice vessels—A globular body with three (or four) spouts placed together on the top of the body. The spouts are short, wide-mouthed, and slightly flared. Diameter of vessel is about 20 cm. Rare. (Fig. 72, 7.) U-shaped double-orifice vessels—Two mouths or spouts which are joined together by a connecting section. 10 to 12 cm. high and 17 and 24 cm. long. Rare. (Fig. 72, 7.) Composite-silhouette vessels: Vessels which have an abrupt differentiation in outline as of a beaker and bowl or two bowls, etc., joined together, one above the other. Jar with cambered rim—Small jar or pot forms which constrict at a point about one-third of their total height below the rim and then recurve and again constrict at the orifice. Diameters 12 to 18 cm. Height somewhat greater. Rare. (Fig. 72, a.) = s a a “ss Gs ww é oY J k ; ee a) EY 4 Mm Fic. 72.—Florida Gulf Coast vessel form classification. Composite forms and effigy forms. a, jar with cambered rim; b, c, double-globed jar; d, single-globed jar; é, inset jar-bowl; f, jar with cambered rim and effigy affixed; g, double- globed jar in semieffigy form; h, single-globed jar in semieffigy form; i, quad- ruple-orifice vessel ; 7, U-shaped double-orifice vessel ; k, conjoined pot; J, human- effigy vessel; m, human-figurine vessel. 503 504 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I13 Double-globed jar—Jars which are characterized by two globular or flattened-globular sections, one above the other. These two globes may be contiguous or they may be separated by a section of the vessel of lesser diameter and parallel vertical walls. Some specimens may have cylindrical, flat-bottomed, beakerlike bases. Some have short collars. Size range 9 to 16 cm. in maximum diameter with an average of about 11 cm. Height greater than diameter by ratio of 2-1 to 8-7. Moderately common. (Fig. 72, b, c.) Single-globed jar—Jars which are characterized by a globelike upper por- tion and a narrower, cylindrical, flat-bottomed base. The orifice of the globe may or may not have a collar attachment. Size range of maxi- mum diameter from 15 to 23 cm. In most instances this diameter is about the equivalent of the total height. Common. (Fig. 72, d.) Inset jar-bowl—These vessels have a globular bowl-like base into which has been set an elongated cylindrical upper portion. They differ from long-collared jars in that the inset is quite clear, with the rim ofthe bowl, or lower portion, showing plainly on the outside of the cylindrical inset. Diameter 10 to 16 cm. Height slightly greater. Rare. (Fig. 72, €.) Pedestal vessels: Small jars with definite pedestal bases. Rare. (Fig. 70, é.) Erricy Forms Effigy affixation: These are vessels of various shapes which are distinguished by affixed effigy figures or adornos. Simple bowl with effigies affired—These bowls have what appear to be bird head-and-tail projections on the rim on opposite sides of the vessel. Size range 15 to 31 cm. in diameter. Moderately common. (Fig. 71, a.) Flattened-globular bowl with effigies affived—tThere are a number of varia- tions in the nature and placement of the effigies or adornos in this type. A number of the bowls have two bird heads placed on opposite sides of the vessel rim; some have only a single bird head on the rim; others have a bird head-and-tail arrangement on opposite sides of the vessel. In the case of the latter the head and tail usually project out from the vessel a little below the plane of the rim. A dog, a fish, and a frog(?) are also utilized as effigies, but the bird is by far the most common. Size range 9 to 37 cm. in all-over diameter. Common. (Fig. 71, c.) Collared globular bowl with effigies affixed—A frog bowl with head, tail, and legs shown as affixations or in relief on the four sides of the bowl. 14 cm. in diameter. Rare. Simple jar with effigies afixed—Effigies include unidentified birds, owls, ducks, and unidentified animals. In most instances heads are placed on the rim, although some of the effigy faces are very large and an entire side of the jar is used for portrayal. Size range 10 to 22 em. in diameter. Height is greater. (Fig. 70, f.) Squared beaker with effigies affixed—Has human figure affixed on one side of vessel. Diameter 14 cm. and height 18 cm. Rare (one specimen). (Fig. 71, j.) WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—-WILLEY 505 Double bowl with effigies affixed—A double bowl with a human effigy head placed at the junction of the rims. Diameter 16.5 cm. and height 11 cm. Rare. (Fig. 70, 0.) Jar with cambered rim and effigies afixed—A cambered jar with an animal head placed on the side of the vessel at junction of body and cambered rim. Diameter 13 cm. and height 13.5 cm. Rare. (Fig. 72, f.) Semiefigy: Vessels in which the form has been only partially modified and combined with adornos, modeling, and decorative techniques to give an effigy effect. Single-globed jar in semieffigy form—All are bird forms, including an un- identified bird with a small head and pointed beak, the owl, and the duck. Modeling often includes portrayal of head, back, and wings and clear delineation of the head with the rest of the body merely sug- gested. All vessels of this form type have flat, cylindrical, beaker bases. The upper portion of the jar varies from a slight expansion to a large bowl. Size range 15 to 33 cm. in diameter. Height may be slightly greater or less than diameter. Moderately common. (Fig. 72, h.) Double-globed jar in semiefigy form—These vessels approximate the double-globed contour described under the composite-silhouette vessels. Only two specimens were observed. One is a small jar in which the wings of the bird are suggested by the upper globe or bulge. The other is a large globular vessel with a completely modeled owl effigy nesting on the top of the globe. Respective diameters are 190 and 26 cm. Rare. (Fig. 72, g.) Complete effigy: Vessels in which the form has been largely conditioned by the effigy effect desired rather than limiting the effigy by the vessel specifications. Gourd form—Simpie and globular bowls with a single handle projecting out from the side of the vessel after the fashion of a gourd. Handles are sometimes halved or scooplike, sometimes cylindrical and solid. Over-all diameters range from 11 to 24 cm. Rare. (Fig. 71, b, d.) Human-effigy vessel—These are squat male figures of jar proportions. They are modeled in full-round. Head, facial features, arms, legs, and genitals are depicted in a semirealistic, semistylized fashion. Size range 6 to 19 cm. in diameter. Height is greater than diameter by ratio of 3-2. Moderately common. (Fig. 72, 1.) Human-figurine vessel—This vessel is a small, narrow cylinder attached to the back of a standing male figure. The figure is similar to a large, solid figurine. It is 9 cm. at the shoulders (point of maximum breadth) and 21 cm. high. Only one specimen observed. Rare. (Fig. 72, m.) Eccentric effigy vessel—These are curious shapes. Seashells, plants or fruits, and shoe forms are represented. Rare. Wedge-shaped vessel—These are flattened cones or wedge shapes, having an oblong orifice and a pointed base. They have an average maximum diameter of 10 cm. and an average height of about 15 cm. Rare. Lobed forms: Lobed vessels are those with bodies divided vertically into three or more bulbate sections. Flattened-globular bowl with lobes—Have four or more lobes and present a “melon” effect. Diameter range from 16 to 22 cm. Rare. (Fig. pie: |, g) 34 506 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. II3 Simple jar with lobes—Tall jar with three lobes located midway between base and orifice. Only one specimen observed. Height 22 cm. (Fig. 70, g, h.) Short-collared jar with lobes—Four- and three-lobed specimens observed. The four-lobed vessel has tetrapodal supports, one to a lobe. Diameter from 6 to 16 cm. Rare. (Fig. 70, i.) Pedestal jar with lobes—Single specimen. Diameter 12.5 cm. (Fig. 70, j.) MISCELLANEOUS Miniature vessels: Very small vessels, most of which are rather carelessly made. They conform, in general, to the forms observed in the larger vessels. The short-collared jar, single-globed jar, and pot forms are best rep- resented. Moderately common. APPENDIX B: TRAIT LIST OF FLORIDA GULF COAST CULTURES DEPTFORD Settlements : 1. Small villages on rivers and bays. 2. Shell refuse. Economy: 3. Hunting-fishing-gathering. Organization of society: 4. Probably small autonomous bands. 5. Formal politico-religious organization unlikely. Disposal of the dead: (Data very scanty, open to question. ) 6. Burial in villages. 7. Cremation. 8. No grave goods with individual burials. 9. Caches of destroyed pottery in burial area. Ceramics : 10. Deptford Linear Check Stamped. 11. Deptford Bold Check Stamped. 12. Deptford Simple Stamped. 13. St. Marks Plain. 14. St. Simons Plain. 15. Alexander Incised. Artifacts : 16. Large to medium-size triangular-bladed chipped projectile points, notched or stemmed. 17. Pebble hammers. 18. Whetstones. Perico IsLanp Settlements : 1. Large villages on marshy shores and offshore islands. 2. Shell refuse. 3. Shell embankments and mound arrangements. Economy: 4. Hunting-fishing-gathering. Organization of society: 5. Probably autonomous village units. 6. Formalized politico-religious functions with each village unit. Disposal of the dead: 7. Cemeteries. 8. Burial mounds. 9. Primary flexed burials. 10. Burial offerings absent or rare. Ceramics: 11. Glades Plain. 12. Miami Incised. 507 508 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. II3 13. Perico Plain. 14. Perico Incised. 15. Perico Linear Punctated. 16. Belle Glade Plain. 17. Okeechobee Plain. Artifacts : 18. Busycon hammers or picks. 19. Busycon cups or vessels. 20. Strombus celts. 21. Busycon celts. 22. Plummet-type shell pendants. 23. Shell tools. 24. Shell beads. 25. Bipointed bone projectile points. 26. Bone awls. 27. Bone daggers. SAntTA Rosa-SwirT CREEK Settlements : 1. Small villages on rivers and bays. 2. Shell refuse. 3. Burial mounds located near or on midden. Economy : 4. Hunting-fishing-gathering. 5. Maize agriculture implied. Organization of society: 6. Probably autonomous village units or small groups of villages. 7. Formalized politico-religious functions with each village unit or group unit. 8. Cult of the dead. 9. Social differentiation or stratification. to. Craft specialization. Disposal of the dead: 11. Burial mounds, both circular-conical and oblong-flattened. 12. Extensions on burial mounds. 13. Surrounding embankments with mounds (rare). 14. Mounds constructed intermittently over considerable period of time. 15. Subfloor burial pits. 16. Burial pits inclusive within mound. 17. Primary extended burials. 18. Primary flexed and semiflexed burials. 19. Secondary bundle burials. 20. Secondary single skull burials. 21. Full cremation (rare). 22. Partial in situ cremation. 23. Oyster and other shells with burials. 24. Nonceramic artifacts with individual burials. 25. Pottery with individual burials (rare). 26. Mass cache of ceremonially destroyed pottery. 27. Ceremonial destruction of nonceramic burial artifacts. 28. Cranial flattening (one case). WHOLE VOL. .ARCHEOLOGY OF FLORIDA GULF COAST——-WILLEY Ceramics: 20. 30. St. BO. 33. . Crooked River Complicated Stamped (Early Variety). . St. Andrews Complicated Stamped (Early Variety). . New River Complicated Stamped. . Gulf Check Stamped. Alligator Bayou Stamped. Basin Bayou Incised. Santa Rosa Stamped. Santa Rosa Punctated. Swift Creek Complicated Stamped (Early Variety). West Florida Cord-marked (Early Variety). Crystal River Incised. . Crystal River Zoned Red. . Crystal River Negative Painted. . Pierce Zoned Red. . Franklin Plain. Unusual vessel forms (see discussion). Artifacts : . Large, well-chipped ceremonial blades. . Projectile points, probably large and medium-size stemmed forms. . Polished stone celts. . Hones. . Smoothing stones. . Pebble hammers. . Stone beads. . Bar amulets. . Plummet-type pendants of stone. . Rock-crystal pendants. . Galena. . Mica. . Mica spear-point forms. . Kaolin. . Batonlike objects of kaolin. . Hematite. . Bitumen. . Busycon cups or vessels. . Busycon hammers or picks. . Chisels and adzes of conch columellae. . Strombus celts. . Plummet-type pendants of shell. . Shell gorgets. . Shell beads. . Bone projectile points. . Bone fishhooks. . Bone beads. . Bone gorgets. . Turtle-shell rattles. . Perforated canine teeth. . Shark teeth. . Cut animal jaws. . Monitor pipes of stone. . Monitor pipes of pottery. 599 510 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. II3 Tubular pottery pipes. . Equal-arm elbow pipes of pottery. . Flaring-bowl elbow pipes of pottery (?). . Elbow pipes of stone. . Copper-covered ear spools. . Copper bicymbal ear spools. . Copper ear spools plated with silver. . Copper ear spools plated with meteoric iron. . Conjoined copper tubes. Sheet-copper ornaments. WEEDEN ISLAND Settlements : I. 2. 3. Small villages on rivers and bays. Shell refuse. Burial mounds located near or on midden (possibility of domiciliary or temple mounds ?). Economy : 4. 5s Hunting-fishing-gathering. Maize agriculture implied. Organization of society: 6. ie 8. 9. 10. Probably autonomous village units or small groups of villages. Formalized politico-religious functions with each village unit or group unit. Cult of the dead. Social differentiation or stratification. Craft specialization. Disposal of the dead: rr: Pe: 13: 14. IS. 16. Gh, 18. TO. 20. 21. 22. 23) 24. 25. 26. 27. 28. 20. 30. ONE a7 Burial mounds, both circular-conical and oblong-flattened. Extensions on burial mounds (Weeden Island I only). Mounds constructed intermittently over considerable period of time. Small interior mound (rare, although record on this trait is not adequate). Subfloor burial pits. Log tombs on mound base (?). Burial pits inclusive within mound. Limerock slabs used within mound construction. Primary extended burials (rare). Primary flexed and semiflexed burials. Secondary bundle burials. Secondary single skull burials. Mass secondary burials. Full cremations (rare). Partial in situ cremations. Oyster and other shells with burials. Limerock slabs covering burials. Nonceramic artifacts with individual burials (uncommon). Nonceramic artifacts as general mortuary offerings. Pottery with individual burials (very rare). Mass cache of ceremonially destroyed pottery. Cranial flattening probably more common in Weeden Island II. WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—-WILLEY SII Ceramics : 33. Weeden Island Plain 34. Weeden Island Incised. 35. Weeden Island Punctated. 36. Weeden Island Zoned Red. 37. Carrabelle Incised. 38. Carrabelle Punctated. 39. Indian Pass Incised. 40. Keith Incised. 41. Tucker Ridge-pinched. 42. Hare Hammock Surface-indented. 43. Swift Creek Complicated Stamped (Late Variety). 44. Crooked River Complicated Stamped (Late Variety). 45. St. Andrews Complicated Stamped (Late Variety). 46. Tampa Complicated Stamped. 47. Sun City Complicated Stamped. 48. Old Bay Complicated Stamped. 49. Wakulla Check Stamped. 50. Thomas Simple Stamped. 51. West Florida Cord-marked (Late Variety). 52. Mound Field Net-marked. 53. Hillsborough Shell Stamped. 54. Ruskin Dentate Stamped. 55. Ruskin Linear Punctated. 56. St. Petersburg Incised. 57. Papys Bayou Plain. 58. Papys Bayou Incised. 59. Papys Bayou Punctated. 60. Papys Bayou Diagonal-incised. 61. Little Manatee Zoned Stamped. 62. Little Manatee Shell Stamped. 63. Little Manatee Complicated Stamped. 64. Biscayne (St. Johns) Plain. 65. Biscayne (St. Johns) Check Stamped. 66. Biscayne (Dunn’s Creek) Red. 67. Biscayne Roughened. 68. Biscayne Cord-marked. 69. Pasco Plain. 70. Pasco Check Stamped. 71. Pasco Red. 72. Gainesville Linear Punctated. 73. Residual Plain. 74. Smooth Plain. 75. Plain Red. (Types 43, 44, and 45 are principal markers for Weeden Island I; types 49, 65, 70 are principal markers for Weeden Island II. Types 64, 65, 66, 67, 68, 73, 74, 75 are also found in later periods.) Artifacts : 76. Large well-chipped lance points. 77. Medium-size stemmed, triangular-bladed points, some with deep notches or barbs or with fishtail bases 512 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I1I3 78. Medium-size leaf-shaped points with rounded stems. 79. Small stemless triangular points. 80. Knives, chisels, scrapers, drills, flint hammerstones. 81. Polished stone celts. 82. Smoothing stones. 83. Hones. 84. Pebble hammers. 85. Plummet-type pendants of stone. 86. Stone gorgets (rare). 87. Galena. 88. Galena beads (rare). 80. Mica. 90. Mica spear-point forms. o1. Hematite. 92. Bitumen. 93. Plumbago. 94. Perforated teeth (rare). g5. Cut animal jaws (rare). 96. Turtle-shell rattles (rare). 97. Busycon cups or vessels. 08. Plummet-type pendants of shell (rare). 99. Shell beads. 100. Chisels and punches of conch columellae. 101. Elbow pipes of stone. 102. Elbow pipes of pottery. 103. Bird-effigy pipe of pottery. 104. Copper ornaments (unidentifiable). Fort WALTON Settlements : Small villages on rivers and bays. Shell refuse. Large villages in the interior. Burial mounds located near or on village area (uncommon). Rectangular and circular flat-topped substructure mounds within vil- lage area. Economy: 6. Hunting-fishing-gathering. 7. Agriculture definitely known. Organization of society (supplemented by ethnohistory) : 8. Semiautonomous large villages with satellite villages. 9. Territorial federations with capital towns. 10. Formalized politico-religious functions with each large village and with capital towns on a higher level. 11. Importance of secular and military power. 12. Social differentiation and stratification. 13. Craft specialization. Disposal of the dead: 14. Burial mounds, circular-conical. 15. Cemeteries. Waco a Sy WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—-WILLEY 16. 17. 18. 19. 20. oT. 22. 23. 24. 25. 26. Ppp 28. 20. Burials intrusive into earlier burial mounds (Weeden Island). Burials intrusive into various levels of substructure mounds. Primary extended burials. Primary flexed and semiflexed burials. Secondary bundle burials. Mass secondary burials. Secondary single skull burials. Nonceramic artifacts with individual burials (common). Nonceramic artifacts as general mortuary offerings. Mass caches of ceremonially destroyed pottery. Pottery with individual burials (common). Pottery vessel inverted over skull. Urn burial (rare). Cranial deformation. Ceramics: 30. aT: Lake Jackson Plain. Fort Walton Incised. Point Washington Incised. Pensacola Plain. Pensacola Incised. . Pensacola Three-line Incised. . Pensacola Red. . Marsh Island Incised. Moundville Engraved. Artifacts : 39. Large well-chipped blades or lance points. . Medium-size triangular-bladed stemmed and barbed points. . Small stemless triangular points (?). Polished stone celts. . Polished stone celts, thin rectangulate forms. Hematite. . Limonite. . Busycon cups (rare). . Shell beads. Spike-form ear pins of shell. . Plummet-type pendants of shell. 50. Shell gorgets. 51. Shell chisels and other columellae tools. 52. Bone awls. 53. Bone fishhooks. 54. Bone beads. 55. Copper spear-form objects (aboriginal copper ?). 56. Pottery trowels or anvils (?). 57. European materials (rare). ENGLEWOOD Settlements : 1. Small villages on rivers and bays. 2. Shell refuse. 3. Burial mounds located near or on midden. S51 514 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. II3 Economy : 4. Hunting-fishing-gathering. 5. Maize agriculture implied. Organization of society: 6. Probably autonomous village units or small groups of villages. 7. Formalized politico-religious functions with each village unit or group unit. 8. Cult of the dead. 9. Social differentiation or stratification. 10. Craft specialization. Disposal of the dead: 11. Burial mounds, circular-conical. 12. Small interior mound. 13. Subfloor burial pit. 14. Burial pits inclusive within mound. 15. Primary flexed burials. 16. Secondary bundle burials. 17. Secondary single skull burials. 18. Mass cache of ceremonially destroyed pottery. Ceramics 19. Englewood Incised. 20. Englewood Plain. 21. Sarasota Incised. 22. Lemon Bay Incised. 23. Biscayne Series types in association. 24. Glades Series types in association. Artifacts : 25. Busycon cups. 26. Busycon hoe. 27. Red ocher. SAFETY HArBor Settlements : 1. Small villages on rivers and bays. 2. Shell refuse. 3. Burial mounds located near or on midden. 4. Rectangular flat-topped structure mounds within village area. Economy : 5. Hunting-fishing-gathering. 6. Maize agriculture. Organization of society: 7. Large villages with satellite smaller villages. 8. Territorial federations with capital towns (?). 9. Formalized politico-religious functions within each village or group of villages (degree of centralization probably not as great as Fort Walton). 10. Importance of secular and military power. 11. Cult of the dead. 12. Social differentiation and stratification. 13. Craft specialization. WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—WILLEY Disposal of the dead: 14. 15. 16. 17. 18. 10. 20. 21. 22. 23. 24. 25. 26. 27. 28. 20. 30. Burial mounds, circular-conical. Cemeteries. Burials intrusive into earlier burial mounds. Burials intrusive and inclusive in substructure mound. Subfloor burial pits (probably rare). Burial pits inclusive within mound. Primary extended burials. Primary flexed and semiflexed burials. Secondary bundle burials. Secondary single skulls burials. Mass secondary burials. Full cremations (rare, except at one site). Partial in situ cremations (rare, except at one site). Nonceramic artifacts with individual burials (common). Pottery with individual burials (common). Pottery vessel inverted over skull. Mass caches of ceremonially destroyed pottery. Ceramics: ta 32. 33. 34. 35. 36. Safety Harbor Incised. Pinellas Plain. Pinellas Incised. Lamar Complicated Stamped. Biscayne Series types associated. Glades Series types associated. Artifacts: 37- 38. 39. 40. . Plummet-type pendants of stone. . Red ocher. Large and medium-size stemmed and barbed points. Large stemless points. Small triangular stemless points. Knives, scrapers, drills. Kaolin. Busycon picks and hammers. . Busycon cups. . Plummet-type pendants of shell. . Shell beads. . European metal artifacts reworked. . European artifacts (common). LEON-JEFFERSON Settlements : I. 2. Small to large villages. Built around nucleus of Spanish mission-forts. Economy : 3. Hunting-fishing-gathering (probably diminished). 4. Agriculture. Organization of society: 5. 6. Village unit with politico-religious functions (mission and fort). Authority in hands of Europeans. 515 516 SMITHSONIAN MISCELLANEOUS COLLECTIONS Disposal of dead: (No data. Presumably Christian cemetery burial.) Ceramics: 14. vse Mission Red Filmed. . Miller Plain. . Aucilla Incised. . Leon Check Stamped. . Jefferson ware. Lamar Bold Incised. Ocmulgee Fields Incised. Alachua Series types associated. Continuity of some Fort Walton types. Artifacts : 16. 17 18. 19. 20. oi 22) 23. 24. 25. Large triangular projectile points. Large notched projectile points. Small triangular projectile points. Scrapers. Discoidal stones (not chunkees). Grinding or smoothing stones. Hones. Pounders. Mauls. European artifacts (common). VOL. 113 VII. A SUMMARY OF THE ETHNOHISTORY HISTORY OF EUROPEAN EXPLORATIONS The peninsula of Florida, owing to its proximity to the early Spanish bases established in the West Indies, was the first area of native North America to be subjected to European exploration and conquest. Many of these adventurous forays struck the Gulf Coast, and because of this the Indians were aroused against the white in- vaders very early in the sixteenth century and fought with ferocity to maintain their holdings and their way of life. Subsequently, dur- ing Spanish colonization, these Florida Gulf Coast Indians were among the first in the North American continent to suffer disruption, population decimation, and general cultural break-down. These events, dramatizing the impact of two worlds, took place between the years A. D. 1500 and 1700. At the beginning of the sixteenth century the Florida Gulf Coast cultures were thriving and vigorous; two cen- turies later they had all but vanished except for the archeological record which they left behind them. It is difficult to say which of the Europeans was the first to reach the shores of Florida. We know, however, that the event must have taken place very shortly after the discovery of the New World and the implantation of the Spanish Colonies in the West Indies. To Ponce de Leon, in 1513, goes the official credit of Florida discovery, although it is almost certain that he was preceded by other less- heralded explorers. It is questionable as to just what section of Gulf Florida or with what Indian group De Leon touched. Swanton and others (Swanton, 1946, pp. 34 ff.) feel that it was most likely south Florida and among the Calusa rather than farther north. He was received with hostility in 1513 and again in 1521, the second voyage ending in his death. Afterward, coasting vessels under the commands of Diego Miruelo, in 1516, and Alvarez de Pineda, in 1519, followed along the Gulf Coast shore, but their records tell us of little except the antagonistic attitudes of the aroused Indians. Pineda does, however, describe 40 villages, indicative of quite a population concentration, in the vicinity of what was probably Mobile Bay. The Narvaez expedition of 1528 gave the Gulf Coast tribes their first taste of European soldiery in the mass. Narvaez is believed to have landed just north of Tampa Bay, whence he struck off inland 517 518 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I13 afoot with 300 men. He crossed the Withlacoochee and Suwannee Rivers and eventually reached Apalachee towns where his men found corn and other food supplies. The party was driven out of the Apalachee country by the Indians, and they marched westward to Pensacola Bay, in which vicinity they reported long-haired aborgines who put up a battle with slings and darts. On the other side of the peninsula other Spaniards, under Ayllon and Quexos, were busy laying the groundwork for bad relations be- tween White and Indian as early as 1520-21. These explorers dis- covered the St. Johns River and carried off a group of Indians as slaves to the West Indies. Ayllon and Quexos also attempted un- successful colonies on the South Carolina and Georgia coasts. In May of 1539 the famed De Soto expedition landed at the south- ern end of Tampa Bay. The party consisted of 600 men, their camp followers, horses, hogs, mules, dogs, and heavy provisions. It was Spain’s first full-dress attempt to repeat in the North American woodlands what had been carried out so successfully in Mexico and Peru. De Soto and many of his followers were, in fact, veterans of the Peruvian venture. As it turned out, their stay in Gulf Florida was not an extended one, but the accounts of the expedition, to- gether with those of the Narvaez party, give us our only descriptions of Indians and Indian culture in the area for the period. They landed near an Indian town called Ucita which is believed to be the present location of Shaw Point in Manatee County. This site, as is the case with many points or passages of the De Soto party, is disputed, but it seems fairly certain that the landfall was made some- where in the Tampa Bay section. Ucita, or Ocita, is thought to have been a Timucuan town. From here they made contact with an inland chief who was said to dominate the tribes of the coast. This relationship among the tribes is interesting and may signify a general trend for the times. In the northwest as well, the more powerful centers and towns were in the interior, not immediately upon the coast. From Ucita, De Soto marched north and crossed the Alachua area which was occupied by a powerful Timucuan group, the Potano. Other Timucuan tribes were met with, and the Spanish had their first big fight with the Indians, probably the Uriutina, at a place called Two Lakes. They crossed the Suwannee River on Septem- ber 25, after having left Ucita on July 15 of that year. The last Timucuan town encountered, after resuming their march northward, was called Agile or Aucilla and was located on that river. After crossing the Aucilla they found themselves in Apalachee territory, presumably meeting up with some of the same Indians that had WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—-WILLEY 519 known the Narvaez party of 1528. Their definiteness about the mat- ter of the tribal boundary between the two groups is of interest as it shows that the Indians of the area had clear convictions concern- ing territorial or “national” divisions. Shortly after, on October 6, they arrived at Iniahica, the principal Apalachee town which was located in the vicinity of Tallahassee, possibly at the site of Lake Jackson (Le-1). (See pp. 95-98.) Throughout, the march had been an inland not a coastal one with the main body of the army passing along a route at least 30 to 50 miles from the Gulf. The De Soto company spent the winter of 1539-40 at Iniahica while their unwilling hosts awaited their departure with unconcealed anticipation. Apparently, only the size and strength of the army kept it from being set upon and wiped out. During the wait, subsidiary parties of exploration were sent out to seek a harbor as a base for supplies for the coming year. This was done by selecting such a site to the westward at either Pensacola or Mobile. Except for these minor survey parties, the Florida Gulf Indians were spared further attentions of the De Soto party. In March of 1540 the main army set out to the northwest, crossing the Ocklockonee and, eventually, the Flint River. Hereafter, they encountered the Muskogean groups of Georgia who were, at least at first, much less hostile to the in- vaders than the Timucuans or Apalachees. De Soto’s subsequent travels led inland and he never returned to the Gulf Coast. After De Soto, there were sporadic contacts with the Indians of the Gulf Coast for the next several decades. Many of these were by way of wrecked treasure ships returning from Panama to Spain. From these the Indians salvaged gold and other articles. The boy Fontaneda was the victim of one of these wrecks in the sixteenth century and lived for years among the Indians of south Florida. Attempts at colonization and Christianization of the Indians were also made in the mid-sixteenth century. One of these, instigated by Menendez, was with the Tocobaga in Old Tampa Bay in 1567. This small expedition ended disastrously for the Spanish. A much larger venture was undertaken under the leadership of De Luna. This was farther to the north in the neighborhood of Mobile Bay. Probably the largest group of colonists yet to land in North America, the De Luna party comprised over 1,000 persons, including Mexican In- dians. Working north from the Mobile-Pensacola region these col- onists spent over a year in Alabama, but the project was finally abandoned and the survivors returned to Hispaniola. The first successful Florida colony was founded on the east side of the Peninsula in 1565 at St. Augustine. Its settlement came about 520 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 113 as a countermeasure to a French thrust of 1564. In that year Laudon- niére had established a town on the lower St. Johns River, but the Spanish immediately retaliated by wiping out the settlement and setting up their own not far distant. The French interlude in Florida, though brief, is responsible for the best ethnological accounts of the Indians of the time. These accounts pertain to the Timucuan tribes that inhabited the northeastern part of the peninsula and were the linguistic kinsmen of the Timucuan groups encountered by Narvaez, De Soto, and others in Gulf Florida. The Spanish mission system grew out of the St. Augustine base. In the north the Franciscans were quite successful, first founding missions among the Timucua and eventually reaching the Apalachee in 1633. Farther south the Indians were less amenable. We know that in 1612 a punitive expedition was sent against the Tocobaga by way of reprisal for attacks by those Indians against some of the missionized natives of the north. There was also trouble in the north as well. Revolt flared among the Apalachee in 1647, and in 1656 the Timucua rose in northeast Florida with the result that the trouble spread to the west involving the Apalachee. These revolts may mark the crucial breaking point of the old northwest Florida native cul- tures. The rebellions were put down with some bloodshed, and many of the Indians fled the area forever. Following the revolt of 1656, conditions in the Apalachee country and along the northwest coast seem to have been fairly stable and peaceable. Forces were at work outside of the area, however, that were to bring about the destruction of the western missions and terminate the era of relative peace. As a result of sporadic border warfare between the Spanish and the lower Creeks on the Chattahoochee River in what is now south Georgia, many of the Creeks left their homeland, joining their relatives on the Ocmulgee River in central Georgia. Here they were consolidated into an effective fighting force by Col. James Moore who led the English-Creek raid upon the Apala- chee-Spanish missions in 1704. The missions never recovered from this blow. Following more fighting in 1706-7, many of the Apalachee, Apalachicola, and Chatot were uprooted from northwest Florida and moved to the Savannah River where they settled under the watchful eye of the English in the Carolinas. Some fugitives from the re- moval fled west and took protection under the French at Mobile. The English and Creeks also carried their war against the Timucua missions, destroying many of them. The period was one of confu- sion and flux. In 1715 the Yamasee moved south out of Georgia and joined the Timucua in northeast Florida. At the same time some WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—-WILLEY 521 of the Apalachee and Apalachicola returned to their old homes in Florida. The tribal identities of the old northwest Florida tribes begin to disappear early in the eighteenth century. Some of the Indians merged with the Creeks who were moving south out of Georgia ; others went west and joined the Apalachee and Apalachicola villages around Mobile and Pensacola, preserving for a while their tribal organiza- tions. The old cultural traditions of the Gulf Coast were, however, lost or submerged in the confusion of war and change and never reappeared. For our consideration of the development and history of native Indian cultures in Gulf Florida the early European explorations and colonizations may be divided conveniently into two periods. The earlier period begins at about 1500, the time at which the first Spaniard appeared in Florida. Although these strangers must have struck fear into the hearts of the natives, it is doubtful if these essen- tially casual adventurers had the effect of seriously upsetting the aboriginal patterns of life. Even the relatively huge De Soto party served to irritate rather than to destroy or profoundly change the Indian tribes.7®? One of the most important of the circumstances conditioning relations between Indians and Europeans during this early period of contact was the attitude of the Spaniards. The latter were at this time interested in a duplication of their Mexican or Peruvian triumphs which had been based upon quick material gain (gold) and rapid control of native governmental machinery (conquest of Aztec and Inca Empires). Neither proved possible in aboriginal eastern North America. Gold was rare or absent, and the political systems of the southeastern woodlands were not structured to permit a sudden coup d’etat type victory whereby a small group of out- siders could, as it were, decapitate the government and assume con- trol of the body politic. In view of these interests of the invaders and the conditions surrounding the invaded, the outcome was in- evitable: frustration and exhaustion for the former and only tem- porary alarm and disruption for the latter. The second period of contact had an orientation significantly dif- ferent from the first. The goals of the Spaniard were no longer quick plunder and glory but colonial establishment and religious con- version of the inhabitants. The effect of this second phase upon the 79 The factor of disease is the one possible contradiction to this. We know that a century later European diseases did have disastrous effects upon native Florida populations. There is no record of what happened in 1539-43, or imme- diately afterward, but it is likely that conditions for contagion would not have been as great as they were later under colonization. 35 522 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I13 Indians was markedly greater than anything that had happened to them during the previous century of contact. In the Gulf area from 1633 onward native cultures underwent serious change. Some of it proceeded peacefully, although there were brief periods of revolt. Disease and the Indian death rate undoubtedly increased. This Spanish mission phase lasted until just after the close of the seven- teenth century when it was brought to a violent end as the result of intertribal warfare stimulated by the competition of the European powers who were rivals for the lands and peoples of the New World. The reality of these two periods of European contact in Gulf Florida, the first, purely exploratory and adventuristic, the second, colonial and systematic, is clearly reflected in the archeological cul- tures of the northwest coast. The earlier Fort Walton culture, though in occasional contact with the Spaniards, is thoroughly native and integrated; the later Leon-Jefferson culture, on the other hand, is one which has been turned from its old channels and is under- going rapid acculturative change. TRIBES The Apalachee—The Apalachee were probably the most powerful of the tribes living in the northwest Gulf Coast area of Florida. Linguistically, their affiliations are with the southern division of the Muskogean stock. The importance and stability of this tribe are attested by the numerous early historical accounts covering the 200- year period following the arrival of the first Europeans. At the time of the Narvaez (1528) and De Soto (1539-40) expeditions the Apala- chee occupied a large part of northwest Florida. Swanton (1922, pp. 109-110) defines their territory as follows: The Apalachee proper occupied, when first discovered, a portion of what is now western Florida between the Ocilla (Aucilla) River on the east and the Ocklockonee and its branches on the west. They probably extended into what is now the State of Georgia for a short distance, but their center was in the region indicated, northward of Apalachee Bay. Tallahassee, the present capital of Florida, is nearly in the center of their ancient domain. Later, he enlarges this territory slightly by placing their western boundary at the Apalachicola rather than the Ocklockonee River (Swanton, 1946, p. 89). A number of Apalachee towns and communities are mentioned in the sixteenth-century chronicles. Some of these seem to be principal towns, separated from each other by several leagues. Of these, Iniahica was the most important and was located near the present WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—-WILLEY 523 site of Tallahassee. Other important Apalachee towns mentioned by the De Soto narrators are Calahuchi, Uzela, and Ochete. All these seem to have been in the interior. Another major town, Aute, visited by the Narvaez party, lay some 9 days’ march to the south- west of the other Apalachee towns, being near the sea. It is also likely that there were Apalachee towns other than those of which we have record. These principal towns seem to have been the various nuclei around which the settlement pattern of the Apalachee was structured. Un- fortunately, we have no very good descriptions of these towns which would indicate the disposal of mounds, temples, public buildings, and dwellings. Cabeza de Vaca, in the narrative of the Narvaez ex- pedition, states that one of the major Apalachee towns contained about 40 small houses. He describes the houses as being made of straw. The village area, as he pictures it, seems to have a clearing in a dense, swampy forest. He also mentions that houses were “scattered all over the country” (Swanton, 1922, p. 113). This last would imply smaller outlying settlements at some distance from the main centers. It seems confirmed by the De Soto chronicler, Elvas, who says, in speaking of the country around Iniahica: At the distance of a half a league to a league apart there were other towns which had much maize, pumpkins, beans, and dried plums of the country, whence were brought together at Anhayca (Iniahica) Apalache what appeared to be sufficient provision for the winter. (See Swanton, 1922, pp. 116-117, footnote p. 117.) The storing of food supplies, taken from surrounding villages, in the principal town of Iniahica definitely suggests the pattern of the politico-religious center sustained by outlying farming communities. The economy of the Apalachee was primarily agricultural. Corn, beans, and squash were the principal cultivated foods, and these were supplemented by native fruits. The fruit trees do not seem to have been cultivated in a strict sense but may have been tended in some fashion. Elvas states: These ameixas (persimmons) are better than those of Spain, and come from trees that grow in the fields without being planted. Elvas refers to these as “dried plums,” Ranjel mentions “dried veni- son” stored in the villages, and we know beans and corn were also stored. It would appear that the food preservation techniques of the Apalachee were quite adequate. Supplementary diet from hunting and fishing is reflected in the Ranjel statement: The Province of Apalache is very fertile and abundantly provided with supplies with much corn, kidney beans, pumpkins, various fruits, much venison, many 524 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I13 varieties of birds and excellent fishing near the sea;..... (Swanton, 1922, p. 116.) Political unity is implied for the Apalachee in all the early state- ments, but we do not know how this was effected or exactly what it meant. There was general agreement among the Indians that Iniahica was their principal town, but it certainly was not the only Apalachee town of importance. Elvas records that “the lord of all that country and Province (of the Apalachees) resided” at Iniahica (Swanton, 1922, p. 116). However, the nature of the authority this “lord” wielded, whether or not authority rested in the hands of one man or more, and the manner in which the authority was enforced are not revealed. It is most likely that the “Province of Apalachee” was a confederacy of closely related tribes or subtribes, each, per- haps, with its own principal town. The dominance of Iniahica, as the most populous and strongest unit, was recognized by the other groups within the confederacy, but there is no evidence of tribute by other towns to Iniahica other than the transporting of food supplies to the main town from nearby villages (apparently not principal towns). It is not clear as to whether or not there was warfare among the various Apalachee towns. In the Cabeza de Vaca account of the 1528 expedition there is an inference to what might be internecine enmity of this sort, but, for the most part, peace and unity seemed to have prevailed, at least in the sixteenth century. Little cultural detail is afforded in the early Spanish descriptions. We know nothing concerning the burial customs of the Apalachee, and very little of their manufactures. Cabeza de Vaca mentions deer skins and “mantles made of thread.” Over 100 years later Apalachee clothing is described as being made of bark, skins, and roots of trees. Feather headdresses and gold are listed in Cabeza’s account as having been in the possession of Timucua Indians who claimed to have received them from the Apalachee. It is likely that the gold was from Spanish wrecks. Mortars for grinding corn are men- tioned, but there is no description of them. Probably they were made of wood. The Apalachee are described as excellent archers by both of the early expeditions. Cabeza states: Their bows are as thick as an arm, from eleven to twelve spans long, shooting an arrow at 200 paces with unerring aim. (Swanton, 1922, p. 114.) Cabeza noted that Spaniards wearing armor were, nevertheless, killed by arrows shot with “force and precision.” The nature of the pro- jectile points is not disclosed, however. In connection with warfare, Garcilasso says that the Indians took the scalps of the Spaniards, which were much prized, and hung them upon the arms of their bows. WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—-WILLEY 525 It is presumed that life must have continued much the same for the Apalachees until the seventeeth century. After the establish- ment of the mission systems we know that there were important changes in the old culture. Except indirectly, the historical sources of the seventeenth and eighteenth centuries give us little on the exact nature of the changes. These are seen best in the contrasting archeo- logical patterns of the Apalachee area. It is a certainty that the Indian Mission culture of the archeologically defined Leon-Jefferson Period can be identified as Apalachee of the period of 1633 to 1704 or slightly later. From this it can be reasonably inferred that the al- most wholly indigenous archeological culture of the Fort Walton Period, which immediately precedes the Leon-Jefferson, corresponds to the Apalachee as they were found by Narvaez and De Soto in the sixteenth century. Other tribes of northwest Florida——tThere are records of a num- ber of other tribes occupying the area to the west of the Apalachee, but these accounts do not, for the most part, treat of these Indians during the sixteenth century. One of these groups was the Chatot. These people are classified as an independent linguistic group of the southern division of the Muskogean stock. Their original homeland probably was along the Apalachicola River, but they do not appear in the literature until 1639 at which time they were living somewhere to the west of the Apalachicola, probably near the middle course of the Chipola. Spanish missions were established among them in 1674. After this their history was more or less a common one with the Apalachee missionized Indians. (See Swanton, 1922, pp. 134-137; 1946, pp. 107-108.) The Sawokli are another group who seem to have had an old claim to a portion of northwest Florida. Linguis- tically, the Sawokli are classed as a subdivision of the Hitchiti group, southern division, Muskogean stock. Their original home is believed to have been somewhere along the Gulf Coast to the west of the Apalachicola River. By 1675 they were located on the Apalachicola River a few miles below the present Georgia-Florida line. They were missionized and, later, became involved in the Creek wars along with the Apalachee and their neighbors. (See Swanton, 1946, pp. 179-181.) The Apalachicola were a related Hitchiti-speaking tribe or small group of tribes. They were located on the river of the same name near the present site of River Junction in the late seventeenth century. They were attacked and scattered by the English and Creeks in 1706-7 (Swanton, 1946, pp. 92-93). Farther to the west, around Pensacola Bay, were the Pensacola. Apparently, these Indians were in that vicinity during the sixteenth century at the time of Narvaez’ 526 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I13 voyage. Cabeza de Vaca describes them as being tall and well built and living in permanent houses made of matting. The chief was lodged in a special house and was dressed in a cloak of “marten- ermine’ skins. They had slings, darts, and bows and arrows for weapons. Some are described as wearing their hair long. Many years later, in 1686, the Spaniards recorded a tribe by the name of Panco- colos or Panzacola (Pensacola) at Pensacola Bay. These Pensacola were then at war with the neighboring Mobile. Shortly after this they seem to drop from history, although there are some later ac- counts which may refer to them. Linguistically they are Muskogean, being a subdivision of the Choctaw group (Swanton, 1922, pp. 143-150). The remaining tribes that lived for a short time in northwest Florida appear to have been relatively late comers. Swanton (1946, map I1) locates these as the Yamasee, Yuchi, Tawasa, Pawokti, Tamathli, Mikasuki, and Osochi. Most were Muskogean or Yuchean peoples who moved into Florida from the north late in the seven- teenth or in the eighteenth century. The Osochi may have been a Timucuan tribe from peninsular Florida. Archeological identification of the tribes that lived on and west of the Apalachicola River can be made only in a very general way. It is likely that the ancestors of the Chatot, Apalachicola, Sawokli, and Pensacola were the bearers of the Fort Walton culture as it existed west of the Apalachee region. Probably these tribes made up small nations, each with its own territory, much as did the Apalachee. They had a common linguistic heritage and the cultural differences that existed among them, as reflected in the archeology of the Fort Walton Period throughout northwest Florida, were not great. After the Spanish mission period of the seventeenth century these tribes may have developed a culture very similar to that we have termed the Leon-Jefferson. Leon-Jefferson sites have not, as yet, been re- ported west of the Apalachee territory, although they may be dis- closed in the future. On the basis of what we know of the history of the Spanish missions in north Florida it is to be expected that mission influence upon Indian cultures would have been less pro- nounced west of the Apalachicola River. The Timucua.—South and east of the Aucilla River, extending over two-thirds of the way down the entire Florida peninsula, is the area once occupied by the Timucuan-speaking tribes. This relatively large region did not have the semblance of political unity that characterized the Apalachee nation. It seems, rather, to have been composed of a number of tribes or chieftainships, many of which were at war with WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—-WILLEY 527 each other. Further, the archeology reveals little cultural unity for the area. The Gulf Coastal strip, extending from the Aucilla River down to Charlotte Harbor (the central Gulf Coast and Manatee archeological regions) was fairly homogeneous in the later periods (Weeden Island and Safety Harbor) ; however, this cultural homo- geneity cannot be extended across the peninsula to interior Central Florida or the St. Johns area where the Timucuan linguistic stock also prevailed. It will be remembered that both Narvaez and De Soto landed in Timucuan territory and marched through that country before reaching the land of the Apalachees. Their routes lay back from the coast partly within the Gulf Coast area but mostly within central Florida. The prin- cipal early tribes in the country through which they passed, as located by Swanton (1946, map II), are, from south to north: the Pohoy or Ocita (Ucita) ; the Mococo; the Tocobaga ; the Acuera ; the Ocale ; the Potano; the Utina ; the Onatheaqua; and the Yustaga.®° As with the Apalachee, the cultural data on these Timucuans are rather scanty as far as the Spanish sources are concerned. There is, in fact, nothing on the tribes of the immediate coast north of Tampa Bay. What information we have comes largely from the tribes situated 30 to 50 miles in the interior. Data on the Timucua as a whole are, how- ever, richly supplemented by the French accounts of 1564-65. These mainly refer to Timucuan tribes living on or near the St. Johns River. To just what extent can these eastern data be used to apply to Timucuan tribes farther west and south? We have made the observation that there are differences existing between the archeology of the western and eastern Timucuan areas. These differences pertain largely to styles of manufactures. Many similarities between east and west do exist, and the similarities seem to be of a fundamental sort, applying to basic technologies, to burial customs, and to the construction of mounds, the latter reflecting some- thing of the religious, social, and political mores of the people. In the light of such similarities it is reasonable to believe that the observations which the French made on the eastern Timucuans have at least partial relevance in a consideration of their western relatives. The first description of a Timucua community is that of Ucita on Tampa Bay, the landing place of the De Soto expedition in 1539. This account from the De Soto narratives states: The town (of Ucita) was of seven or eight houses, built of timber, and cov- ered with palm leaves. The chiefs’s house stood near the beach, upon a very 80 A complete list of Timucua provinces, tribes, towns, chiefs, and later mis- sions, covering east Florida as well as the Gulf Coast, is given by Swanton (1922, pp. 322-330). 528 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. II3 high mount, made by hand for defense; at the other end of the town was a temple, on the roof of which perched a wooden fowl with gilded eyes. (Swanton, 1922, p. 353-) Significantly, we note substructure mounds at the site and a differ- entiation between the chief’s house and the temple. It is not made clear whether or not the temple building also stood upon a mound. Farther north, in the Uriutina country, open courts or “town yards” were noted by the Spanish (Swanton, 1922, p. 353). Presumably, these “courts” were centers around which houses or other buildings were grouped. In northeast Florida the French give more detailed descriptions of towns and houses, and these differ in some respects from what the Spanish saw at Ucita. These eastern Timucuan houses also were made of timbers and covered with reeds or palmettos, but Le Moyne de- scribes the chief’s house as being in the center of the village with the houses of the principal men grouped around it. According to him the dwelling of the chief was semisubterranean. The Le Moyne drawings show most of the Timucua houses as circular in form, but the squarish ground plan was also used by the Indians. These oblong houses were, apparently, both gabled and dome-roofed and often served as “town houses.” Concerning the town house, Swanton (1922, p. 353) quotes from Spark: Their houses are not many together, for in one house an hundred of them do lodge; they being made like a great barne .... hauing no place diuided, but one small roome for their king and queene. The absence of mention of substructure mounds or mounds of any kind is curious, as the archeological record reveals temple-type mounds in northeast Florida in the late periods, contemporaneous with those of the Safety Harbor culture in the west.8t The town house feature may be distinct from anything in the west or northwest ; however, the records of Apalachee dwellings are not sufficiently detailed to enable us to compare this point. The Timucua often fortified their towns. Le Moyne describes this: A position is selected near the channel of some swift stream. They level it as even as possible, and then dig a ditch in a circle around the site, in which they set thick round pales, close together, to twice the height of a man; and they carry this paling some ways past the beginning of it, spiralwise, to make a narrow entrance admitting not more than two persons abreast. The course of the stream is also diverted to this entrance; and at each end of it they are accustomed to erect a small round building, each full of cracks and holes, and 81 The St. Johns II Period, estimated as terminating around 1650. WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—WILLEY 529 built, considering their means, with much elegance. In these they station as sentinels... .. (Swanton, 1922, p. 379.) The French were very explicit concerning the food economy of the Timucua. Corn or maize was the main staple, and Laudonniére main- tained that they planted twice a year, in March and again in June. Ac- cording to Le Moyne they planted late in the fall and let the corn remain in the ground to ripen the following summer or spring. Pump- kins and beans are also mentioned as important crops. Land for plant- ing was cleared by burning and tilled with mattocklike instruments made of wood. Agricultural labor was performed by both sexes and was directed as a group effort by the chief or “king.” Common and probably private storehouses were provided for the agricultural pro- duce. Both Laudonniére and Le Moyne state that the three winter months were spent in small, individual family huts in the woods, dur- ing which time the Indians subsisted largely upon fish and game. Fish and animal foods were also taken, prepared, and used within the larger community settlements. Le Moyne’s famous drawing of the babracot, or wooden rack, placed upon forked stakes over a fire, shows fish, a deer, an alligator, and a snake, among other animals, being smoked and dried. Such prepared animal foods were then placed in storehouses. Besides the more nutritious plant foods it appears the Timucua also included various wild roots and fruits in their diet. Tobacco was used and, undoubtedly, cultivated.*° The principal Timucuan sociopolitical unit was the tribe, and the tribe was headed by a chief who wielded great authority. The dig- nity and rank of the chief was emphasized with a pomp that is sur- prising considering the relatively simple economic and technological levels of the culture. The larger tribes were composed of a great number of towns which had been combined into a domain. Appar- ently each town was a self-sustaining economic unit with its own chiefs or leaders, but these head men were, in turn, under the com- mand of the principal town of the tribe and the tribal high chief. The French speak of five of these “supertribes” or confederacies. On the lower St. Johns were the Saturiwa; the Timucua proper, or Utina, were situated around Santa Fe Lake in Alachua County ; the Potano were in the Alachua plains, and to the northwest of these eastern Timucua were the Onatheaqua and the Yustaga. It is probable that some of these tribal holdings were about as extensive as the territory claimed by the Apalachee in northwest Florida. South and west of 82 Swanton (1922, pp. 357-362) gives considerable detail on Timucua foods, farming, hunting, and fishing. 530 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I13 the principal Timucua tribal states were others, perhaps less powerful, such as the Ocita and the Tocobaga of Tampa Bay. Concerning social organization Swanton says: Two different classifications seem to be represented here, of which the second is plainly along the line of clans, and the groups probably were in fact clans similar to those of the Creeks. The first, however, indicates a kind of aristo- cratic system which appears to have been based on male descent and recalls somewhat the special privileges accorded to children and grand children of “Suns” among the Natchez. Perhaps these “lineages” were actually associated witheclanss aie (Swanton, 1922, p. 370.) 8% The power and dignity of the chiefs was reflected in all phases of life. His special function as a religious leader, as distinct from the tribal shaman or sorcerer, is indicated by Le Moyne’s description of sun worship. In these rites the chief leads the prayers and directs the offerings to the sun god. The chief’s central position in cult life is further exemplified by Le Moyne in his description of human sacrifice : Their custom is to offer up the first-born son to the chief. When the day for the sacrifice is notified to the chief, he proceeds to a place set apart for the purpose, where there is a bench for him, on which he takes his seat. .... The ceremonies being through, the sacrificer takes the child, and slays it in honor of the chief, before them all, upon the wooden stump. (Swanton, 1922, p. 382.) Warfare was developed around the significance and prestige of the chief. It was highly ceremonialized and based upon raiding, capture, and ritualistic torture of prisoners. In civil councils or ceremonies the chief’s prerogatives were equally marked. A marriage ceremony for a great chief is described by Le Moyne: When a king chooses to take a wife, he directs the tallest and handsomest of the daughters of the chief men to be selected. Then a seat is made on two stout poles and covered with the skin of some sort of rare animal, while it is set off with a structure of boughs, bending over forward so as to shade the head of the sitter. The queen elect having been placed on this, four strong men take up the poles and support them on their shoulders, each carrying in one hand a forked wooden stick to support the pole at halting. Two more walk at the sides, each carrying on a staff a round screen elegantly made, to protect the queen from the sun’s rays. Others go before blowing upon trumpets made of bark, which are smaller above and larger at the farther end and having only the two orifices one at each end. They are hung with small oval balls of gold, silver, and brass, for the sake of a finer combination of sounds. Behind follow the most beautiful girls that can be found, elegantly decorated with necklaces and armlets of pearls, each carrying in her hand a basket full of choice fruits, 83 Swanton (1922, pp. 362-371) considers social organization in detail. WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—-WILLEY 531 belted below the navel and down to the thighs with the moss of certain trees, to cover their nakedness. After them come the bodyguards. With this display the queen is brought to the king in a place arranged for the purpose, where a good-sized platform is built up of round logs, having on either side a long bench where the chief men are seated. The king sits on the platform on the right hand side. (Swanton, 1922, p. 372.) The litter and throne are symbols of high status in a well-defined class or caste system. Disposal of the dead among the eastern Timucua was, apparently, by simple burial. Le Moyne’s drawing of ritual mourning for a dead chief shows a small hillock, beneath which the body has been buried. Around the hillock are a circle of arrows placed in the earth, and on top of the hillock is a shell drinking cup. The hillock itself is not de- scribed in Laudonniére’s text (Swanton, 1922, p. 373), but possibly it represents a burial mound although the heap of earth is, to judge from the proportions of the drawing, very small. It is said that the house and belongings of the chief were burned after his death. Priests (sor- cerers?) were buried in their houses which were afterwards burned. In the west burial customs were somewhat different. In 1528 the Narvaez party, somewhere in the central part of the peninsula, came upon an Indian village (apparently Timucuan) where they saw sev- eral human corpses, each in a box and each covered with a painted deer hide. This suggests the storage of the bodies of the dead, prob- ably prior to secondary burial. Swanton (1922, p. 374) quotes from an early manuscript concerning burial among the Tocobaga: When one of the principal caciques dies, they cut him to pieces and cook him in large pots during two days, when the flesh has entirely separated from the bones, and adjust to one another until they have formed the skeleton of a man, as he was in life. Then they carry it to a house which they call their temple. This operation lasts four days and during all this time they fast. At the end of the four days, when everything is ready, all the Indians of the town get together and come out with the skeleton in procession, and they bury it with the greatest show of reverence. This is certainly secondary burial, a form of interment quite common among the Indians of the central Gulf Coast during both the Weeden Island and Safety Harbor Periods. At Safety Harbor proper (P1-2), probably a Tocobaga site, secondary burials were found in a burial mound of sand (see pp. 135-142). It is not clear from the above de- scription whether or not the bones were buried in a mound or a cemetery. The manufactures of the Timucua included wooden benches, beds, and stools for the houses, pottery, baskets, wooden vessels, carrying baskets, woven sieves, fans, mats, painted gourds, dressed skins, stone 532 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOE. 113 axes, feather fans, shell tools, bows, and arrows. Pottery pictured by the French shows a variety of forms. Among these are strap-handled jars, a definitely late period form in the archeological sequences for Florida. The stone axes or celts are, of course, common to the archeological collections from all parts of the Timucua area, east as well as west, and in all the middle and late periods. Projectile points are described by the French as being made of the “teeth of fishes” (probably shark teeth), of stone, of metals (probably European trade metal), and of hardened wood. Arrows were carried in skin quivers. Canoe travel, in square-ended dugout canoes, was the only method of transport other than walking. The Indians possessed the domesti- cated dog. The Timucua themselves are described by Ribault as being “‘of good stature, well shaped of body as any people in the world; very gentle, courteous, and good-natured, of tawny color, hawked nose, and of pleasant countenance.” The breechclout of painted deerskin was cus- tomary for adult males while women wore a skirt made of Spanish moss or of skin. Garments imported from the western Timucua, which came to the eyes of the French, were made of feathers and col- ored rushes. The eastern Timucua men wore their hair in a curious “upswept” hairknot as is seen in all the Le Moyne drawings. Feather and metal ornaments were worn in the hair or on the head, and woven hats are also depicted. Other ornaments for the person include ear plugs, fresh-water pearls, shell beads and gorgets, fish-tooth bracelets, and pendants, diadems, and gorgets of gold, silver, and copper. These last probably were made of metals taken from Spanish treasure ships, although some may have been wholly aboriginal. Tattooing of the skin and painting of the body and face seem to have been fairly common. (See Swanton, 1922, pp. 345-352.) Most of the artifacts and ornaments listed by the French explorers can be identified among the archeological collections from the Timucua area. There is little question but that the Timucuan peoples around Tampa Bay were the bearers of the Safety Harbor type culture in the seven- teenth century. In fact, the Safety Harbor site proper is generally conceded to be the Tocobaga capital. It is less certain if Safety Harbor type culture was in existence 50 to 100 years earlier at the time of the De Soto landings, although a correlation of the Ucita site with the Safety Harbor Period is the most likely possibility. The only other alternative would be that Weeden Island culture was still extant in 1539, a correlation that is less probable. We know that at Ucita De Soto observed the chief's house based upon a mound, undoubtedly a substructure or “‘temple”-type mound. Although substructure mounds WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—-WILLEY 533 may appear as early as the Weeden Island Period, they are much more securely identified with the Safety Harbor Period. The Safety Harbor Period is, of course, the period of European contact in both the central Gulf Coast and Manatee regions; no other period in these regions incorporates European trade goods. Trade goods are found in the majority of Safety Harbor sites and more abundantly in some of these than in others. These sites remain, however, aboriginal rather than Europeanized. Mound burial, for example, continues throughout the Safety Harbor Period as far as we can determine. In other words, the Indian Mission culture, compar- able to the Leon-Jefferson Period of the northwest, is lacking on the central Gulf Coast, although we are quite certain that some of the Safety Harbor sites were occupied contemporaneously with Leon- Jefferson. This accords with the historical records of the establish- ment and maintenance of the mission chain in Florida. While the northern tribes were brought successfully into the fold, the southern groups either resisted or were largely ignored by the Spanish during the sixteenth and seventeenth centuries. After 1700 some of the Tocobaga and Pohoy came north and settled near missons in the vicinity of St. Augustine where most of them died of pestilence in 1726. This removal, presumably, terminated the Safety Harbor culture. Very little is known of Timucuan peoples of the Gulf Coast either north or south of Tampa Bay. Goggin (n.d.2) is of the opinion that the coastal strip to the south, in effect the largest part of the Manatee archeological region, was not occupied during the latter half of the Safety Harbor Period. He bases this upon absence of any historical accounts of peoples in this region plus the fact that most of the Safety Harbor sites south of Manatee County are with- out European trade materials and apparently represent only the earlier phases of the Safety Harbor Period (prior to A. D. 1600). To the north of Tampa Bay and environs our site surveys and reviews reveal only two late period sites, Bayport (He-1) and Buzzard’s Island (Ci-2). The scarcity of late period sites from the central Gulf Coast north of Pinellas County and the absence of any early historical accounts of Indians in this region may mean that the region was deserted, or nearly so, during the sixteenth and seventeenth centuries. POPULATION ESTIMATES The best population estimates which can be made for the historic tribes of the Florida Gulf Coast are reckoned from census figures 534 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 113 provided by soldiers, administrators, and missionaries in the seven- teenth and early eighteenth centuries. There are almost no data of this sort from sixteenth-century exploration accounts. Estimates which I offer for the mid-sixteenth century are my own and are pro- jections backward from seventeenth-century totals. In northwest Florida many of the population figures given by the Spanish missionaries for the Apalachee during the seventeenth cen- tury were exaggeratedly high. Swanton questions these and offers a total of 5,000 persons for this tribe as of the year 1676 (Swanton, 1946, p. 91). Mooney, estimating for 1650, places the figure at 7,000 (Mooney, 1928, p. 8). At the time of the defeat of the Apalachee by the English and the Creeks in 1704, their population is believed to have declined to about 2,000 (Swanton, ibid.). The 5,000 to 7,000 estimate for the mid-seventeenth century should, I think, be increased somewhat for the mid-sixteenth century, although I doubt if the differ- ence was great. The Indian cultures of 1650 were just beginning their transition from aboriginal to Europeanized, and the results of the white man’s rule had not yet built up to their cumulative effect. I offer an estimate of from 6,000 to 8,000 Apalachee as of 1550. The Chatot were, apparently, less numerous than the Apalachee. In 1674 there were probably about 500 persons in the tribe (Swanton, 1946, p. 108). One hundred years or more before, there may have been 1,000. In 1750 the Sawokli numbered 50 men (Swanton, 1946, pp. 180- 181). This figure probably stands for five times as many people in the tribe, or a total of 250. We know that by 1750 the northwest Florida tribes had all suffered population decimation. A mid-seventeenth- century estimate of 500 does not seem excessive, projected to a 1550 guess of 1,000 persons. The Apalachicola, similarly, numbered about 200 Indians in 1715 (Swanton, 1946, pp. 92-93). In the mid-seventeenth century their population was probably in the neighborhood of 500, with a 1550 total of 1,000. We have almost no data for the Pensacola. A sheer guess of 1,000 persons for 1550 is offered. These figures total 10,000 to 12,000 Indians for northwest Florida at about the year 1550. The sum approximates that given on Kroeber’s revision of Mooney’s chart for the combined tribes of northwest Florida (Kroeber 1939, p. 138). If this is a 1650 rather than 1550 estimate, I am inclined to say that it is too high. If, however, Kroeber is offering a “climax” period population total his summary of 12,000 fits with that given here (see pp. 468-470). WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—WILLEY 535 On the Gulf Coast of the peninsula we are unable to do as well by way of estimates. The Timucua figures advanced by Mooney are based upon the eastern and northern Timucua. The Tocobaga, one of the few Timucuan groups of the Gulf Coast upon which we have any figures, were said to have been able to muster 1,500 warriors to meet the Spanish in 1567. This, if reliable, was probably the full military strength of the tribe which would give us a total population in the neighborhood of 7,500 persons. Swanton (1946, p. 196) casts some doubt on the 1,500 figure, and I am also of the opinion that it is excessive. Mooney estimated 1,000 Tocobaga for 1650. No figures of any kind are available for the Mococo, but there were said to be 300 Pohoy in the year 1680. I offer the archeological estimate of 5,000 for the Tampa Bay region as of 1550 with a subsequent decline in the seventeenth century (see pp. 487-488). PU es : Py oN eget f y i yeh arth (ame ie ho fie spel Brey Fy) ae vit pare ullgipt naa ny pe cody > Os i ' ios 7 Pei roles ie G4 ple oo are at cee lari dvenoh Dey, g ie _ re. wnt) oko Tea en Ras pipe» alae ith bs ae gone ul eras peer Del eam (anemone Sle erway all Ladle e TOL LIT AT nae ams te i tte a) a os : ‘ net PRA ee i Bua it ened il nae ? ; ‘ : Wet. yD ee. a : vic rit Tie) have oa 4 . ? , a i + A rn Ce ae oe ee i ‘ iw -? yt ‘hae a \ 4 re + \ 4 ) ' . "? (4 ‘ | @ - yeas 7 . me ' he i ' he : a & J | | ih CNS y i } @ w ks , , ‘ { | * Fi » s ; * - I: a, J 7 a - - 1k a . - % . Fe * Oi ae ® aa nny ; : < : u 7 : N A A ie eee) oo) ii VIII. CONCLUSIONS CULTURAL CONTINUITIES Settlements—The earliest communities of the Florida Gulf were shellfishing stations along the shore or the immediate tributary bays and streams. These are marked by shell-midden piles. Except for a single site of the preceramic period (see pp. 327-328), those of which we have knowledge can be divided into two groups. In the northwest are the small shell-refuse sites of the Deptford Period. These are without works such as artificial mounds, embankments, or fortifica- tions. Around Tampa Bay and farther south are the shell-midden remains of the Perico Island Period. The Perico Island Period is thought to have been coeval, in its earlier phases, with Deptford. Perico Island sites are larger than those of the Deptford Period and often of great depth. The affiliations of Perico Island ceramics are with the Glades area, and the whole culture has ties with south Florida rather than the central west or northwest Gulf Coasts. At some of the Perico Island sites there are artificial mounds and em- bankments. A problem is raised as to whether these features can be considered as coeval with the Deptford Period or whether they are somewhat later in time and are the result of Santa Rosa-Swift Creek or Weeden Island influences upon Perico Island. I am inclined to be- lieve that the latter interpretation is the correct one. This would mean that the Perico Island culture continued later than the Deptford Period, and, in doing so, underwent changes in settlement pattern that reflect those of the succeeding culture period of the northwest and central west coasts, the Santa Rosa-Swift Creek. Santa Rosa-Swift Creek settlements are small coastal and riverine middens, probably only a little larger than those of the Deptford Period but not as large as those of Perico Island. Santa Rosa-Swift Creek sites are much more numerous than those of the preceding periods, however, and they are characterized by a new feature. It is during this period that the first burial mounds appear. These mounds are generally made of sand and are located near but not immediately upon the living sites. Not every Santa Rosa-Swift Creek midden was accompanied by a burial mound, and it is estimated that an average of three small villages combined to build a single mound. Sites of the Weeden Island Periods are virtually identical with those of the preceding Santa Rosa-Swift Creek culture. They are 36 537 538 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I13 small middens, usually less than 100 meters in diameter and about .50 meter in depth. They are more numerous than the Santa Rosa- Swift Creek sites and are found in all three subareas or regions of the Gulf Coast: the northwest, central coast, and Manatee. The sand burial mound, introduced in the Santa Rosa-Swift Creek Period, is common ; and it is again estimated that a burial mound was maintained by approximately three villages. The continuity of settlement pattern seen running through Santa Rosa-Swift Creek and Weeden Island is broken in the Fort Walton Period in northwest Florida. Small coastal village sites are much the same size and have much the same ecological relationships as in pre- ceding periods. They are, also, about as numerous as Weeden Island village sites. The important difference is noted in the appearance of flat-topped temple mounds and temple-mound sites. If an average of three villages joined together to construct a burial mound in the Santa Rosa-Swift Creek and Weeden Island Periods, it is certain that several times this number converged to build a Fort Walton Period temple- mound site. These sites were undoubtedly central settlements or “capitals.” This is inferred from the archeology and is supported by sixteenth-century accounts of the Indians in this region. In the Fort Walton Period the burial mound virtually disappeared, but southward, in central west Florida and the Manatee region, the old Weeden Island pattern of the burial mound as the nucleating ceremonial site continues in the Englewood and Safety Harbor Periods. Temple mounds appear in the south at this time, but are not common. Safety Harbor village middens are about the same size and have about the same type of locations as those of the Weeden Island Periods. The final phase of aboriginal settlement is seen in the Leon-Jefferson Period of northwest Florida. By this time the Indian populations were grouped around Spanish missions and forts. In the southern part of the Gulf area the missions were not established with the same success as they were in the north. To recapitulate the continuity of prehistoric Gulf Coast settlements we can begin by following the simple village site from earliest to latest times. This is the basic community and undoubtedly was the funda- mental economic unit. These villages were probably small tribes or subtribes in which most of the individuals were held together by blood ties. In Gulf Florida these villages were usually located near the sea or on important rivers and were fishing stations. Around Tampa these sites were larger in the early periods than they were in the northwest, probably as the result of a greater abundance of shellfish in the south. In the Santa Rosa-Swift Creek and Weeden Island Pe- WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—-WILLEY 539 riods the basic village community continues much the same except that communities begin to cooperate in small groups of two, three, or four for the purpose of building ceremonial sites or burial mounds. Still later, in Fort Walton times, the small village community con- tinues, but social integration has reached a point where a considerable number of such communities band together to build and supply temple- mound capitals. These capitals were, in part, supported by the In- dians who dwelt there permanently, but they were also sustained by other Indians of the greater tribe who lived at some distance from the temple-mound capital. This idea of the temple-mound capital had penetrated into the central Gulf Coast in the Safety Harbor Period, but it was not as well established there as it was in the Fort Walton Period of the northwest coast. Native settlement patterns organized out of small villages but around temple-mound centers collapsed with the founding of Spanish missions in northwest Florida during the Leon-Jefferson Period. The missions then became the nucleating centers. Economy.—The very nature of village or living sites throughout the prehistoric sequence in Gulf Florida indicates that shellfish and other marine foods were an important part of the Indian diet at all periods. In the Deptford and Perico Island Periods the economy was oriented toward the sea plus the hunting and gathering of some land animals and plants. In Santa Rosa-Swift Creek times we have postulated the appearance of maize agriculture. This is purely in- ferential and follows out of more general interpretations of eastern United States prehistory. Even with maize, and perhaps squash and beans, hunting and fishing remained as important food-gathering tech- niques throughout Santa Rosa-Swift Creek. Economy in Weeden Island times was, presumably, about the same as Santa Rosa-Swift Creek. In the light of an increase in the number of sites it may be that agriculture was more widely practiced than previously ; however, shellfishing, fishing, and hunting continued as supplementary modes of subsistence. The Fort Walton Period we know was agricultural. Sixteenth-century Spanish accounts tell of crops of maize, beans, and squash and of storehouses filled with such produce. The gathering of fruits and wild plants, fishing, and hunting were all practiced, but there are several things which suggest that agriculture became more important during this period than it had been before. First, there is the change in settlement pattern with the rise of the big towns or temple-mound centers. These new population concentrations, larger than those of the previous periods, could have been supported in the Gulf Florida environment only by intensive agriculture. Sec- 540 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. II3 ond, there seems to have been a population shift in Fort Walton times from coast to interior. The largest sites of the period are found inland in the better farm lands. This was attested by the sixteenth-century explorers and is also reflected in the archeology. Such a shift from littoral to interior might have been conditioned by other factors, but the rising importance of agriculture must have played a part. And third, the Fort Walton culture has extra-Floridian relationships with other prehistoric cultures which compose a late southeastern horizon characterized by an intensive agricultural economy. Food economy for the Safety Harbor Period was probably some- where intermediate between that of Weeden Island and Fort Walton, while Leon-Jefferson economy must have been much the same as Fort Walton. Society——The social organization of the early periods, Deptford and Perico Island, was probably coextensive with the village units. That is, each living community was politically and religiously, as well as economically, independent. At this time there seems to have been little group expression of a ceremonial nature. We are judging, of course, from the archeological record in which we see little in the way of advanced artistic or craft expression and no large monuments of any kind. In the subsequent Santa Rosa-Swift Creek and Weeden Island Periods the burial mounds are evidence of an organizational structure over and above the simple village unit. Provided we are correct in our assumptions, intervillage cooperation was involved for the task of constructing and dedicating a mound. The religious or sacred quality of these burial mounds cannot be doubted. Considering the economic and technical level of the Santa Rosa-Swift Creek and Weeden Island Periods, the Indians were lavish and extravagant in their veneration of the dead. Their mortuary goods were among their finest craft products. Such emphasis suggests profound religious motives in the society. There are virtually no archeological clues as to the nature or prevalence of warfare. If common it is likely to have consisted of village raids and counterraids organized on a small scale. Although sixteenth-century accounts of north Florida Indians are replete in their descriptions of class differentiation, the archeologi- cal story covering the centuries previous does not point this up. Dif- ferences in burial treatment are noted in some of the mounds with certain individual skeletons being accorded a central position in the mound, a special grave, and somewhat richer burial furniture. This is more true of Santa Rosa-Swift Creek than the later Weeden Island Periods. Nearly all Weeden Island burial mounds contained WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—WILLEY 541 mass deposits of grave goods, seemingly placed as a common offer- ing to all the dead within the mound rather than for any individual. There is a continuity in the actual treatment of the body between Santa Rosa-Swift Creek and the Weeden Island Periods. Secondary inhumation was practiced in both but was probably more common in Weeden Island. We know very little of earlier burials except that the Perico Islanders practiced primary inhumation and that there are two instances of cremated cemetery burials in the Deptford Period. This latter practice was only rarely carried over into the later periods. There are evidences of greater change in the organization of so- ciety between Weeden Island and Fort Walton than between Santa Rosa-Swift Creek and Weeden Island. The new settlement pattern, structured around the temple-mound sites, had a basis in larger po- litical groupings than were known in previous periods. The Spanish explorers remark upon the concepts of nationality that the Indians of the Gulf Coast possessed, and these early travelers also describe the crossing of well-defined territorial boundaries of large tribes or nations. Groups of several small villages were clustered around each of the principal towns, and among the Apalachee we know that several of these towns were bound together by a supertribal govern- ment which had a recognized capital. Warfare, between tribes, was an established institution and governed by strict rules. Among the Timucua, religious prerogatives were closely bound up with mili- tary and civil privileges and duties, all centering in the hands of the chiefs. Government could, in this sense, be considered theo- cratic, but it is likely that the Fort Walton Period rulers exercised greater secular powers than their Weeden Island predecessors. Rites for the dead are, relatively, less elaborate for Fort Walton than in the preceding periods. Burials were often placed in cemeteries or in the floors or sides of the temple mounds. There is more individual emphasis on grave goods than in Santa Rosa-Swift Creek or Weeden Island Periods; but, with the disappearance of the burial mounds, one cannot help but feel that burial ceremonialism was less of a social integrative force than formerly. The Safety Harbor Period probably represented a transitional stage between Weeden Island and Fort Walton societies. Leon-Jefferson demonstrates great change on the social organiza- tional level. It is likely that many forms of tribal and group allegiance dissolved when the Indians came under the spiritual and political over- lordship of the Spanish. Chieftainship of a sort remained but without the religious sanctions that bolstered the position under aboriginal conditions. 542 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I13 Technology.—The appearance of maize agriculture at the beginning of the Santa Rosa-Swift Creek Period, if we are correct in placing it at this point in the culture sequence, was the single most revolu- tionary technological change for the Indians of Gulf Florida. It is paralleled by a number of cultural changes, but by only one other important technological innovation. This last is the ground- and polished-stone ax or celt. The ground-stone ax is the classic Neolithic type specimen in the Old World, generally considered the clearing and cultivating tool and virtual symbol of early agriculture. In the archeology of the eastern United States some forms of ground-stone axes, particularly grooved axes, are a part of the Archaic horizon which is generally thought to precede Eastern maize horticulture. We know, too, that stone grinding and polishing techniques were part of Archaic stage crafts in many parts of the eastern United States. Nevertheless, in most places in the East, and certainly in Florida, the polished-stone celt is not a part of the Archaic horizon. Its sudden appearance on the Santa Rosa-Swift Creek time level marks a definite technological advance. Most of the elements of Gulf Coast aboriginal technology known in the middle and late periods were present at the inception of our sequence.** The “paleolithic” technology of stone chipping for the manufacture of points and knives, the “neolithic” art of pottery making, and the skills of working in shell, wood, and probably bone, were all Deptford or Perico Island Period accomplishments. From this time forward these technologies can be traced through their respective continuities of change and modification. These are changes which result from the artisan’s more complete mastery of his craft, from the introduction and development of minor techniques, and through the media of varied stylistic expressions. Continuity can be followed with relative ease in ceramic tech- nology. In the quality of pottery ware the basic standards were es- tablished in the Deptford and Perico Island Periods. These stand- ards resulted in the production of a coiled, sand-tempered, even and granular pottery which had been fired with erratic air control to a fairly hard consistency. The sand-tempered Deptford and Perico wares represent a definite break away from the old fiber-tempered ceramic of the Southeast which may have preceded Deptford and Perico Island on the Gulf Coast. From the Deptford Period on, 84 Metals, worked by cold-hammering techniques, are an exception, as these do not appear until the Santa Rosa-Swift Creek Period. It is likely that the Florida metal specimens of this and the Weeden Island Period were imports, not local products. WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—-WILLEY 543 pottery ware qualities remain constant until the Fort Walton horizon. Santa Rosa-Swift Creek pottery is a little finer and thinner than Deptford, but ware differences are generally slight, and Weeden Is- land presents no marked temper or paste contrasts to Santa Rosa- Swift Creek. Much of the Fort Walton Period pottery is also sand- tempered, but coarser grit temper makes its appearance in this period. The most noticeable innovation, however, is the use of crushed shell as temper. These changes characterize Fort Walton and Leon-Jeffer- son but not Englewood or Safety Harbor which continue in the Weeden Island ware traditions. Vessel forms show more inclination to rapid change than the ma- terials of manufacture and conditions of firing. Deptford vessels were virtually all simple pot forms with unmodified rims, and occa- sional tetrapodal supports,®° and Perico Island vessels were open bowls with slightly incurved rims. Both of these forms are seen in all the later periods; however, during the succession of periods a great many new forms are introduced. For example, in Santa Rosa- Swift Creek the old Deptford pot form is quite common, but, in addition, we have a great number of new shapes including flattened- globular bowls, boat-shaped bowls, beakers, composite-silhouette vessels, and collared jars. There are many variations on these new basic forms such as vessels with squared bodies and bases. Besides the principal forms, the Santa Rosa-Swift Creek Period is noted for unusual or exotic vessel forms. Two of the most noticeable are the multiple-orifice vessels and the efhgy forms. One feature which may be a carry-over from the Deptford Period is the use of tetra- podal supports. These were rare in Deptford but frequent in Santa Rosa-Swift Creek. In Weeden Island I the potters continued to create the strange, exotic nonutilitarian forms first noted in Santa Rosa-Swift Creek. By Weeden Island II Period there are fewer of these odd forms, but most vessels conform to the basic innovations of Santa Rosa-Swift Creek. There is one notable discontinuation, however ; tetrapodal supports became extinct in the Weeden Island Periods. With the Fort Walton Period, forms are more limited in variety than was the case for Santa Rosa-Swift Creek or Weeden Island. The two new and characteristic Fort Walton forms are the bottle and the casuela or carinated bowl. It is not clear as to just what degree these two new forms have been influenced by and are a development out of previous Weeden Island vessel shapes. Certainly 85 There is one boat-shaped bowl and one double vessel dating from the Deptford Period. See Carrabelle (Fr-2) site discussion (pp. 267-268). 544 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I13 the bottle bears some relationship to the collared jar and the casuela bowl to the flattened-globular bowl. At the same time, there is also reason to believe that both may be new introductions to Gulf Florida. Data are still insufficient to resolve the problem of ceramic con- tinuities between Weeden Island and Fort Walton. One feature of the Fort Walton Period is, however, definitely new. This is the strap-handle attachment. Safety Harbor Period vessels conform to both Weeden Island and Fort Walton shape styles while Leon- Jefferson vessel shapes combine ideas seen in Fort Walton with others, like the annular-based plate, that appear for the first time in this very late period. Pottery decoration presents some very interesting sequence con- tinuities. In the Deptford Period the almost universal mode of decoration was by stamping. The stamped designs were simple linear arrangements or check patterns. There is one Deptford example of a design worked out in linear punctation (see Carrabelle site (Fr-2), pp. 267-268). In the southern part of the Gulf Coast area stamping seems to have been known only on trade pottery during the Perico Island Period. The little Perico Island pottery that is decorated has been treated with fine-line incisions or linear puncta- tions similar to those noted on the single Deptford Period exception. In the succeeding Santa Rosa-Swift Creek Period check and simple stamping continue, with some modifications, from the Deptford Period; but the outstanding stamped type is of a complicated recti- linear or curvilinear design pattern. This is the pottery of the Swift Creek tradition, and this type of decoration is closely associated with the pot-form vessel. The other major element in Santa Rosa-Swift Creek Period ceramic decoration is the complex of incision, puncta- tion, rocker stamping and occasional red painting. All appear to be new to the Gulf Coast area and not derived from the Perico Island incision or linear punctation. These modes of vessel decoration have a close association among themselves and are also associated with the new ves- sel forms which appear at this same time. Thus, in both vessel shapes and vessel decoration the Santa Rosa-Swift Creek Period was one of fusion of two rather strikingly different sets of ideas: on the one hand, the conoidal-based pots decorated with the stamping tech- nique; and on the other, globular bowls, beakers, collared jars and unusual forms decorated with incision, punctation, rocker stamping or red zoned painting. One other element of pottery decoration which appears for the first time in this period, and which probably has antecedents separate from either the stamped or incised and WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—-WILLEY 545 rocker-stamped traditions, is negative painting. This is true negative color application involving the resist-dye process. In the Weeden Island Period both the incision punctation and the complicated stamping continue. The first is predominant, and com- plicated stamped pottery lasts until the beginning of Weeden Is- land II. Both techniques are stylistically modified, but the develop- mental relationships to Santa Rosa-Swift Creek pottery are obvious. The typical Weeden Island ware is incised or punctated. Rocker stamping, as it was seen in the Santa Rosa-Swift Creek type, Alli- gator Bayou Stamped, has virtually disappeared, with only a sem- blance of the technique retained in the Weeden Island type, Little Manatee Zoned Stamped. Negative painting does not appear in the Weeden Island Periods or again in the Florida Gulf sequence. One prominent mode of decoration becomes popular in the Weeden Island Periods. This is modeling. Usually it takes the form of zoomorphic adornos attached to vessel rims or walls. More rarely, it is ex- pressed in complete effigy shapes. A few of the latter were noted in Santa Rosa-Swift Creek, and it may be that modeling-type decora- tion had its beginnings in that period. Its great vogue, however, was in the Weeden Island Periods. A consideration of the modeling technique leads us into the prob- lem of continuity of design as well as decorative techniques. The common form of Weeden Island Period modeled adornos is the bird. Usually, there are bird heads affixed to the rim of a bowl or jar. Quite often, what appear to be the wings or tail of the bird are indicated on the vessel in incision or punctation techniques. Such designing is highly stylized and difficult to interpret in naturalistic terms. Similar designs, perhaps a little less stylized, also occur in the preceding Santa Rosa-Swift Creek Period, presumably as proto- types for the bird-motif designing that is so typical of Weeden Island. It is quite likely that the Weeden Island artist transposed the bird motif into the modeling technique. Besides the bird design another design idea was also transferred from Santa Rosa-Swift Creek to Weeden Island. This is the rendering of design in the negative. Often a Weeden Island design proper is left plain and only the background is filled in with incision or punctation. This is a method of design expression used in Santa Rosa-Swift Creek incised, rocker-stamped, and punctated types as well as the negative painting of that period. By the dawn of the Fort Walton Period the stamped techniques of decoration had almost completely disappeared in northwest Florida. In central west Florida and the Manatee regions check 540 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. II3 stamping continued into the Safety Harbor Period, and occa- sional late complicated stamped types are also seen. Most Fort Walton pottery is, however, decorated by incision and punctation, modeling, or both. As with the vessel shapes, we are undecided as to whether this represents a continuity from the Weeden Island Period or is a new introduction from outside of the Gulf area. Probably both continuity and outside contact are involved. Some Fort Walton designs, such as the negative meander, suggest con- tinuity from the earlier local periods. Bird-head rim effigies are also very similar to the Weeden Island ones. In the Leon-Jefferson Period incised and punctated techniques and designs carry over from Fort Walton, although these are reinforced by ideas impinging upon northwest Florida from the south Georgia area. A revival of the complicated stamped technique and complicated stamped designs reminiscent of those in Swift Creek and related types can best be explained as a reintroduction of pottery decorative ideas that pre- viously had been extinguished in Florida. The sporadic appearance of these complicated stamped types in Safety Harbor Period sites, farther to the south, are accounted for in the same way. Technology in stone offers a less clear developmental picture. Adequate period identification data on most projectile-point collections makes it difficult to give a concise summation. In chipped stone large and medium-size stemmed projectiles were made during Dept- ford times, and these continue in the later periods. In Santa Rosa- Swift Creek more and finer blades and points were manufactured. Some of the large lanceolate blades of the period are among the finest chipped-stone tools or weapons made at any time or place in Florida. Large fine blades are also found in Weeden Island Period mounds. Stemmed points, triangular and ovate-triangular bladed, belong to the Weeden Island horizon; small, triangular, unstemmed points appear for the first time; and scrapers, drills, and knives were also manufactured. Both large and small triangular points are present in Fort Walton and Safety Harbor sites. In Safety Harbor chipped-stone work is more common than in any of the other periods. Ground-stone work of any consequence has its beginnings in the Santa Rosa-Swift Creek Period. The appearance of the stone celt has been mentioned as an important technological introduction. These celts, made of hard imported stone, were of the pointed-poll variety. The body of the celt is rounded or ovate in cross section. The pointed-poll celt carries on into the Weeden Island and Fort Walton Periods but is rare in Safety Harbor. In the Fort Walton Period a new celt form appears in addition to the pointed-poll type. This is a WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—WILLEY 547 proportionately thinner, flatter, and more rectangulate shape. Stone plummet-type pendants, grooved for suspension at one or both ends, are very common to Santa Rosa-Swift Creek, somewhat less so to Weeden Island, infrequent in Safety Harbor, and absent in Fort Walton and Leon-Jefferson. In general, articles like stone beads, bar amulets, stone gorgets, stone pipes, and rock-crystal ornaments were more usual in Santa Rosa-Swift Creek than in Weeden Island. The presence of raw sheet mica in burial mounds is characteristic of both Santa Rosa-Swift Creek and Weeden Island, while the use of sheet mica for fashioning symbolic or ornamental objects is a Weeden Island trait. Mica is not found in Fort Walton, Safety Harbor, or Leon-Jefferson sites, nor is ground-stone work of any sort, except for the celts, typical of these later periods. One rare exception is the stone discoidal which occurs in Fort Walton for the first and only time in the sequence. Shellwork has its inception in the Perico Island Period, but, as far as we know, was not characteristic of the contemporaneous Deptford Period in the northwest. Perico Island shell weapons or tools include the Busycon hammer or pick, Strombus hammers, and Strombus and Busycon celts.** Shell was also utilized during this period for the manufacture of plummet-type pendants and shell beads. All these artifacts are Glades area types. Busycon hammers, Strombus celts, and the plummet pendants appear in the Santa Rosa-Swift Creek Period. Besides these, Santa Rosa-Swift Creek sites also show chisels and adzes made from conch-shell columellae, shell gorgets, shell beads, and conch-shell drinking cups. All are Glades area arti- fact types although we have no specific record of them for the Perico Island Period. There is less shellwork in the Weeden Island Periods than in Santa Rosa-Swift Creek. Celts and hammers are absent, and shell beads, punches, and chisels were found in only a few sites. The shell cup is the only ubiquitous shell artifact of the period. In Fort Walton there are none of the larger shell tools, and the shell drinking cup is a rare item. Shell chisels or punches, spike-form ear pins, and beads are all present, however. Safety Harbor differs markedly in that Busycon picks and hammers, shell plummet-type pendants, shell beads, and cups of shell were found in most of the mounds of the period. In retrospect, this is a curious, broken con- tinuity for shellwork. Apparently shell tools were very early in the 86 Jt is likely that the Strombus celt was a Perico Island Period imitation of the stone celts of the north. This also suggests that the Perico Island Period lasted until Santa Rosa-Swift Creek times, the earliest appearance of the stone celt in the north. 548 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I13 south but did not appear in the northern part of the area until Santa Rosa-Swift Creek times. They are not important in Weeden Island or Fort Walton sites, but in the south they are revived in the Safety Harbor culture. This suggests a double diffusion from the south of shell implements or the techniques of making these implements, the first dating began from the Perico Island Period and a secondary late diffusion on the Safety Harbor level. Craft work in bone, horn, or teeth does not, in general, characterize the Gulf Coast area. In other areas of the Southeast bone implements and decorated bone were noteworthy developments of the Archaic horizon. Possibly such finds will eventually be made in west Florida. What bonework is known now comes mainly from the Santa Rosa- Swift Creek and Perico Island Periods. This consists of bipointed projectiles, awls, and daggers, all made from mammal bones. Some bone projectiles, bone fishhooks, bone gorgets, turtle-shell rattles, per- forated teeth, shark teeth, and cut animal jaws are associated with Santa Rosa-Swift Creek sites. There is only an occasional bead or fishhook found in the later periods. Probably one of the most highly developed crafts of the Gulf Coast aborigines was wood carving. Our knowledge of it is limited to a secondary record, the impression of carved designs on pottery. Some wood carving was known in the Deptford Period for the checked and simple stamped pottery was, apparently, impressed with carved wooden stamps. No inferences can be made about Perico Island woodwork, but in the Santa Rosa-Swift Creek Period carving in wood may have enjoyed a sudden boom. Pottery of the period is decorated with intricate stamped designs produced with the aid of a carved wooden stamping unit. The carver’s art may have declined in the Weeden Island, Fort Walton, and Safety Harbor Periods. At least, the intricate complicated stamped designs on pottery disappear at the close of Weeden Island I. Some carving was done, however, as we have direct evidence in a small fragment of gracefully carved charred wood taken from a Safety Harbor Period mound. Pottery decorated with inferior stamped designs appears in the Leon- Jefferson Period. Metalwork makes its first appearance in the Santa Rosa-Swift Creek Period. This is cold-hammered copper, meteoric iron, and silver. Ear ornaments, copper tubes, embossed plates, and cut-out designs on plates comprise the metal artifacts. The metals from which they were made are, obviously, imports. It is questionable as to whether or not the actual objects were made in Gulf Florida or made elsewhere and imported. Their similarity to Ohio Valley forms suggest the latter. WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—WILLEY 549 Copper or metal objects are much less common in the Weeden Island Periods. European, or European-transported, metals come in during the Fort Walton and Safety Harbor Periods. Sometimes these have been reworked locally by hammering and engraving. Some of the re- worked metal objects from these periods show considerable skill, and this argues that perhaps metal craftsmanship was known in Gulf Florida in the earlier periods. The argument, however, is weakened by the lack of a strong continuity in aboriginal metal products from the Santa Rosa-Swift Creek Period, through Weeden Island and into the late periods. PEOPLES (By MarsHatt T. NEwMaAn) An analysis of Floridian Indian physical types in time and space is difficult with the present data. These data consist of Hrdlicka’s two reports (1922, 1940) ; Von Bonin and Morant’s (1938) biometric analysis of Hrdlicka’s 1922 measurements; and our own hasty ex- amination of several cranial series in the United States National Museum. Hrdlicka’s earlier study was based upon 173 (121 male, 52 female) undeformed Florida skulls and a few long bones, most of which came from the lower portion of the Central region and the upper portion of the Manatee region of the Gulf Coast. The total series for his later study was expanded to 426 (232 male, 194 female) undeformed skulls, with the bulk of them still from the middle Gulf Coast. Thus the Northwest, Central peninsular, Atlantic Coast, and Southern areas were meagerly represented. This spotty distribution, coupled with his almost complete disregard of chronological pegs upon which he might have hung his series in the 1940 report, lays open to question certain parts of Hrdlicka’s conclusions on the peopling of Florida. It was Hrdlicka’s opinion that the native population of Florida was a robust Indian group in which two distinct, although intermixed, physical types could be discerned. The numerically predominant type, in terms of his series at least, was brachycephalic, extremely high- headed, of medium stature (males 165-168 cm. or 5 feet 5-6 inches; females 152-154 cm. or 5 feet), and possessing exceptionally rugged faces and lower jaws. Hrdlicka identified this type with the Timucuans of the north half of the State. He considered these people as simply a southern extension of the Gulf or Centralid type of the southeastern States, but attributed the exceptional bony ruggedness of the Flo- ridians to their marine diet which was undoubtedly extra rich in phos- phates. The minority type was essentially mesocephalic and charac- 550 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. II3 teristically less massive. It occurred only as a rarity in Gulf Coast and St. Johns area series, but on the slim basis of three small series,®” Hrdlicka felt these mesocephals were more prevalent in the northwest and south of Florida. The lack of a “pure” mesocephalic series in his collections was explained by long-time intermixture between the two Floridian types, although he (Hrdlicka, 1922, p. 89) showed some confidence that future excavations would reveal a “fairly pure’ meso- cephalic group. He related this mesocephalic type to Muskogean- speaking intruders, and specifies the Seminoles. The type was con- sidered a modification of the Algonkin or Silvid group to the north. Hrdlicka believed that the brachycephalic group represented the indigenous population, and that the mesocephals entered the peninsula at a later time. He guarded himself, however, by offering two other explanations, which he thought less probable. The first was that the brachycephalic people invaded the lower southeastern States from Mexico, intermixed with the longer-headed Muskogean-speakers in Alabama and Georgia, and in that way brought the mesocephalic ele- ment into Florida. The second explanation considers the remarkable head height, and inferentially some of the other Floridian character- istics, a local development induced possibly by special environmental factors. Von Bonin and Morant, in their reanalysis of Hrdlitka’s 1922 measurements, have added three points of interest to the present syn- thesis. In the first place, sensing the arbitrary nature of Hrdlicka’s two Floridian types, they compared his total Gulf Coast series with a pooled series of all others from Florida, including the Seminoles. None of the mean differences were found to be statistically signifi- cant. Secondly, they calculated standard deviations for the total Florida series, and found its variability to be quite unexceptional. From the data at hand, they saw no justification for subdividing the total series into types, although they felt that more abundant material might make it possible to distinguish regional and time differences. Thirdly, they used the Coefficient of Racial Likeness to compare the total Florida male series with the Central Californian (mostly Yokuts) series. This led them to note that both Central Californians and Floridians were placed by Von Eickstedt (1934) in his “gruppe Margide” or Marginal type. While there is an attractive neatness in having geographically and culturally marginal areas occupied by 87 Of 8 skulls from Santa Rosa Island near Pensacola, 1 is dolichocephalic, with the rest showing indices over 80. Among 8 skulls from Canal Point east of Lake Okeechobee, 2 are dolichocephalic; 3 more have indices under 8o. In a Seminole series of 13, only 1 shows an index over 8o. WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—WILLEY 551 racially marginal peoples, it will take more than the pat “efficiency” of the Coefficient of Racial Likeness and the gross, sweeping coverage of Von Eickstedt to demonstrate such a proposition. One rather ob- vious objection is that while Central Californians were notably low- headed, Floridians were exceptionally high. Stewart (1940) has shown the considerable diagnostic value of head height as a racial determinant among Indians, pointing out that generally high-heads were early and low-heads late in the sequences of Indian physical types. Further researches may indicate other objections to a close racial affinity between so-called Marginal peoples, although such re- lationships are hard to deny on the basis of present knowledge. Returning to Hrdlicka and his two Floridian physical types, we must agree with Von Bonin and Morant on the arbitrary nature of the mesocephalic group. Without arguing against some separateness for his Seminole series, we feel that the other Florida mesocephals are more likely individuals on the lower end of the index range in the brachycephalic group than representatives of another type. Our rea- sons for doing so are twofold. First, Hrdlicka made little attempt to document the distinctiveness in other ways of the few mesocephals he selected ; and second, within a series showing a mean length-breadth index of 79-81—a good mean range for Florida—the individual varia- bility would be from about 74-87. If the longer-headed individuals fail to show differences of a parallel nature in the face and other areas, their separation as a distinct type is hard to justify. All this is not to deny the very real possibility of several different physical types in Florida’s racial prehistory. Rather we submit that Hrdlicka’s mesocephalic type requires validation beyond the “opinion of ex- perience.” In addition, the identification of the mesocephalic type with Muskogean-speakers may be questioned. Certainly bearers of the Fort Walton culture, the Apalachee, were the first quite definite Muskogean peoples in Florida. Culturally their influence extended south only to about the Aucilla River. For them to have directed mesocephalic genes to south Florida, without perceptibly influencing the central part of the State, stretches one’s credulity. If we are cor- rect in the foregoing, Hrdlicka’s mesocephalic type as well as his source for it are to be seriously questioned. Hrdlicka (1922, p. 89) stated that there was no evidence of any earlier Floridians than the brachycephalic population. The recent re-evaluation of the controversial Vero and Melbourne finds by Stewart (1946) raises the very good possibility of an early dolicho- cephalic population living on a Paleo-Indian or, at least, Archaic level. Even if a Pleistocene dating for Vero and Melbourne is questioned, 552 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I13 they are probably entitled to greater antiquity than the earliest sites from which Hrdlicka obtained his series. Now if there was a Paleo- Indian or Archaic population of dolichocephals in Florida, there is no apparent trace of it in Hrdlicka’s series, where only 4 percent are long- headed and in our opinion simply represent one morphological extreme of the brachycephalic range. This brings us to a point where we can develop a hypothesis of our own, more in accord with the present cultural chronology in Florida, and with the sequence of physical types elsewhere in the Southeast. The Vero and Melbourne skulls suggest an early dolicho- cephalic population in Florida, just as has been amply substantiated in Alabama (Newman and Snow, 1943) and Kentucky (Snow, 1948). One of the most likely places to look for more skeletal remains of these people would be in the large Archaic prepottery middens in the St. Johns area. The time of entrance of the brachycephalic people into peninsular Florida is hard to gauge, but it probably begins on a Santa Rosa-Swift Creek or St. Johns I level. Certainly they had arrived in South Florida by the equivalent Glades I-II periods, judg- ing by Hrdlicka’s measurements and our inspection of the sizeable Perico Island series. Hrdlicka’s measurements on the Belle Glade skulls (Glades II-III) also indicate brachycephals in the Glades area. Within the Central Gulf Coast region, the small and rather frag- mentary Weeden Island series cannot be distinguished from the Perico Island skulls, either by Hrdlicka’s measurements or our own inspection. The same is true for the larger Safety Harbor series. Thus, as far as we can determine, there does not seem to be any change in physical type along the Central Gulf Coast and Manatee region from Perico Island to Safety Harbor times. The more exact racial position of these series—as well as the whole Florida brachycephalic group—is a problem we cannot solve at present. With little doubt, the similarities in conformation of the vault link the Florida brachy- cephals most closely with the Gulf or Centralid people of the South- east. The differences in the facial skeleton, which are largely in the direction of greater massiveness in the Florida group, may be of racial or subracial proportions, or as suggested by Hrdlicka (see Pp. 549) may be due to a diet especially rich in phosphates. Needed to settle this question is a detailed comparison of Florida series and those from the Muskogean-occupied portions of the Southeast. This comparison would contrast the early brachycephals in the South- east, presumably as exemplified by the Florida series, and the later brachycephals. Since cranial deformation is a culture trait recorded in bone, it has WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—WILLEY 553 become part of the province of the physical anthropologist to observe it. Hrdlicka’s earlier (1922, pp. 83-85) summary observes that all Florida cranial deformation is of two types, intentional fronto- occipital or presumed intentional occipital. He also adds that both the frequency and degree of head flattening diminish on the Gulf Coast from northwest to southeast. To this we can add the chron- ological generalization that the trait seems to be late rather than early. Only one occurrence was noted in Moore for a Santa Rosa- Swift Creek-Weeden Island I burial mound ; two come from Weeden Island mounds, undifferentiated as to period; three flattened skulls are reported from Weeden Island II mounds;** and two Fort Walton Period burial sites are known to show the trait. Cranial de- formation was not recorded for Safety Harbor or Englewood Periods. It is not clear, however, how many of these instances are cases of frontooccipital (definite) deformation and how many are only of the simple occipital (probably accidental) types. A review of Hrdlicka’s later crania studies on Florida (1940) reveals no in- stances of frontooccipital flattening of an undisputed intentional nature. That true frontooccipital flattening did occur seems at- tested to, however, by an illustration of a skull from the Sowell Mound (Weeden Island Period) in Moore (1902, fig. 67). As far as available data go it is, perhaps, most significant that cranial deformation never became a widespread trait in Florida in spite of its common occurrence in the middle and late horizons of the Southeast. EXTRA-AREAL CONTACTS AND AFFILIATIONS The south—To the south of the Florida Gulf Coast area is the Glades archeological culture area. It comprises the southern end of the Florida peninsula from a point at Boca Grande Pass on the west to Fort Bassenger on the Kissimmee River and, thence, to St. Lucie Inlet on the east coast (Goggin, 1947b, p. 119). The Glades is a low- lying region of coastal mangrove swamps and interior grassy marshes. It was, and in some places remains today, a semitropical refuge wilderness. For early American maize horticulturists the Glades country was not a favorable environment; but for a people whose economy was based upon hunting and fishing it offered resources and advantages. Most archeologists are of the opinion that the Glades area was a cultural cul de sac, receiving but not reciprocating in the diffusion of 88 These counts do not include cases of slight, questionable flattening in the Weeden Island mound proper (Pi-1) or the Thomas site (Hi-r1). 37 554 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I13 ideas among the prehistoric Indians of Florida and the Southeast. Continued research in the Glades and neighboring areas has demon- strated this to be largely true, although there are some exceptions. Connections between the Glades area and the Gulf Coast probably go back to preceramic levels. Relationships between Gulf Coast and Glades at this early time, if such existed, must have been those of a general cultural homogeneity. We would expect that these early fishers, hunters, and shellfish gatherers were closely related peoples who migrated from north to south along the shore until a number of villages, each possessing a culture very similar to the next, were founded all along the Gulf Coast and well into the Glades area. The earliest specific evidence we have for contact between Gulf Coast and Glades follows the preceramic and fiber-tempered pottery eras and manifests itself on a Deptford-Perico Island-Glades I horizonal equation (see fig. 76 for chronological alignments). On this time level it is presumed that all the culture periods involved were still prehorticultural. Regional differences were just beginning to emerge and stabilize out of the cultural homogeneity of the pre- ceramic and fiber-pottery periods, and contact and interchange between Glades and Gulf Coast was probably by way of diffusion of ideas and trade rather than actual mass migrations of people. Glades culture at this time extended farther north than it did in later periods, over- lapping into what we have defined as the Gulf Coast area. Perico Island is this northernmost extension and variant of the Glades cul- ture and as such equates with the Glades I and Glades II Periods as they have been established farther south. The boundary line between the Perico Island or early northern Glades-like culture and the Deptford culture of northwest Florida is not definitely known, but Perico Island sites are found as far north as Pinellas County. A few Deptford sherds have been found in Perico Island sites, but no Perico Island sherds or Glades-type shell tools are known from the Deptford sites to the north. Cultural contact as it existed must have been from north to south to judge by the limited evidence available. A little later the exchange of artifacts and ideas between Gulf Coast and Glades seems to have received stimulation. This is seen in contacts between the Perico Island culture, which evidently lasted on after the close of the Deptford Period, and the new period of the north, Santa Rosa-Swift Creek. The burial-mound idea probably was diffused south at this time, and a single complicated stamped sherd of the Santa Rosa-Swift Creek Period was also found at the Perico Island site. Going in the other direction we have, apparently, the strongest influences out of the Glades and into the Gulf Coast at any WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—WILLEY 555 period in the sequence. Shell hammers or picks, shell celts, and other shell artifacts, typical of the Glades area, are found in Santa Rosa- Swift Creek burial mounds well up the Gulf Coast. Boundaries be- tween Glades and northern cultures still are not well defined, but there is evidence for assuming that the Santa Rosa-Swift Creek culture was rapidly moving south and encroaching upon territory formerly domi- nated by the Perico Islanders. Following the Santa Rosa-Swift Creek Period some contact be- tween Glades and Gulf Coast was maintained during the Weeden Island and Englewood Periods. This is seen almost solely in the exchange of pottery between the southern Weeden Island sites and the Glades. Glades II and III Period pottery types are found in such Weeden Island sites as the Thomas mound (Hi-1) in southern Hillsborough County while occasional Weeden Island and Englewood sherds have shown up in Glades sites as distant as Belle Glade in Palm Beach County (Willey, n.d.). In the late periods Glades-Gulf Coast contacts again pick up. Not only is Glades pottery common in Safety Harbor Period middens and burial mounds, and Safety Harbor ceramic types frequently found in Glades sites, but Busycon and Strombus shell implements are also present in Safety Harbor sites. It is probably at about this time that many of the big flat-topped shell mounds and complex shell embankment sites were constructed along the southern shores of Tampa Bay and along the Glades area coast. It is likely that the idea of big flat-topped mounds of the temple type was a diffusion southward from Fort Walton via Safety Harbor. The big shellworks around Tampa Bay may have been built by Safety Harbor people, or they may represent a late movement of Glades tribes northward which brought a reintroduction of the temple-mound build- ing complex modified by the. addition of the complex embankment features that are so typical of the Glades area. The east—The culture areas or regions east of the Gulf Coast have been defined by Goggin (1947b) as: Central Florida, which lies immediately to the east of the central Gulf Coast; the Kissimmee region, lying immediately east of the southern portion of the central Gulf Coast and the Manatee region; the Northern St. Johns, along the Atlantic and the St. Johns River ; and the Melbourne region, which lies between the Northern St. Johns and the Glades area. Of the four, the Northern St. Johns region is the most distinctive and the most divergent from the Gulf Coast. Central Florida is, essentially, a geographical transition between the St. Johns and Gulf Coast. The Melbourne region is very close to the Northern St. Johns, and serves 550 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I1I3 as a traditional zone between it and the Glades area. The Kissimmee is a small region most closely related to the Glades area. As with the Glades area, the earliest connections between the Gulf Coast and east Florida are obscure. It is quite likely that preceramic and Orange Period relationships existed, with a very similar culture in both areas. The first definite tie-ups come on a Santa Rosa-Swift Creek level. In Central Florida, our knowledge of which comes mostly from Alachua County, Goggin (n.d. 2) has defined a Pre-Cades Pond Period (see fig. 76) which equates with Santa Rosa-Swift Creek. Pottery of this period shows relationship to Franklin Plain and the Early Variety of Swift Creek.*° The burial mound with primary burials is also a feature of the Pre-Cades Pond Period. In general, the drift of influence was probably from west to east at this time. Later, during the Cades Pond Period, con- nections with the Gulf Coast Weeden Island I culture are reflected in such Weeden Island ceramic traits as compartment vessels, Weeden Island Incised and Plain types, and Swift Creek Complicated Stamped, Late Variety. These wares may be actual trade pieces from the west. St. Johns I types are also present in Cades Pond sites, and the distinctive type, Oklawaha Plain (pl. 41, a), character- istic of Central Florida and the St. Johns area, shows close linkages with Weeden Island I Period ceramics. The burial mounds of the Cades Pond Period contain secondary burials.°° The Central Florida Hickory Pond Period correlates with Weeden Island II of the Gulf Coast. This is a burial-mound period in which Weeden Island trade wares are present. Ceramic influence was, apparently, passing both ways at this time. St. Johns Check Stamped, a type developed on the St. Johns River, is found in both Hickory Pond and Weeden Island II sites, while local Central Florida types such as Gainesville Linear Punctated and Prairie Cord-marked turn up in Weeden Island II sites on the Gulf Coast as trade wares. Comparable cross ties have been established for the later Alachua and Fort Walton Periods. 89 A mound excavated by James Bell (1883, pp. 635-637) in the southern part of Alachua County yielded pottery showing these Santa Rosa-Swift Creek affinities (U.S.N.M. No. 45802). 90 Cades Pond Period sites were excavated by Bell (1883, pp. 635-637) near Gainesville, Alachua County. Mound 3 of this group was a 12-foot-high burial mound containing 1,000 or more secondary burials along with Weeden Island I and Cades Ponds sherds (U.S.N.M. No. 43176). Bell’s mound 5, of the same group, also fits the Cades Pond-Weeden Island I time horizon (U.S.N.M. Nos. 43176-43177-43179). The Snowden mound, excavated by J. P. Rogan, also dates from this same period, judging by the collection (U.S.N.M. No. 88079). The Snowden mound was about 3 miles southeast of Gainesville. WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—WILLEY 557 South of Alachua and Marion Counties there is a small section of interior Florida centering upon Lake County and including part of the adjoining counties of Sumter and Orange. In his first con- sideration of Florida culture areas and subareas, Goggin (1947b) included this Lake County section in the Central Florida area; later (Goggin, n.d. 2) he related it as a subdivision of the Gulf Coast area. I have not, in this report, considered it as a part of the Gulf Coast. Like north-central Florida, it is transitional between the Gulf Coast and the St. Johns areas. However it is considered from a regional-taxonomic point of view, the Lake County section has significance in showing the eastward extent of Gulf Coast influences. Data for the Lake County region come largely from Moore (1895a) and Featherstonhaugh (1899). Moore excavated several sites in the region, among them the Hopson, Tavares, Helena, and Old Oka- humpka mounds. These were all low sand and earth mounds contain- ing secondary burials. Polished-stone celts, stone projectiles, stone and shell pendants of various kinds, copper beads and decorated sheet- copper objects, sheet mica, and pottery were found in the mounds. The traits of “killing” vessels by perforation and of making vessels with holes for mortuary purposes were also noted. Pottery is not typical of either Weeden Island or Santa Rosa-Swift Creek. In the Hopson mound were a rather crude double bowl, on the order of a Weeden Island compartment tray, and a rectangular boxlike bowl. Both vessels were plain. More direct contact with the Gulf Coast is seen in a sherd from the Tavares site which is typical Weeden Island (or Papys Bayou) Incised (Moore, 1895a, pl. LXXXVI, fig. 4). Goggin (n.d.2) equates the Hopson mound with Santa Rosa- Swift Creek and the others with Weeden Island I. I agree that the mounds equate with the Gulf Coast burial-mound periods, either Weeden Island or Santa Rosa-Swift Creek, but hesitate to refine the dating on any of them beyond this identification. Featherstonhaugh (1899) excavated near Lake Apopka, Orange County, at a group known as the Brooker mounds. He describes the site as four mounds which were grouped within a few hundred feet of each other. One was a burial mound; the others he believed to be house platforms. He found several hundred secondary burials in the burial mound along with polished celts, a duck-head stone pendant, shell plummet pendants, shell beads, shell pins, thin plates of copper, and great masses of broken pottery. Featherstonhaugh’s collections from the site, now in the United States National Museum, contain, among other items, a single-cymbal copper ear spool (No. 173819) and a collection of sherds (Nos. 173821-173824). The latter includes 558 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I13 Biscayne and Papys Bayou Series types. Biscayne and Wakulla Check Stamped date the mound on the general Weeden Island II horizon, but it is possible that the mound has been used in a later period as well, as Featherstonhaugh describes Venetian glass beads and iron tools. As stated, the preceramic and early ceramic levels of the St. Johns region have no well-defined counterparts on the Gulf Coast. It is possible, though, that certain pottery traits of the Orange, or fiber- tempered ceramic, Period do carry over into the later Gulf Coast periods. One indication of this is the linear-punctation technique that is so common in Weeden Island ceramics. This technique is not present in Deptford or Santa Rosa-Swift Creek but is a feature of the Orange Period. The transference of the linear-punctation feature from the Orange Period on the St. Johns to Weeden Island on the Gulf Coast is a possibility to be considered ; however, intermediate stages of this transference and development have not yet been revealed. It is with the subsequent St. Johns I Period that the specific con- nections between Gulf Coast and St. Johns areas appear in great numbers. This linkage is seen in burial mounds, secondary burials, stone celts, the use of copper, mica, galena, Strombus celts, Busycon cups, common mortuary deposits of pottery, the “killing” of vessels, and a variety of pottery traits. During the earlier part of the St. Johns I Period, the phase contemporaneous with Deptford and Santa Rosa-Swift Creek, Deptford pottery appears as trade in the St. Johns region along with the local St. Johns and Oklawaha types. Tetrapodal supports on pottery also come into the St. Johns at this same time. One of the distinctive features of the local St. Johns ware of the early St. Johns I is the boat-shaped or boxlike bowl form, and the probabilities favor the development of this trait in the St. Johns region. For it is here that the boat-shaped vessel form, an uncommon one in other parts of the Southeast, enjoys its greatest vogue. The boat-shaped bowl probably passed from east to west at this early period as it is known in at least one Deptford site in northwest Florida. Early Swift Creek Complicated Stamped pottery occurs as a minority type in the early phase of the St. Johns I Period, but, according to Goggin (n.d.2), it comes in slightly later than the Deptford types. Contemporaneous with the Swift Creek types we have the greatest clustering of Hopewellian traits of any period in the Northern St. Johns sequence. This is, of course, paralleled by the strong Hope- wellian cast of the Santa Rosa-Swift Creek Period in west Florida. These influences are seen, particularly, in the relative abundance of ornamental metalwork in these periods. These Hopewellian influences WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—-WILLEY 559 probably passed to the St. Johns by way of the Gulf Coast Santa Rosa-Swift Creek culture, or they may possibly have come from the north. The isolated geographical position of the St. Johns region favors a western entry through the Gulf area. In the latter phase of the St. Johns I Period the Gulf Coast influ- ences are out of Weeden Island, and there are numerous examples of western trade wares in the east. It is at this time in the west (Weeden Island II) that the eastern type, Oklawaha Plain, is both accepted as a trade ware and imitated. The later St. Johns II Period is characterized, as is Weeden Island II, by the rise of the small check stamped pottery. On the Gulf this is Wakulla Check Stamped, and in east Florida it is the St. Johns or Biscayne Check Stamped. These widely popular types probably had their origins on the St. Johns River and spread from there to the Gulf and even farther west. Toward the end of the St. Johns II Period temple mounds with ramps and late ceramics of the Fort Walton and Englewood styles are seen on the northern St. Johns. The St. Johns sequence closes with a Spanish Mission-Indian culture, the St. Augustine Period, which is comparable and related to the Leon-Jefferson Period in northwest Florida. South of the Northern St. Johns is the Melbourne region which is largely a pale reflection of happenings to the north. There are, appar- ently, no burial mounds in the Malabar I Period, which corresponds with St. Johns I. Burial mounds appear in Malabar II along with the small check stamped pottery and the stone celt (Goggin, n.d.2). This region, because of its geographical position and general back- wardness had neither much effect upon, nor was it closely bound up with, the course of developments on the Gulf Coast. The north.—We have seen how, following the preceramic and fiber- tempered pottery periods, the Gulf Coast and east Florida were similarly influenced. In some cases this was a matter of exchange be- tween the Gulf area on the one hand and Central Florida and the St. Johns regions on the other. But it is also obvious certain other elements were probably derived from a third source. To attempt to explain some of these diffusions we look northward to the Georgia area. During the preceramic and fiber-tempered periods the Georgia coast and the lower Savannah River displayed many similarities to the comparable Mount Taylor (preceramic) and Orange (fiber- tempered ceramic) Periods of the St. Johns (see fig. 76). In the succeeding Deptford Period,** coastal Georgia seems to have been a 91 A. J. Waring, Jr., has recently isolated a new chronological period between the Georgia Stallings Island and Deptford Periods. Pottery of this new period 560 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 113 center for certain ceramic ideas. These are expressed in the Deptford Complex which is best known from the Georgia coast and South Carolina. Deptford is, in effect, a South Atlantic complex, and every- thing points to its having diffused into Florida from the north and northeast. The geography of the situation suggests that Deptford influence probably reached the Northern St. Johns prior to its arrival on the northwest Gulf Coast. In view of this it is curious that Deptford has not yet been demonstrated to be full period in east Florida as it is in the west. The feature of tetrapodal supports is associated with the Deptford Complex in northwest Florida and appears on this time horizon in east Florida. On the Georgia coast, the type Deptford Simple Stamped shows the podal appendages, although the Deptford check stamped types do not (Caldwell and Waring, 1939). It is possible although disputable that the tetrapodal-support idea made its way into Florida from the north, with the Deptford Complex. The origin and distribution of this ceramic feature in the southeastern and east- ern United States is still a puzzle. It will be discussed further in the succeeding section on “The West and the North.” The problem of the first appearance of the burial mound in the Southeast is unsettled. This question, as it affects Gulf Florida, is in- volved with the origins of Swift Creek and related types of compli- cated stamped pottery. Swift Creek Complicated Stamped of the Early Variety is one of the diagnostics of the Santa Rosa-Swift Creek Period in Gulf Florida, and is also found in the early phase of the St. Johns I Period farther east. The area of intensity for the Swift Creek is south and central Georgia (fig. 73, e-7), overlapping into adjacent northwest Florida. In both the St. Johns and Gulf Coast areas com- plicated stamped pottery does not appear until after the advent of the first burial mounds. On the other hand, in central Georgia its earliest occurrences are thought to be unaccompanied by the burial- mound trait. In consideration of this I am inclined to believe that the burial-mound idea did not come into Florida from the immediate north or by way of Georgia; but complicated stamped pottery, which seems to have had an early history separate from the burial mound, appar- ently did spread from a hearth in Georgia southward into Florida. After the Santa Rosa-Swift Creek Period, relationships between Gulf Florida and Georgia were maintained in the realm of trade in raw materials. Mica and stone for celts were two of the main im- is sand-tempered ware, and, seemingly, represents an intermediate stage between the decorated fiber-tempered pottery and the Deptford types. (Waring, personal communication, 1947-48.) Fic. 73.—Related pottery types from Louisiana and Georgia. a, b, Marksville Stamped; c, d, Marksville Incised; e, 7, Swift Creek Complicated Stamped. (Redrawn from the following sources; a, b, c, after Setzler, 1933, pls. 1, 4, 3; d, after Ford and Willey, 1940, fig. 36; e-j, after Keily, 1938, pl. 11.) 561 562 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I13 ports from the uplands, while Gulf shells were sent north in return. Weeden Island Period sites were established well up the major river systems into Georgia as the northernmost outposts of the culture. Such sites have been found in the Okefenokee Swamp in south Georgia,®? and the big Kolomoki site in southwest Georgia was in- habited at least for a time by Weeden Islanders (personal communi- cation, A. R. Kelly and W. Sears, July 1948). Farther to the north Weeden Island influences in vessel and rim forms are seen in late Swift Creek levels near Macon. During the Fort Walton Period there are a number of similarities between the Gulf Coast and the contemporaneous Lamar Period of Georgia, but it is most likely that this is the result of both areas re- sponding to a third major influence rather than the direct effects of one upon the other. Very late, in the Leon-Jefferson Period, ceramic influence attributable to Georgia is seen in northwest Florida. The appearance of late complicated stamped wares in Florida at this time is a part of such influence. There is historic corroboration of this in that Creek tribes are known to have moved into north Florida at the close of the seventeenth century. The west and the north—It was from the west and the northwest that the major cultural influences were exerted upon Gulf Florida. It has been seen how south and east Florida, especially in the later periods, were on the receiving rather than the giving end of the main currents of diffusion as these passed between the Gulf Coast and the remainder of Florida. To the immediate north, the cultures of Georgia were largely influential in forming the ceramic patterns of the Gulf Coast during the earlier periods after which their influence lessened. It was from the west along the Gulf and the Gulf Coastal Plain that an important series of new ideas and, possibly, peoples came that were to link the Florida Gulf Coast with the vigorous and rapidly changing culture centers of the Mississippi Valley. It is likely that these western influences began as early as the Deptford Period, and they were probably derived out of the Tchefuncte culture of Louisiana (see fig. 76) or from similar and equally early manifesta- tions in Alabama (Wimberly and Tourtelot, 1941) and Mississippi (Jennings, 1941). The Tchefuncte culture is characterized by incised, punctated, linear-punctated, and rocker-stamped pottery, by small, collared jars with tetrapodal supports, the coiling principle in pottery making, the use of smoking pipes, and the burial mound with both 92 Collections in the Peabody Museum, Yale University, from Bug-a-boo Island (Nos. 21579-21580) and Buzzard mound (No. 21575), both in the Okefenokee, are clearly Weeden Island. WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—WILLEY 563 primary and secondary burials. Pottery types of the north Alabama Alexander Series are found in association. (See Ford and Quimby, 1945.) It is from Tchefuncte that I am disposed to derive the tetra- podal vessel supports for Deptford and subsequent periods of the lower eastern Southeast. The podal support trait, as it appears in Deptford, is simple and rare. In the coeval Tchefuncte culture it is elaborated and common (see Ford and Quimby, 1945, fig. 18). In Tchefuncte supporting feet occur on vessel bases and with vessel shapes that are in a different tradition than either the subconoidal pots of Deptford and similar Woodland styles or the deep open bowls of the Archaic fiber-tempered pottery. The burial-mound idea may also have been transferred from west to east on the Tchefuncte-Deptford level. If so, it did not become popular on the Florida Gulf for some time as no Deptford Period burial mounds are recorded either there or in Georgia. It is reasonable to believe, however, that Deptford burial mounds will be found and, as such, should represent the in- ception of the mound idea in Georgia and Florida. Along with the burial mound, or at about the same time, the customs of secondary burial and ceremonial breakage of grave pottery probably were dif- fused into Gulf Florida from the west. It must be emphasized that the question of the burial-mound com- plex in Gulf Florida is but one aspect of the larger historical problem of the origin and diffusion of mound burial in the eastern United States and that the interpretations given here are admittedly tenta- tive. More finely calibrated sequence dating, either relative or ab- solute, must be developed in eastern archeology before such a complex problem and all its ramifications can be satisfactorily resolved. This applies to the widespread net of relationships linking Tchefuncte, Adena, and other burial-mound cultures as well as to the nature of the specific connections that existed between Tchefuncte and the cultures of Gulf Florida and south Georgia. With the beginnings of the Santa Rosa-Swift Creek Period the affiliations of the Florida Gulf and Lower Mississippi Valley become very close. Tribal migrations from west to east may have taken place at this time. If the burial-mound complex was established in west Florida toward the close of the Deptford Period, and more or less simultaneously with the rise of the Swift Creek pottery styles, Marksvillian ceramic influences from the west were then intrusive into this early Swift Creek burial-mound culture. Or if the Marksvillian ceramic traits came into Gulf Florida as a part of the burial-mound complex, Swift Creek styles were added to this complex after its arrival and establishment. Data on this point are 564 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I13 not sufficient te enable us to choose between the alternatives. The Marksvillian and Troyvillian (fig. 74) ceramic influences in Florida are very specific (fig. 73, a-d), although there are differences between the Santa Rosa Series and the Louisiana pottery types. The Santa Rosa pottery was definitely Florida-made, not imported ware. The cambered rim decorated with incised cross-hachure, which is typically Marksvillian and Hopewellian, is not, for example, found in Santa Rosa-Swift Creek; and there are other differences which suggest a parallel development in the two regions rather than a complete dependence of one upon the other. Some of these differences may hark back to earlier influences into Florida out of Tchefuncte. For instance, the type Santa Rosa Stamped is much closer in decorative features to Tchefuncte Stamped than it is to Marksville Stamped, and Crystal River Incised has a much greater resemblance to the Tchefuncte type, Orleans Punctated (Ford and Quimby, 1945, pp. 62-63, pl. 6) than it has to any Marksville type. Monitor pipes of the Santa Rosa-Swift Creek Period can be related to similar pipes in the Marksville Period, and copper ear spools of the cymbal and bicymbal types are also both Marksvillian and Santa Rosa-Swift Creek. Other Gulf Florida copper forms, such as the conjoined tubes and the repoussé placques or ornaments are more Ohio Hopewellian than Marksvillian. At least copper has never been found in such abundance in the reported Marksville mounds. This may indicate channels of Hopewellian influence into Florida other than the Marksville. Possibly such copper artifacts were traded directly south out of the Ohio Valley through Alabama and Georgia rather than down the Mississippi River and then along the Gulf Coast to Florida. One fascinating element in Santa Rosa-Swift Creek which has no parallel in Marksville, or in any Hopewellian culture, as far as we know, is the negative-painting or resist-dye process of decorating pottery. The Gulf Coast occurrences of negative-painted pottery are the earliest in the east, antedating by what must be several centuries the use of the negative technique in Middle Mississippian pottery design in the Cumberland area and other parts of the Southeast. The most reasonable explanation is that the resist-dye process was trans- ferred from wood, textiles, or gourds to pottery in Gulf Florida. Knowledge of the technique in connection with perishable materials may have come into the Southeast from east Mexico. The Vera Cruz- Tamaulipas area is the nearest locality in which negative-painted pottery is found. Fic. 74—Related pottery types from Louisiana. a, b, Troyville Stamped; c, d, Churupa Punctated; e, f, Yokena Incised. (After Ford and Willey, 1939.) 565 566 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. II3 In this discussion we have refrained from placing a geographical boundary between the Santa Rosa-Swift Creek and Marksville cul- tures, but it is almost certain that the two merge somewhere in the 150 miles that separate Pensacola Bay from the delta of the Mississippi. This intervening coastal strip should be investigated with the problem of such a contact in mind. In the succeeding Weeden Island Periods the Santa Rosa-Swift Creek settlement pattern of the small villages and the associated burial mounds is retained; but, contemporaneously, a similar com- munity type was undergoing profound changes in the Troyville and Coles Creek cultures of Louisiana. Large temple-mound sites were founded in the Troyville Period and became widespread during Coles Creek, while the burial mound either completely disappeared or is very rare for these periods. The extent to which the temple-mound complex became established in Gulf Florida during the Weeden Island Periods has not been fully determined, but it is certain that it was not an important feature. Possibly the idea was just beginning to be accepted as there are some Gulf Coast mounds of modest size which may be Weeden Island temple mounds, and in southwest Georgia the big Kolomoki temple mound may be a Weeden Island structure. The strong ceramic affinities of the Gulf Coast and the Lower Mississippi Valley during the Weeden Island and Troyville-Coles Creek Periods are best seen in the types Weeden Island Incised and Weeden Island Punctated and the Louisiana type, French Fork In- cised. In the Louisiana sequence the French Fork type has its in- ception in the Troyville Period, reaches its maximum occurrence just at the close of this period, and decreases and finally disappears during the Coles Creek Period. Other types further strengthen this Florida-Louisiana relationship. Mazique Incised and Rhinehardt Punctated in Louisiana have parallels in Carrabelle Incised and Carrabelle Punctated, and Weeden Island Zoned Red has a striking likeness to Woodville Red Filmed. Additional linkages can be made between Weeden Island Plain and Coles Creek Plain, Keith Incised and Beldeau Incised, and Wakulla Check Stamped and Pontchartrain Check Stamped. (See fig. 75.) These western ceramic affiliations of Weeden Island exist not only between Florida and the Lower Mississippi Valley but extend up the drainage of the Red River in Louisiana into southwestern Arkansas. In the latter area the Weeden Island-like pottery, as found at the Crenshaw site (Dickinson, 1936), underlies late Caddoan pottery. Krieger (1946, fig. 26) has dated it as of about A.D. 1200-1300 (see Crenshaw, fig. 76). on i Ae ne A ee er GF a) Fic. 75.—Related pottery types from Louisiana. a, c, Coles Creek Plain; a French Fork Incised; d, Mazique Incised; e, Woodville Red Filmed; f, Rhin hardt Punctated; g, Coles Creek Incised ; ‘h, Beldeau Incised. (After Ford ahd Willey, 1939.) 567 568 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I13 That separate parallel developments of the Weeden Island or French Fork style took place in Gulf Florida, southern Louisiana, and southwestern Arkansas is not a plausible assumption. Although there are regional differences, there is too much of a unity in this Weeden Island-French Fork-pre-Caddo style group to account for it by parallel developments from a similar ancestry. Of the three areas where the Weeden Island-like wares occur, the greatest diversity and elaboration of what might be called the central stylistic themes are found in Gulf Florida. This suggests a Floridian origin and development with a subsequent westward diffusion. There are pos- sible types for the Weeden Island or French Fork style in Marksville but those of Santa Rosa-Swift Creek are more convincing. For ex- ample, the type Crystal River Incised is a more logical ancestor to Weeden Island Incised than any of the Louisiana Marksville types. If we are correct, and the principal flow of ceramic ideas during the Troyville and early Coles Creek Periods was from east to west, then this stream of influence and its direction certainly continued in late Coles Creek and Weeden Island IT times. Small check stamped pottery appears in southern Louisiana at this period and seems to come from the east where similar types were extremely popular dur- ing the Weeden Island II. As opposed to this the marker type for Coles Creek, Coles Creek Incised, had only a slight effect upon Florida ceramic traditions. This is seen in St. Petersburg Incised, a rather rare type in late Weeden Island II. The distribution of Weeden Island culture actually extends west of the boundaries of the Gulf Coast area as we have defined them. Continued archeological explorations in south Georgia and south Alabama are almost certain to reveal more sites of the type. On the lower Tombigbee River in Alabama, 50 to 75 miles from the Gulf Coast, Moore (1905, pp. 253-262) excavated three Weeden Island burial mounds—Payne’s Woodyard, Carney’s Bluff, and Kimbell’s Field. Along the Alabama coast, De Jarnette and Bucker (ms. of 1937) explored a mound in lower Baldwin County that is undoubtedly Weeden Island. The westward continuation of Weeden Island and the eastward extension of the Troyville-Coles Creek cultures are still to be plotted, but the presence of Weeden Island as a burial-mound and ceramic complex in much of south Alabama is assured. The possibility of a much wider range for Weeden Island ceramic ideas than we have outlined above should be kept in mind. Weeden Island is, essentially, a pre-Middle Mississippian culture. While there is undoubtedly some contemporaneity between Weeden Island II and the earlier phases of Middle Mississippi (Temple Mound I WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—WILLEY 569 horizon) the artistic patterns of Weeden Island were formulated be- fore this. They were, in fact, in formulation during the Santa Rosa- Swift Creek Period. One fact should be emphasized. During the Santa Rosa-Swift Creek and Weeden Island I Periods no other area in the eastern United States produced such an elaboration of pottery vessel forms, decorative techniques and designs, and effigy modeling as did the Florida Gulf Coast. Only the Middle Mississippian and Caddoan potteries rival the Weeden Island in elaborateness and skill, and these styles are clearly later in time. In Santa Rosa-Swift Creek and Weeden Island I we note the following techniques: incision, punctation, linear punctation, rocker stamping, direct painting, nega- tive painting, effigy modeling, and complicated stamping. Several of these technical traits are found in late Middle Mississippian pottery ; and two of them are particularly characteristic of much of it. These last are effigy adorno modeling and negative painting. It is suggested here that both of these techniques may have been derived from a culture with Weeden Island-type ceramics that preceded and in- fluenced later Middle Mississippian wares in the Central Mississippi or Lower Ohio Valleys. Such a transition could have been taking place during the early Middle Mississippian periods in western Ar- kansas and eastern Tennessee and Mississippi or even farther north. A hint of contact is seen at Cahokia, IIl., where occasional sherds which are very much like Weeden Island Incised have been found in association with the earlier or Old Village occupation of that site (Willey, 1945, p. 243). All this is, of course, a hypothesis for which most of the evidence is lacking, but the striking similarity between Middle Mississippian bird and human-head rim effigies and those of Weeden Island prompts it when one realizes that the Weeden Island tradition for this sort of thing is considerably the older.** Influence from the west and the north is plainly dominant in the Fort Walton Period of Gulf Coast prehistory, and these influences, though fainter, are registered in the south in the contemporaneous Safety Harbor Period. Although the idea for the temple mound may have reached west Florida in Weeden Island times, it did not become firmly established until Fort Walton. It is likely that during this later period the temple mound and the ceremonial complexes which surrounded it were introduced under different circumstances than previously. This was probably as part of an actual invasion of 93Tn this connection, another link between a Gulf Coast burial-mound cul- ture and the late Middle Mississippian horizon is the suggestive prototypes for the Southern Cult in the ceramic designs found at Crystal River, a Santa Rosa-Swift Creek Period site (see Willey, 1948c). 38 570 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 113 the northwest Gulf region by a people whose culture was predomi- nantly Middle Mississippian. Something of the possibilities of con- tinuity between Weeden Island and Fort Walton ceramics has already been discussed. Design lay-outs and rim effigy forms may be local Florida transfers from Weeden Island to Fort Walton. Or the Fort Walton rim effigy could be a reintroducton of an old Weeden Island idea that had been integrated into and modified by a Middle Mississippian culture. Whichever was the case we can be fairly sure that some other new ceramic modes were being brought into west Florida during the Fort Walton Period. These new features of vessel form, vessel decoration, and of temper can be traced to south, central, and even northern Alabama. Large bowls used for burials covers, stylized death’s-head (?) designs, shell temper, frog-effigy bowls, the bottle form, and black engraved ware are some of them. It is in the Fort Walton Period that a very few evidences of the Southern Cult (Waring and Holder, 1945; Krieger, 1945) appear in west Florida. These are the engraved eagle and eagle-mask designs on a polished black pottery specimen from the Jolly Bay (W1-15) site in Walton County. Occasional “Cult” pieces are found in the late sites of the St. Johns farther east, but “Cult” influence does not seem to have been strong in Gulf Florida. The West Indies—The problem of Florida-Antillean connections has been considered for many years. It is, I believe, a fair statement to say that the case for cultural influence in either direction has never been satisfactorily demonstrated. This does not deny the possibilities of such contact. It is, indeed, curious that more evidences of trade or intercommunication have not been found.** Fontaneda, the ship- wrecked Spanish boy who spent several years with the Calusa Indians in the sixteenth century, reported that there was a colony of Arawaks living in southwest Florida for a time under the sovereignty of the Calusa leader Calos (Escalente Fontaneda, 1944). In spite of what seem to be the potentialities of the situation there is little in the archeology of either Florida or the Antilles that implies cultural affiliations of the same order which we have been discussing for Gulf Florida and other areas of the southeastern United States. Osgood has approached the problem of Florida-West Indian con- nections on the time level of the Southeastern Archaic culture. He has taken the hypothesis (Osgood, 1942, p. 57), also advanced by others, that the first inhabitants of Cuba were a former North ®4 The only bona fide West Indian trade item found in Florida is a stone ax from a site in Alachua County (Goggin and Rouse, 1948). This ax is of the “eared” variety common to the Lesser Antilles. WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—WILLEY 571 American preagricultural, preceramic people who were gradually forced south in the Florida peninsula and eventually crossed over to Cuba where they ultimately evolved the Ciboney type culture. He dates this movement at about A.D. 500 which is only slightly earlier than the estimated beginnings of the agricultural and ceramic pe- riods in Florida whose populations are presumed to have been respon- sible for displacing the Floridian Archaic tribes. The Archaic Southeastern emigrants apparently had the Greater Antilles to them- selves until they were compressed into western Cuba by the expan- sion of the agricultural, pottery-making Arawaks who invaded the islands from South America. This hypothesis is plausible and should be investigated further. The main difficulty has been in the tracing out of relationships be- tween western Cuba and south Florida by comparisons of the shell tools found in the two areas. These are similar, but the environments are similar ; and the shell implements in question are also so extremely simple that they fail to clinch the argument for a diffusion (Osgood, 1942, pp. 40-43). Even if true, the greater part of the story of Gulf Coast development, as presented here, is not affected by this theory. Only the postulated preceramic period would be involved. Rouse has viewed the problem also from the point of view of Florida-to-the-Antilles diffusions (Rouse, 1940). He was primarily concerned with the origins of the Meillac pottery, a ceramic tradition that is found in the northern Greater Antilles and which is the earliest pottery in those islands where it follows the preceramic horizon. It was his thesis that the relatively simple incised and modeled Meillac ware had been influenced by North American decorative techniques. He drew his Southeastern comparisons not from Florida, alone, but from several areas as well as from different time levels. Rouse’s methodology was not grounded on comparisons of types or styles but with the cross-matching of what he termed “modes.” These modes are isolated elements such as “‘punctation,” “curved incised lines,” or “naturalistic design.’ By treating the data in this rather atomistic fashion, he was able to demonstrate that Meillac pottery had a greater over-all similarity to pottery of the southeastern United States than it did to wares in northeastern South America. This was the crux of his argument, and although it does have a bearing upon Southeastern- West Indian relationships, it does not aid us greatly in working out the time and place of specified diffusions. The frame of reference is too general to make his conclusions more than suggestive. The most intriguing Florida-Antillean similarities are, I think, between Weeden Island pottery and the Carrier ceramic style of the 572 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I13 Indies. The Carrier style is the late, most widely distributed, best- known pottery type of the Greater Antilles. It has in the past usually been referred to as “Arawak pottery.” °° In mode-for-mode or element-for-element comparisons, using the Carrier modes described by Rouse (1940, pp. 59-60), an imposing case can be made for the relationship of Weeden Island and Carrier. Out of the 23 Carrier modes 21 are found in Weeden Island. While there is an annoying, elusive similarity between the two styles, this resemblance is nowhere near as strong as the mode analysis might lead one to believe. A close inspection of the modes or elements held in common weakens, rather than strengthens, the arguments for relationship between the two styles. For example, the modes of “naturalistic design” and “zoomorphic head lugs” are quite distinct as they are found in each style: Weeden Island emphasizes the bird motif ; Carrier, the monkey or the bat. “Curved incised lines,” “curvilinear incised design,” “punctation,’ and “modeling” are designations so general that they have little meaning except for very broad contrasts or comparisons. At the same time, a few of the modes stand up fairly well under this scrutiny. “Line and dot incision,” or terminal punctations in or at the ends of incised lines, in Carrier are deep, hemiconical gouges ter- minating a broad, round-bottomed incision or groove. This incision and punctation is a little bolder than seen in most Weeden Island In- cised, but it compares very favorably with some of the Weeden Island Incised pottery of Weeden Island I or some of the bold line incision techniques of the preceding Santa Rosa-Swift Creek Period. Another mode, “inturned shoulders,” producing the flattened-globular bowl form, is a fairly common New World pottery shape, but the fact that the Florida Gulf Coast seems to be an early Southeastern center for the trait cannot be ignored. In the same way, “ridge on outside rim,” or folded rim, is a simple idea, but, like the flattened-globular bowl, its Southeastern center is the Florida Gulf. The “boat-shaped bowl” is another mode of this same kind. Florida is the center for the form. It is more common on the St. Johns than the Gulf Coast but does occur in the latter area in Weeden Island times and earlier. We have then, a number of features in Carrier ceramics which are also typical or present in Weeden Island. Some of these, like the flattened-globular bowl, the folded rim, and the boat-shaped bowl are ®5 Rouse has established regional styles for the various geographical units of the West Indies. The late period styles or types in Puerto Rico (Esperanza and Capa), Dominican Republic (Boca Chica), and eastern Cuba (Pueblo Viejo) are all similar to the Carrier from Haiti and the Bahamas (Rouse, 1948). WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—-WILLEY 573 very characteristic of the Florida or Florida Gulf Coast area; spread to other parts of the Southeast from there; and may have been intro- duced from the West Indies. These similarities and the distributional facts that accompany them are not sufficient to prove diffusion be- tween Florida and the Antilles, but they are provocative enough to keep the door open to this possibility. Time equations between the two areas would not preclude such diffusions. The Carrier and related styles were flourishing in the West Indies at the time of Columbus or up until about A.D. 1500. The date line 1500 is also the estimated terminal date of the Weeden Island culture in Florida. Beginning dates for both Weeden Island and Carrier can only be estimated, and with less accuracy than the terminal dates. As such they do not serve as reliable data points for plotting the course of area-to-area connections. The best we can say is that if diffusion did take place it could have proceeded from north to south or vice versa. Such a con- tact would have bypassed both western Cuba and southern Florida which lack, respectively, the Carrier and Weeden Island types. There is one other possibility of ceramic influence between the Antilles and the Southeast. Holmes suggested many years ago (Holmes, 1894b) that complicated stamped pottery in the southeastern States was the result of the introduction of West Indian art forms in wood carving onto the North American mainland. He had come upon this idea while examining a unique carved wooden stool or duho from Turk’s Island in the Bahamas. In his opinion the carved designs on the stool were strikingly like those found on Southeastern compli- cated stamped pottery, and he inferred that Southeastern artisans had copied such designs from stools or other wooden objects onto the wooden paddles with which they stamped their pottery. The idea is partially sustained by recent findings in Southeastern archeology which show that the complicated stamped pottery styles appear quite sud- denly and full-blown. Prior to the appearance of the complicated styles, pottery was stamped, but the designs were simple lines or checks. This abrupt appearance of the new pottery designs does argue for a relatively sudden technological transference of the com- plicated carvings from other manufactures. Such a process could have occurred, however, entirely within Southeastern native traditions ; and proof of a West Indian diffusion reverts, in final analysis, to similarity or lack of similarity between West Indian and Southeastern designs. The examples of design which Holmes shows, from the wooden stool and from Southeastern stamped pottery, are certainly suggestive of contact in their general similarity. Holmes’ proposition, like some of the others, should be examined by more thorough and rigorous analysis. 574 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. II3 We have considered the question almost entirely upon the basis of pottery similarity or dissimilarity, but there is one aspect of late Antillean culture that definitely does not reinforce the arguments favoring a connection with Weeden Island. This is the burial-mound trait. Neither in Carrier nor in any other West Indian culture do we have definite evidence of burial mounds.°* This is a basic dis- tinction between West Indian cultures and not only Weeden Island but many contemporary horticultural societies in the southeastern United States. TIME-SPACE INTEGRATION A graphic expression of the temporal and spatial relationships of Florida Gulf Coast culture periods and neighboring archeological culture areas in the Southeast is the chronology chart presented in figure 76. The Gulf Florida sequences are given in the three regional divisions as they have been developed in this report: the northwest, the central, and the Manatee. Gulf Coast and Georgia and Gulf Coast and Louisiana correlations follow the survey conclusions of Willey and Woodbury (1942) and of Ford and Willey (1941). The position of the Leon-Jefferson Period is from H. G. Smith (1948). Correlations of Gulf Coast with the Glades and East Florida are taken from Goggin (1947b; n.d.2), and from numerous conversations with Goggin during 1944-48. The placing of the “Caddo” area column in juxtaposition to the other sequences is after Krieger (1946). The chronological charts of Martin, Quimby, and Collier (1947) and J. B. Griffin (1946) have also been consulted. The estimated absolute dates are taken directly from Krieger (1946) for the centuries from A.D. 1200 to 1700. (See also Krieger, 1947.) These are the best archeological dating estimates for the Southeast, cross-tying as they do with the tree-ring calendar in the eastern Pueblo area of New Mexico. The effect of Krieger’s dating has been to drop back the estimates for the later periods by 100 to 200 years.°* For example, in the Lower Mississippi Valley column the beginnings of the Placque- mine Period have been revised from a date of A.D. 1600 to approxi- mately 1450, and the Coles Creek Period has been lowered from an initial date of around A.D. 1350 to about 1250. Marksville and 96 Gower (1927, pp. 13 and 49) lists burial mounds as “Tainan” or Arawak. There is no clear-cut evidence anywhere in the West Indies for burial mounds as distinguished from rubbish hillocks in which burials were made. 97 As this goes to press Krieger’s recent (1949) dating estimates in eastern Texas place Early Alto (Gahagan) as coeval with Marksville (and Santa Rosa-Swift Creek) and place them all as of about A.D. 500. OASTAL IVANNAH INE, 4 VAMWVAH E ‘MINGTON PTF ORD in relatio 4 575 changes. Valton is ods low- ‘s. 2 and aceramic, cte, have and late It prob- had been 2riods as " ceramic ‘e wares ; diffusion 2s has, in tures in- has been Island I the two ind com- Weeden cultures, \poraries, violated a burial- horizon. the trait time lag the dif- he work- doraneity the same logy has ex which ibility to graphical concepts, esistance ‘d propa- ire traits theastern “ LOWER ANSS1ISSIPP/ FLORIDA GULF COAST VALLEY NATCHEZ, LEON- GLENDORA BAYOU GOULA, ETC. | YEFFERSON SAFETY | SAFETY HARBOUR | HARBOUR ALACHUA FT. WALTON LAMAR (QENE - FULTON PLA QUEMINE i ies LNGLEWOOD ST JOHNS | MALABAR GLADES IT I Z WEEDEN | WEEDEN | WEEDEN | AtACON | SAVANNAH | HICKORY PLATEAU Ts POND tig Z G/BSON COLES CREEK CRENSHAW =} GAHAGAN ALTO TROYVILLE MARK SVILLE TCHEF UNCTE "CADDO AREA EAST FLORIDA GLADES AREA EST DATES CENTRAL OCMULGEE FIELDS COASTAL N. CENTRAL) N. ST JOHNS MELBOURNE ST ST AUGUSTINE | AUGUSTINE SAVAMVAH | CADES Zi; POND WEEDEN | WEEDEN | WEEDEN /SLANO WILMINGTON GLADES IT SZ JOHNS | MALABAR Le PER/CO 1SLANO GLAOES LT DEPTFORD DEPTFORD S77 SIMON S oR ORANGE ORANGE (?) ORANGE PRE - CERANIIC CERAMC PRE- CERAMIC PRE - CERAANC Fic. 76.—Chart of comparative culture sequences showing the Florida Gulf Coast in relation to neighboring culture areas. WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—WILLEY 575 Tchefuncte have been made earlier in accordance with these changes. The same is true for the Florida Gulf Coast where Fort Walton is now given a beginning date of 1500, and the preceding periods low- ered to conform with this (cf. Ford and Willey, 1941, figs. 2 and 6). The earlier periods represented on the chart, the preceramic, Orange, Stallings Island, Deptford, and probably Tchefuncte, have all been foreshortened out of all proportion with the middle and late periods. This was done to make room for them on the chart. It prob- ably would have been closer to the truth if these horizons had been lengthened to three or four times the duration of such periods as Santa Rosa-Swift Creek or Weeden Island. All sequence equations have been made upon the basis of ceramic cross ties. In some instances this has been by means of trade wares ; in others, stylistic resemblances indicative of relatively rapid diffusion have served. The equation of culture periods by ceramic types has, in some cases, been at variance with other aspects of the cultures in- volved. For example, the Troyville Period of Louisiana has been matched on the chronology chart (fig. 76) with the Weeden Island I Period of the Gulf Coast, and the ceramic correlation of the two periods is, indeed, high. The distribution of the burial-mound com- plex, on the other hand, does not conform to this equation. Weeden Island I, and also Weeden Island II, are burial-mound cultures, whereas Troyville and Coles Creek, their Louisiana contemporaries, are not. Sequence alignments by ceramics are even further violated when we consider the Safety Harbor Period which is also a burial- mound culture, although it is on the European-contact horizon. Clearly, in this case, the factor of time lag is operative upon the trait of mound burial. It is also possible, and even likely, that time lag has also affected the distribution of pottery trade wares or the dif- fusion of pottery-making and pottery-decorating ideas. But the work- ing assumption underlying the attempt to arrive at contemporaneity between the sequence periods of two relative chronologies is the same as that upon which the structure of the Gulf Coast chronology has been erected. Pottery has been considered as a sensitive index which changes through time with relative rapidity. This susceptibility to rapid change should minimize the time-lag factor in its geographical distributions. Furthermore, the diffusion of new pottery concepts, from area to area, should have met with less ethnocentric resistance than would have been engendered by diffusions or attempted propa- gations of other ideas. These inequities in the time-space arrangement of culture traits and trait complexes have given rise to another concept in Southeastern 576 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I13 archeology, the culture stage. The formulation of stages must neces- sarily be preceded by the establishment of periods and horizons. The latter are horizontal time zones such as those indicated on the chro- nology chart (fig. 76). A culture stage may be, in certain circum- stances, coextensive with a culture horizon, but usually it is a di- agonal cross-sectioning of the time continuity. Ford and I applied a culture-stage type of synthesis to eastern United States archeology in 1941 (Ford and Willey, 1941). The stages which we employed were defined primarily by mound ceremonialism. Superficially, this was an inconsistency as “burial mounds” are essentially places for the disposal of the dead, while “temple mounds” are pediments for religious or political buildings. Two quite different cultural categories or activities were, thus, being contrasted in the same system. Yet, in this case, the labels “burial mound” and “temple mound” have a fundamental consistency. We were, in effect, contrasting community types by emphasizing their most conspicuous symbols. In this sense “burial mound,” or the Burial Mound stage, represented not just a’ burial practice but a way of life differing significantly from the “temple mound” or Temple Mound stage. A third community type, the Archaic, was also defined as a stage and is, in this same sense, in comparable contrast to the other stages; although the term “Ar- chaic” has a connotation quite different from that of the other two stage names. The three major culture stages of the East may be applied to Gulf Florida and its neighboring areas in a very striking manner. The Archaic stage includes: (1) the pre-Tchefuncte period in the west; (2) the preceramic, Orange, Deptford, and, probably, early Perico Island Periods on the Florida Gulf; (3) the preceramic, Stallings Island, Deptford, and, probably, early Swift Creek Periods in Georgia ; (4) the preceramic and Orange periods in east Florida; and (5) Glades I and probably part of the Glades II Period in the Glades area. The Burial Mound stage begins in the west with the Tchefuncte Period which belongs to the Burial Mound I subdivision of the stage. Possibly components of this stage will be revealed in the Deptford, early Swift Creek, and early St. Johns I cultures; but evidence for this is not clear. The Burial Mound II subdivision of the stage is more apparent. It includes: (1) the Marksville Period in the west ; (2) the Santa Rosa-Swift Creek, late Perico Island, Weeden Island I and II, Englewood, and Safety Harbor Periods on the Florida Gulf ; (3) the late Swift Creek, Wilmington, and Savannah I Periods in Georgia; and (4) the Cades Pond, Hickory Pond, St. Johns I, early St. Johns II, and Malabar II Periods in east Florida. Temple WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—WILLEY 577 Mound I stage is applicable to the Troyville-Coles Creek Periods in Louisiana, the Macon Plateau Period, and, perhaps, the Savannah II Period in Georgia. Only Temple Mound II, however, is seen in Florida where it is represented by the Fort Walton Period, probably the Alachua Period and the late St. Johns II Period. The Temple Mound II stage is also well represented in Georgia by the Lamar and Irene-Lamar Periods. In Gulf Florida, the Safety Harbor culture was undergoing a transitional phase between the Burial Mound and Temple Mound stages. South, in the Glades, burial mounds make their appearance in the late Glades II and Glades III Periods, and temple mounds also appear in the Glades III Period. But other major aspects of Glades culture, including the subsistence orientation, re- main more or less unchanged from Archaic stage standards. SUMMARY RECONSTRUCTION The prehistory of the Florida Gulf Coast can be most readily syn- thesized on three levels or stages of cultural attainment: the Archaic, the Burial Mound, and the Temple Mound. 1. During the Archaic stage Gulf Florida was first occupied by a hunting, fishing, and shellfish-gathering people who occupied villages along the shore. These Indians made flint and shell tools, weapons, and other implements but no pottery. They were undoubtedly related by culture, race, and, possibly, language to other Archaic-stage In- dians of the southeastern United States. At this period, which we will call simply the preceramic, the Gulf Coast was not as densely in- habited as other areas of the Southeast, such as the lower St. Johns River in northeast Florida, the lower Savannah River, and the Tennessee River in Alabama. Fresh-water shellfish seem to have been the principal diet at this early time, and, as the rivers flowing into Gulf Florida were not so large or so rich in mussels as those named, adequate food resources for larger populations were not available. Later, in the Archaic stage a distinctive fiber-tempered pottery was manufactured at many places throughout the Southeast. On the lower Savannah this fiber-tempered pottery period has been called the Stallings Island, on the Georgia Coast, the St. Simon’s, and on the St. Johns, the Orange. Occasional fragments of this fiber ware have been found in Gulf Florida sites, suggesting that such a period may have been in existence here, but data are as yet too few to be definitive. Toward the close of the Archaic stage the idea of ceramics becomes fully established in northwest Florida. This has been called the 578 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I13 Deptford Period after a culture of the same name on the Georgia Coast and lower Savannah River. Deptford pottery is sand-tempered, made in simple conoidal pot shapes, and decorated with check or simple linear-stamped impressions. It is found most abundantly in the eastern section of northwest Florida, diminishing toward the west. This distribution, coupled with its abundance in Georgia and South Carolina, suggests that the knowledge of this type of ware came to Florida from the northeast. Deptford sites are small shell- midden heaps, and it is doubtful if any important changes in the type of economy had occurred since preceramic times. Oysters and other marine shellfish apparently replaced fresh-water mussels as the dietary staple at about this time. This shift was brought about by environmental changes, and, undoubtedly, made the seacoast a more desirable place to live. Contemporaneous with the Deptford Period another culture was flourishing in the south, around Tampa Bay. This is the Perico Island. The Perico Islanders were also potters, and manufactured a sand- tempered ware, less well made than the Deptford and only occasion- ally decorated with incisions and punctations. The Perico Island people must have been a part of the earlier preceramic population who occupied the Florida peninsula. The stimulus for their ceramic craft was probably derived from the Orange Period people on the St. Johns and from the Deptford Indians. The specialization of Perico Island culture, as seen in its pottery and shell tools, probably marks the be- ginning of the Glades cultural tradition which was to continue for centuries in southern Florida. A little later the Perico Islanders were to adopt the burial-mound complex and to trade with more advanced ceramic craftsmen in north Florida, but this takes us ahead of our story into the next stage. It is possible that all these Archaic-stage Indians may have been closely affiliated as to racial group. This is a surmise based upon our knowledge of other Archaic peoples of the Southeast. These were medium- to long-headed Indians. If the old Florida Indians were medium- to long-headed we may expect to find this physical type in the south, an area which also retained many Archaic cultural traits long after they had disappeared in the northern part of the State. As this south Florida Glades area was in historic times occupied by the Calusa, Tekesta, and Ais, linguistic groups unaffiliated with the major stocks of the Southeast, it may be that the Archaic populations of Florida spoke these languages. 2. The first period of the Burial Mound stage in Gulf Florida is the Santa Rosa-Swift Creek. There are hints that certain ceramic traits WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—WILLEY 579 associated with this period, as well as the burial-mound idea, may have reached northwest Florida even earlier, during the Deptford Period ; but the evidence on this is inconclusive. The appearance of the Santa Rosa-Swift Creek culture type may be linked with the arrival of a new people speaking a language different from those of the old Archaic-stage inhabitants. If so, the probabilities favor this language, or linguistic stock, being the Timucuan. Santa Rosa-Swift Creek culture is found throughout most of the Gulf area, occurring as far south as Tampa Bay. During this expansion it probably im- pinged upon and influenced the Perico Island culture. Living sites of the Santa Rosa-Swift Creek Period are also small villages, but these are much more numerous than the Deptford or Perico Island Period sites. Santa Rosa-Swift Creek population has been estimated at about 5,000 to 7,000 persons. The duration of the period is placed at 200 years, from A.D. 800 to 1000. It is thought that the new elements responsible for the Santa Rosa- Swift Creek culture came into Florida from the west as the burial- mound and ceramic complexes of the Lower Mississippi Valley Marksville Period have much in common with Santa Rosa-Swift Creek. In turn, Santa Rosa-Swift Creek became the focus of Burial Mound stage influences for the rest of Florida and Georgia. Although local supporting evidence is lacking, maize horticulture was probably introduced into Gulf Florida as a part of the Santa Rosa-Swift Creek Complex, and from this time on farming was an important economic activity of the Gulf Florida Indians. Western influences were not the only ones that were determinative for Santa Rosa-Swift Creek. In addition to the new pottery styles of the west (incision, punctation, rocker-stamped, and painted tech- niques), a complicated stamped tradition has a major role in the ceramic complex of the period. This is expressed in the Swift Creek ware, and the conception and skillful execution of the decorative de- signs of the style imply a mastery of the wood-carver’s art. Origins of Swift Creek and other complicated stamped pottery styles were in south Georgia. Other streams of influence in Santa Rosa-Swift Creek are the negative-painted pottery type, which may have devel- oped locally as a technological transfer; the polished-stone celt and monitor pipes, which came from the west or the north; use of mica, a northern import ; and copper ornaments of the Ohio Hopewell type. These last were probably trade items. Shell tools also found their way into Santa Rosa-Swift Creek artifact assemblages from the Glades area to the south. 580 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I13 The Weeden Island I and II Periods were local developments out of the Santa Rosa-Swift Creek Period. The nature of sites and settlements and their distribution are much the same during these two periods as they were during the previous one, although population in- creased. An average of 10,000 to 15,000 people for the Gulf Coast is estimated, and the duration of the two periods is put at 250 years for each, spanning the gap from A.D. 1000 to 1500. Burial-mound practices were much the same during the Weeden Island Periods with only slight changes noticeable. Secondary burial is a little more com- mon than during the Santa Rosa-Swift Creek Period. Pottery styles change, but show relationships to the previous period. There is a tendency for the complicated stamped tradition and the vessel forms associated with this tradition to drop out, especially in the Weeden Island II Period. Stonework is a little scarcer than previously, pipes are less common, copper artifacts are very rare, and shell tools and ornaments are not as frequent as in Santa Rosa-Swift Creek. Weeden Island trade contacts with both the north and the south seem less strong than they had been; with the west, contact continues and is reflected mainly in ceramics. There is also a very real possibility that Weeden Island ceramic ideas, as represented in effigy modeling, had an influence upon the development of Middle Mississippian styles. This is suggested by stylistic similarity and the earlier position of Weeden Island in relation to Middle Mississippi; but the course of such con- tact has not been plotted. It is during the Weeden Island Periods that contacts with the West Indies are most likely to be found, if such existed during the sequence of pottery periods on the Gulf Coast. Evidence on this question is not sufficiently strong to postulate such diffusions, in either direction. In general, the Weeden Island Periods were the culture climax or optimum prehistoric periods for the Gulf Coast. The subsistence technology at that time represented the maximum effective adjustment to the environment. Food economy was undoubtedly balanced between fishing, hunting, and shellfish gathering on the one hand and horti- culture on the other. 3. The Temple Mound stage in Gulf Florida was chronologically late in its arrival compared to some other areas of the Southeast, and it was undoubtedly a part of the lower Southeastern spread of the Middle Mississippian intensive agriculturists. It is postulated that a new people came into northwest Florida at this time with the Fort Walton culture. These were Muskogean peoples who, moving south through Alabama and western Georgia, dislodged the Timucuans and pushed them east into the Florida peninsula. The Apalachee, Sawokli, WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—WILLEY 581 Chatot, and Pensacola were, following this argument, the bearers of the Fort Walton culture. At the same time, the Englewood and Safety Harbor cultures of the central Gulf Coast and Manatee regions were evolved locally out of Weeden Island and late Weeden Island contacts with Fort Walton. The Safety Harbor peoples were the sixteenth- and seventeenth-century Timucuans of the area. Changes in community type were effected in the northwest during Fort Walton with the appearance of the temple mound and the disap- pearance of the old burial-mound ceremonialism. There are few temple-mound sites on the immediate coast, but more are known in the interior. This interior country, particularly in Leon and Jefferson Counties, is more suitable for intensive agriculture than the sandy soils of the coastal strip, and it is likely that with the institution of a new socioeconomic and political system population shifted inland to an environment more favorable to the modifications in food economy. This is partially corroborated by the accounts of the sixteenth-century Spanish who report that the larger towns of the northwest were in the interior. A population estimate for the Fort Walton culture has been set at between 7,500 and 12,000 persons. For Safety Harbor an estimate of 6,000 persons is made, giving a total Gulf Coast popu- lation for the sixteenth century of somewhere in the neighborhood of 15,000, about the same as the estimated figure for each of the Weeden Island Periods. Duration of the Fort Walton Period is given as A.D. 1500 to 1650, the later date coinciding with the estab- lishment of the Spanish missions in the Gulf Coast territory and the beginning of the Leon-Jefferson Period. Leon-Jefferson lasted from 1650 to about 1725. During this time the Indian populations were considerably reduced. Changes in ceramic types between Weeden Island and Fort Walton are fairly abrupt. The new styles are closely related to those of central Alabama although there is some evidence of old local con- tinuity. Other artifact types are changed somewhat. [Fort Walton burial is in cemeteries or in the floors and flanks of platform mounds rather than in burial mounds, but the earlier prevalent custom of sec- ondary burial treatment continues. Englewood and Safety Harbor ceramic complexes are derived more directly out of Weeden Island, and the burial-mound practice seems to be universal in the south. Temple mounds do, however, come in during the Safety Harbor Period. The Leon-Jefferson culture shows the breaking up of the Indian patterns and the substitution of Spanish colonization. Native handi- 582 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. II3 crafts continue to some extent, and the pottery of the period can be derived from Fort Walton and from late influences out of Georgia. TRENDS A number of trends can be recapitulated from this record of Gulf Coast prehistory : 1. Food economy. Hunting, fishing, and gathering were succeeded by an economy that employed these older techniques together with maize horticulture. A late-period tendency is toward a more inten- sive horticulture, although the old food-gathering methods were still employed. This trend follows the general trend for the southeastern United States, and is historically conditioned by the diffusion of maize horticulture. 2. Settlements. Early settlements were small coastal villages. During the Santa Rosa-Swift Creek and Weeden Island Periods these villages remain much the same in size and location but they became much more numerous. In the late periods, especially Fort Walton, the small village community continues, but large towns also arise. There is a trend toward settlement in the interior rather than the coast at this time. These settlement trends are obviously conditioned by those of food economy. 3. Population. The population trend is steadily upward until a climax is reached during the Weeden Island Periods. It is doubtful if there were more people in the area during the Fort Walton and Safety Harbor Periods, but they were concentrated in larger communities. 4. Political organization. It is inferable that the trend was from numerous small autonomous units to larger political aggregates and, eventually, tribal federations. This trend was undoubtedly con- ditioned by population growth but is also explainable by historical factors. It is duplicated throughout the Southeast and, probably, the world, at least as far as early Neolithic sequences are concerned. 5. Ceremonialism. There is little reflection of ceremonialism in the Archaic periods, but the succeeding Burial Mound stage periods are rich in a ceremonialism which centers around a cult of the dead. There is a late shift away from this in Fort Walton when cere- monialism can more reasonably be said to have existed around the temple sites and the religious and political leadership which they symbolize. This trend is also noted throughout the Southeast. 6. Ceramic arts. Technical and artistic improvement is continued up through the Weeden Island I Period. From here on, technical WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—WILLEY 583 levels are maintained, but artistic quality may be lessened. This last judgment is, of course, subjective. It seems likely, though, that pot- tery was less intimately tied up with religious ritualism in the late periods, and this may account for an esthetic decline. 7. Other crafts. These are best developed in the Santa Rosa-Swift Creek Period after which work in stone, shell, and metals is less common. This pattern of occurrence, rather than trend, is probably explained by the following. 8. Trade. Outside exchange in raw materials and finished products was most vigorous during the Santa Rosa-Swift Creek Period. At least it is at this time that the most exotic goods are found in Gulf Florida sites. Trade, per se, should be distinguished from influence in this context. This pattern of occurrence is duplicated on the Hopewellian horizon in many parts of the eastern United States. AON, eRe an i ee nO ee aR f Ann ay HOLD ak bakers ; ua ,, v ‘i : Cree) | eat 4 ish SED tee i j j ) ‘ . ( ie ey 4 i { f {i ,. Oy oe a 1 \ ! } , [ cre: r ap ; | bind men era beteit Sit ase ber diy Sal ai vn eo ue ee hone ahaa re mi bi u a han ee oh pveltanyy i Te win) r¥ An it ; 1" Ay on iy ae ) Ht ‘hey i fc =f | PET avy off Nt ee sh +t ; ‘i i's f hig i 16 ; \ * : ae —_ et \ i) a } a \ rs 1 j / 8 \0A * r? #y) “\ . ¥ Bi MD ee ; elias ' ‘ he 1w) j bh Peery oer ra Yvan ; ] ae in ml Der a MeN i yt "} : vr * j a) uy ' > (S)) panes BIBLIOGRAPHY ALLEN, JoHN H. 1846. Some facts respecting the geology of Tampa Bay, Florida. Amer. Journ. Sci., ser. 2, vol. 1, pp. 38-42. ANONYMOUS. 1890. Pottery find in Florida. Amer. Antiquarian, vol. 12, pp. 185-186. 1925. Work in Florida. Expl. and Field-work Smithsonian Inst. in 1924, Smithsonian Misc. Coll., vol. 77, No. 2, pp. 93-98. 1926. Investigation of shell and sand mounds on Pinellas Peninsula, Florida. Expl. and Field-work Smithsonian Inst. in 1925, Smithsonian Misc. Coll., vol. 78, No. 1, pp. 125-132. 1937. Florida second biennial report of State Board of Conservation, Bi- ennium ending June 30, 1936, pp. 109-152. Tallahassee. 1939a. Notes on two interesting mounds excavated in Hillsborough County. Third Bien. Rep. Florida State Board Conserv., Biennium ending June 30, 1938, pp. 25-30. Tallahassee. 1939b. Aboriginal stone quarries of Hillsborough County and sources of abrasives and pigment. Third Bien. Rep. Florida State Board Conserv., Biennium ending June 30, 1938, pp. 31-32. Tallahassee. BarTRAM, WILLIAM. 1940. The travels of William Bartram. Edited by Mark Van Doren, New York. Brecker, R. B. 1944. An early Indian clam bake. Proc. Florida Acad. Sci., vol. 7, No. 1, Pp. 23-27. BELL, JAMES. 1883. Mounds in Alachua County, Florida. Ann. Rep. Smithsonian Inst. for 1881, pp. 635-637. Boyp, M. F. 1939. Mission sites in Florida. Florida Hist. Quart., vol. 17, No. 4, pp. 255- 280. Brinton, DANIEL GARRISON. 1859. Floridian peninsula, its literary history, Indian tribes, and antiquities. Philadelphia. 1867. Artificial shell deposits of the United States. Ann. Rep. Smithsonian Inst. for 1866, pp. 356-358. BROWER, CHARLES DE WOLFE. 1906. Shell heaps of Florida. Rec. Past, vol. 5, pt. 9, pp. 331-338. Brown, JoHN S. n.d. Shell heaps of Cedar Keys, Florida. Ms. Bur. Amer. Ethnol. vault. (Circa 1920.) CALDWELL, J. R., and McCann, CATHERINE. 1941. Irene Mound Site. Univ. Georgia Press, Athens. CALDWELL, J. R., and Warne, A. J., JR. 1939. Newsletter of the Southeastern Archaeological Conference, vol. 1, No. 6. (Mimeo.) Lexington, Ky. 39 585 586 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I13 CaLkKIns, W. W. 1877-1880. Notes of personal investigations among the shell mounds of Florida. Proc. Davenport Acad. Nat. Sci., 1876-78, vol. 2, pp. 225- 229. Davenport, Iowa. CLAFLIN, WitL1AM H., Jr. 1931. The Stallings Island Mound, Columbia County, Georgia. Peabody Mus. Pap., vol. 15, No. 1. Cambridge. Coe, Jorrre L. n.d. Report on mound work in the region of Lake Apopka, Florida. Ms. Bur. Amer. Ethnol. vault. (September 1933.) Conrap, T. A. 1846. Observations on the geology of a part of eastern Florida, with a catalogue of recent shells of the coast. Amer. Journ. Sci., ser. 2, vol. 2, pp. 36-48. Cooke, C. W., and Mossom, S. 1929. Geology of Florida. 20th Ann. Rep. Florida State Geol. Surv., pp. 31- 279. CUSHING, FRANK HAMILTON. 1897. Explorations of ancient key-dwellers remains on the Gulf Coast of Florida. Proc. Amer. Philos. Soc., vol. 35, pp. 320-448. Phila- delphia, Pa. Dati, WILLIAM HEALEY. 1887. Notes on the geology of Florida. Amer. Journ. Sci., ser. 2, vol. 34, art. 19, pp. 161-170. Dickinson, S. D. 1936. Ceramic relationships of the pre-Caddo pottery from the Crenshaw Site. Bull. Texas Arch. and Paleont. Soc., vol. 8. Ecker, A. 1878. Zur Kenntniss des Korperbaues friitherer Einwohner der Halbinsel Florida. Archiv. fiir Anthrop., vol. 10, pp. 101-114. EScALENTE FoNTANEDA, D. 1944. Memoir. Written in Spain about 1575. Translated and annotated by Buckingham Smith, 1854. Miami, Florida (Gulf Coast). FArrBANKsS, C. H. 1940. Newsletter of the Southeastern Archaeological Conference, vol. 2, No. 2. (Mimeo.) Lexington, Ky. 1942. The taxonomic position of Stallings Island. Amer. Antiquity, vol. 7, No. 3, pp. 223, 231. 1946. The Kolomoki mound group, Early County, Georgia. Amer. An- tiquity, vol. 11, pp. 258-260. Farser, G. C. 1887. Mound in Florida. Amer. Antiquarian, vol. 9, No. 5, pp. 307-308. FEATHERSTONHAUGH, THOMAS. 1897. Note on Florida archaeology. Amer. Anthrop., vol. 10, p. 200. 1899. Mound builders of central Florida. Harrisburg. n.d. An investigation of the Brooker mounds, Florida. Ms., U.S. Nat. Mus., Access. No. 24012. FENNEMAN, NEvIN M. 1938. Physiography of the eastern United States. (First edition.) New York. WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—WILLEY 587 FEWKES, JESSE WALTER. 1924. Preliminary archeological explorations at Weeden Island, Florida. Smithsonian Misc. Coll., vol. 76, No. 13, pp. 1-26, 21 pls. Forp, J. A., and Quimsy, G. I. 1945. The Tchefuncte culture, an early occupation of the Lower Mississippi Valley. Mem. Soc. Amer. Archaeol., No. 3. Forp, J. A., and WILLEy, Gorpon R. . 1939. Newsletter of the Southeastern Archaeological Conference, vol. 1, Nos. 3 and 4. (Mimeo.) Lexington, Ky. 1940. Crooks Site, a Marksville Period burial mound in La Salle Parish, Louisiana. Anthrop. Study No. 3, Dep. Conserv., Louisiana Geol. Surv. 1941. An interpretation of the prehistory of the eastern United States. Amer. Anthrop., vol. 43, No. 3, pp. 325-363. GILLMAN, HENRY. 1878. Crania utilized as cinerary urns in a burial mound in Florida. Amer. Natur., vol. 12, pp. 753-754. Philadelphia, Pa. 1879. Remarkable burial custom from a mound in Florida: the crania utilized as a cinerary urn. Proc. Amer. Assoc. Adv. Sci., sect. B., vol. 27, pp. 309-312. Salem, Mass. Gocecin, Joun M. 1930. A ceramic sequence in South Florida. New Mexico Anthrop., vol. 3, pp. 36-40. 1940. The distribution of pottery wares in the Glades archeological area of South Florida. New Mexico Anthrop., vol. 4, pp. 22-33. 1944a.A tentative formulation of pottery types for the Glades area. (Mimeo.) New Haven. 1944b. Archaeological investigations on the Upper Florida Keys. Tequesta, No. 4, pp. 13-38. Coral Gables, Fla. 1947a. Manifestations of a south Florida cult in northwestern Florida. Amer. Antiquity, vol. 12, No. 4, pp. 273-276. 1047b. A preliminary definition of archaeological areas and periods in Florida. Amer. Antiquity, vol. 13, pp. 114-127. 1948a. Some pottery types from central Florida. Gainesville Anthrop. Assoc. Bull. No. 1. 1948b. Florida archeology and recent ecological changes. Journ. Wash- ington Acad. Sci., vol. 38, No. 7, pp. 225-233. n.d.1 The archaeology of the Glades area, south Florida. Ms. n.d.2 Culture and geography in Florida prehistory. Ph.D. thesis, Yale University, New Haven. Gocctn, Joun M., and Rousg, IRvING. 1948. A West Indian ax from Florida. Amer. Antiquity, vol. 13, No. 4, pl. I, pp. 323-325. Gower, CHARLOTTE D. 1927. The northern and southern affiliations of Antillean culture. Mem. Amer. Anthrop. Assoc. No. 35. GREENMAN, EMERSON F. 1038. Hopewellian traits in Florida. Amer. Antiquity, vol. 3, No. 4, pp. 327-332. 588 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. II3 GRIFFIN, JAMEs B. 1943. An analysis and interpretation of the ceramic remains from two sites near Beaufort, South Carolina. Bur. Amer. Ethnol. Bull. 133, No. 22. 1945. The significance of the fibre-tempered pottery of the St. Johns area in Florida. Journ. Washington Acad. Sci., vol. 35, No. 7, pp. 218- 223. 1946. Cultural change and continuity in eastern United States archaeology. In Man in northeastern North America, Pap. R. S. Peabody Foundation, vol 3, pp. 37-95. Andover, Mass. GriFFIN, JoHN W. 1943. The Antillean problem in Florida archaeology. Florida Hist. Quart., vol. 22, No. 2, pp. 86-91. 1946. Historic artifacts and the Buzzard cult in Florida. Florida Hist. Quart., vol. 24, pp. 295-301. 1947. Comments on a site in the St. Marks National Wildlife Refuge, Wakulla County, Florida. Amer. Antiquity, vol. 13, No. 2, pp. 182- 183. n.d. Prehistoric Florida—A review. Ms. prepared for F-C. Cole Anniv. Vol. Haac, W. G. 1939. Newsletter of the Southeastern Archaeological Conference, vol. 1, No. 1. (Mimeo.) Lexington, Ky. 1942. Early horizons in the Southeast. Amer. Antiquity, vol. 7, No. 3, PP. 209-222. Harper, Rotanp M. 1914. Geography and vegetation of northern Florida. 6th Ann. Rep. Florida State Geol. Surv., pp. 165-437. HEILPRIN, ANGELO. 1887. Explorations on the west coast of Florida, and in the Okechobee Wilderness. Trans. Wagner Free Inst., vol. 1, pp. 1-134. Phila- delphia, Pa. Hewitt, J. F. 1898. Archaeology of the west coast of Florida. Medical Bull., vol. 20, 135 pp. Philadelphia, Pa. Hotmes, WILLIAM HEnrY. 1894a. Earthenware of Florida: Collections of Clarence B. Moore. Journ. Acad. Nat. Sci. Philadelphia, vol. 10, pp. 105-128. 1894b. Caribbean influence on the prehistoric ceramic art of the southern States. Amer. Anthrop., vol. 7, pp. 71-78. 1903. Aboriginal pottery of the eastern United States. 20th Ann. Rep. Bur. Amer. Ethnol. 1898-99, pp. 1-237. 1914. Areas of American culture characterization tentatively outlined as an aid in the study of the antiquities. Amer. Anthrop., vol. 18, PP. 413-446. HroiicKa, ALES. 1907. Skeletal remains suggesting or attributed to Early man in North America. Bur. Amer. Ethnol. Bull. 33. 1922. Anthropology of Florida. Publ. Florida Hist. Soc., No. 1. WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—WILLEY 589 1940. Catalog of human crania in the United States National Museum collections: Indians of the Gulf States. Proc. U.S. Nat. Mus., vol. 87, pp. 315-464. JENNINGS, JESSE D. 1941. Chickasaw and earlier Indian cultures of northeast Mississippi. Journ. Mississippi Hist., vol. 3, No. 3, pp. 155-226. JENNINGS, J. D., and Farrsanxs, C. H. 1939. Newsletter of the Southeastern Archaeological Conference, vol. 1, No. 2. (Mimeo.) Lexington, Ky. Ketry, A. R. ‘ 1938. A preliminary report on archeological explorations at Macon, Ga. Bur. Amer. Ethnol. Bull. 119, No. 1. KENWORTHY, CHARLES J. 1883. Ancient canals in Florida. Ann. Rep. Smithsonian Inst. for 1881, Pp. 631-635. KIMBALL, Mary R. 1872. Florida Indians. Bull. Essex Inst., vol. 4, pp. 18-19. Salem, Mass. Kriecer, ALEX D. 1944. The typological concept. Amer. Antiquity, vol. 9, No. 3, pp. 271-288. 1945. An inquiry into supposed Mexican influence on a prehistoric “cult” in the southern United States. Amer. Anthrop., vol. 47, No. 4, pp. 483-515. 1946. Culture complexes and chronology in northern Texas. Univ. Texas Publ. No. 4640. Austin. 1947. The eastward extension of Puebloan dating toward cultures of the Mississippi Valley. Amer. Antiquity, vol. 12, pp. 141-148. Kroeser, A. L. 1939. Cultural and natural areas of native North America. Univ. Calif. Publ. in Amer. Archaeol. and Ethnol., vol. 38. Kunz, G. F. 1887. Gold and silver ornaments from mounds of Florida. Amer. Anti- quarian, vol. 9, p. 2109. Lawson, R. H. 1918. Explorations at Pensacola, Florida. Archaeol. Bull., vol. 9, No. 6, pp. 67-68. LeBaron, J. FRANCIS. 1884. Prehistoric remains in Florida. Ann. Rep. Smithsonian Inst. for 1882, pp. 771-790. Lerpy, JOSEPH. 1889. Notice of some fossil human bones. Trans. Wagner Free Inst., vol. 2, PP. 9-12. LEeMoyne pE MorcuEs, JACQUES. 1946. The narrative of Jacques Le Moyne de Morgues. The New World, pp. 33-97. Stefan Lorant, editor. New York. Lewis, T. M. N. 1931. A Florida burial ground. Wisconsin Arch., n. s., vol. 10, No. 4, pp. 123-129. Lewis, T. M. N., and KNEBEerc, MADELEINE. 1946. Hiwassee Island. Univ. Tennessee Press. 590 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. II3 Martens, J. H. C. 1931. Beaches of Florida. 21st and 22d Ann. Reps. Florida State Geol. Surv., pp. 69-119. Martin, Paut S., Quimpy, G. I., and Cottier, DonaLp. 1947. Indians before Columbus. Chicago. MAstIus, VERA. n.d. Excavations at South Indian Field, Brevard County. Ms. Mooney, JAMES. 1928. The aboriginal population of America north of Mexico. Smithsonian Misc. Coll., vol. 80, No. 7. Moorg, C. B. 1894a. Certain sand mounds of the St. Johns River, Florida, pt. 1. Journ. Acad. Nat. Sci. Philadelphia, vol. 10, pp. 130-248. 1894b. Certain sand mounds of the St. Johns River, Florida, pt. 2. Journ. Acad. Nat. Sci. Philadelphia, vol. 10. 1895a. Certain sand mounds of the Ocklawaha River, Florida. Journ. Acad. Nat. Sci. Philadelphia, vol. ro. 1895b. Two mounds on Murphy Island, Florida. Journ. Acad. Nat. Sci. Philadelphia, vol. ro. 1896. Certain river mounds of Duval County, Florida. Journ. Acad. Nat. Sci. Philadelphia, vol. ro. 1900. Certain antiquities of the Florida west coast. Journ. Acad. Nat. Sci. Philadelphia, vol. 11. 1901. Certain aboriginal remains of the northwest Florida coast, pt. 1. Journ. Acad. Nat. Sci. Philadelphia, vol. 11. 1902. Certain aboriginal remains of the northwest Florida coast, pt. 2. Journ. Acad. Nat. Sci. Philadelphia, vol. 12. 1903. Certain aboriginal mounds of the Florida central west coast. Journ. Acad. Nat. Sci. Philadelphia, vol. 12. 1905. Aboriginal remains of Black Warrior River, Lower Tombigbee River, Mobile Bay and Mississippi Sound, and miscellaneous investigations in Florida. Journ. Acad. Nat. Sci. Philadelphia, vol. 13. 1907. Moundville revisited: Crystal River revisited: Mounds of the lower Chattahoochee and lower Flint Rivers: Notes on the Ten Thou- sand Islands, Florida. Journ. Acad. Nat. Sci. Philadelphia, vol. 13. 1910. Antiquities of the St. Francis, White, and Black Rivers, Arkansas. Journ. Acad. Nat. Sci. Philadelphia, vol. 14. 1918. The northwestern Florida coast revisited. Journ. Acad. Nat. Sci. Philadelphia, vol. 16. NeEwELL, H. Perry, and Kriscer, A. D. 1949. The George C. Davis site, Cherokee County, Texas. Amer. Antiq., vol. 14, No. 4, pt. 2. Newnmav\, M. T., and Snow, C. E. 1943. Preliminary report on the skeletal material from Pickwick Basin, Alabama. Bur. Amer. Ethnol. Bull. 129, pp. 395-507. Oscoop, CoRNELIUS. 1942. The Ciboney culture of Cayo Redondo, Cuba. Yale Univ. Publ. in Anthrop., No. 25. (Gulf Coast.) QuimBy, GEORGE, JR. 1942. The Natchezan culture type. Amer. Antiquity, vol. 7, No. 3, pp. 255- 275. WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—-WILLEY 591 RaIneEy, F. G. 1935. Indian burial site at Crystal River, Florida. Florida Hist. Quart., vol. 13, No. 4, pp. 185-192. AU, GC. 1878. Observations on a gold ornament from a mound in Florida. Ann. Rep. Smithsonian Inst. for 1877, pp. 298-302. Rouse, IRVING. 1939. Prehistory in Haiti, a study in method. Yale Univ. Publ. in Anthrop., No. 21. 1940. Some evidence concerning the origins of West Indian pottery making. Amer. Anthrop., vol. 42, No. 1, pp. 49-80. 1948. The Arawak. In Handbook of South American Indians, vol. 4, pt. 3, pp. 507-546, Bur. Amer. Ethnol. Bull. 143. Scuorr, H. L. n.d. Letter on Pool Hammock, Florida, near Laurel, Sarasota County. Bur. Amer. Ethnol. vault. (January 10933.) ScHootcrarT, H. R. 1854. Antique pottery from the minor mounds occupied by the Indians in feasts to the dead, on the sea-coast of Florida and Georgia. Indian Tribes of the United States, pt. III, pp. 75-82. SETZLER, FRANK M. 1933. Pottery of the Hopewell type from Louisiana. Proc. U. S. Nat. Mus., vol. 82, No. 2963, art. 22, pp. I-21. SHEPARD J. 1886. Shell heaps and mounds in Florida. Ann. Rep. Smithsonian Inst. for 1885, pt. I, pp. 902-906. Simons, M. H. 1884. Shell heaps of Charlotte Harbor, Florida. Ann. Rep. Smithsonian Inst. for 1882, pp. 794-706. Simpson, J. C. 1941. Source material for Florida aboriginal artifacts. Proc. Florida Acad. Sci. 1940, vol. 5, pp. 32-34. SLEIGHT, F. W. 1943. Archaeological needs for Florida. Amer. Antiquity, vol. 8, No. 4, Pp. 387-392. SMALL, JoHn K. 1924. The land where spring meets autumn. Journ. New York Bot. Gard., vol. 25, pp. 53-94. 1927. Among floral aborigines. Journ. New York Bot. Gard., vol. 28, pp. 1-20 and 25-40. SmitH, HALE G. 1948. Two historical archaeological periods in Florida. Amer. Antiquity, vol. 13, No. 4, pt. I, pp. 313-319. SmirH, Ruea M. 1933. Anthropology in Florida. Florida Hist. Quart., vol. 11, No. 4, pp. 151-172. Snow, C. E. 1948. Indian Knoll skeletons from site Oh2, Ohio County, Kentucky. Rep. in Anthrop., vol. 4, No. 3, pt. 2. Univ. Kentucky. 592 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I13 STEARNS, R. E. C. 1870. Rambles in Florida. Amer. Nat., vol. 3, pp. 349-360, 307-405, 455-470. 1872. Remarks on mounds and shell heaps of Tampa Bay. Proc. California Acad. Sci., ser. 1, vol. 4, pp. 214-215. STERNBERG, G. M. 1876. Indian burial mounds near Pensacola, Florida. Proc. Amer. Assoc. Adv. Sci., vol. 24, pt. 2, pp. 282-292. Salem, Mass. STEWART, T. D. 1940. Some historical implications of physical anthropology in North America. Smithsonian Misc. Coll., vol. 100, pp. 15-50. 1946. A reexamination of the fossil human skeletal remains from Mel- bourne, Florida, with further data on the Vero skull. Smithsonian Misc. Coll., vol. 106, No. 10. Stretinc, M. W. 1930. Prehistoric mounds in the vicinity of Tampa Bay, Florida. Expl. and Field-work Smithsonian Inst. in 1929, pp. 183-186. 1931. Mounds of the vanished Calusa Indians of Florida. Expl. and Field- work Smithsonian Inst. in 1930, pp. 167-172. 1935. Smithsonian archeological projects conducted under the Federal Emergency Relief Administration, 1933-34. Ann. Rep. Smithsonian Inst. for 1934, pp. 371-400. 1936. Florida cultural affiliations in relation to adjacent areas. In Essays in Anthropology, in honor of Alfred Louis Kroeber, pp. 351-357. Stone, Doris Z. 1939. Relationship of Florida archaeology to that of middle America. Florida Hist. Quart., vol. 17, No. 3, pp. 211-218. Stupss, Sipney A. 1940. The future of Florida archaeological research. Proc. Florida Acad. Sci., vol. 4, pp. 266-270. SWANTON, JOHN R. 1922. Early history of the Creek Indians and their neighbors. Bur. Amer. Ethnol. Bull. 73. 1946. The Indians of the Southeastern United States. Bur. Amer. Ethnol. “Bullk 137: THoMAS, CyRUSs. 1891. Catalogue of prehistoric works east of the Rocky Mountains. Bur. Amer. Ethnol. Bull. 12. 1894. Report on the mound explorations of the Bureau of Ethnology. 12th Ann. Rep. Bur. Amer. Ethnol. Unitep STATES DEPARTMENT OF AGRICULTURE YEARBOOK. 1941. Climate and man. VERNON, RoseErt O. 1942. Geology of Holmes and Washington Counties, Florida. Florida State Geol. Surv. Bull. No. 21. Vocpes, A. W. 1879. Notes on a lost race of America. Amer. Nat., vol. 13, pp. 9-II. Von Bonrn, G., and Morant, G. M. 1938. Indian races in the United States. A survey of previously published measurements. Biometrika, vol. 30, pp. 94-129. WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—WILLEY 593 Von Ercxstept, E. F. 1934. Rassenkunde und Rassengeschichte der Menschheit. F. Enke Verlag. Stuttgart, 935 pp. Warnwricat, R. D. 1916. Two months research in the sand and shell mounds of Florida. Archaeol. Bull., vol. 7, No. 6, pp. 139-144. 1918. Archaeological exploration in southern Florida, 1917. Archaeol. Bull., vol. 9, Nos. 3 and 4, pp. 28-32, 43-47. WALKER, S. T. 1880a. Preliminary explorations among the Indian mounds in southern Florida. Ann. Rep. Smithsonian Inst. for 1879, pp. 392-413. 1880b. Report on the shell heaps of Tampa Bay, Florida. Ann. Rep. Smith- sonian Inst. for 1879, pp. 413-422. . 1883. The aborigines of Florida. Ann. Rep. Smithsonian Inst. for 1881, pp. 677-680. 1885. Mounds and shell heaps on the west coast of Florida. Ann. Rep. Smithsonian Inst. for 1883, pp. 854-868. Warine, A. J., JR., and HoLpEer, PRESTON. 1945. A prehistoric ceremonial complex in the southeastern United States. Amer. Anthrop., vol. 47, No. 1, pp. 1-34. Wess, Wituiam S., and DEJARNETTE, Davin L. 1942. An archeological survey of Pickwick Basin in the adjacent portions of the States of Alabama, Mississippi, and Tennessee. Bur. Amer. Ethnol. Bull. 120. WILEY, Gorpon R. 1939. Ceramic stratigraphy in a Georgia village site. Amer. Antiquity, vol. 5, No. 2, pp. 140-147. 1945. The Weeden Island Culture: a preliminary definition. Amer. An- tiquity, vol. 10, No. 3, pp. 225-254. 1948a. Culture sequence for the Manatee region of West Florida. Amer. Antiquity, vol. 13, No. 4, pp. 209-218. 1948b. The cultural context of the Crystal River negative painted style. Amer. Antiquity, vol. 13, No. 4, pt. 1, pp. 325-328. 1948c. A proto-type for the Southern Cult. Amer. Antiquity, vol. 13, No. 4, pt. I, pp. 328-330. n.d. Excavations in southeast Florida. Ms. WILEY, Gorpon R., and PHILLIPS, PHILIP. 1944. Negative-painted pottery from Crystal River, Florida. Amer. An- tiquity, vol. 10, No. 2, pp. 173-185. Witey, Gorpon R., and Woopsury, R. B. 1942. A chronological outline for the northwest Florida coast. Amer. Antiquity, vol. 7, pp. 232-254. Wimperty, STEVE B., and Tourtetot, Harry A. 1941. The McQuorquodale Mound: A manifestation of the Hopewellian phase in south Alabama. Geol. Surv. of Alabama, Mus. Pap. 10, 42 pp. WISSLER, CLARK. 1938. The American Indian. New York. 594 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 113 Wyman, J. 1870. Explorations in Florida. Third Ann. Rep. Trustees Peabody Mus., Harvard Univ., pp. 8-9. 1875. Fresh-water shell mounds of the St. Johns River, Florida. Mem. Peabody Acad. Sci., vol. 1, No. 4. Salem, Mass. EXPLANATION OF PLATES Plate i. Pit excavation at Carrabelle (Fr-2), Franklin County. Top: Excavating the isolated strati-block of pit Il; bottom: test pit on completion, show- ing midden and shells to a depth of 1 meter. Shell middens on the northwest Gulf Coast. Top: East Point midden, Franklin County ; bottom: beginning a test pit at Mound Field (Wa-8), Wakulla County. Lake Jackson site (Le-1), Leon County. Top: View of the lake bed from the bordering hills above the site; bottom: mound 6 (artificial height about 2.5 meters). Burial types from the Thomas (Hi-1) and Cockroach Key (Hi-2) burial mounds. Top: Flexed and vertical bundle burials from Thomas; center: flexed burial from Cockroach Key; bottom: secondary bundle burials, vertical and horizontal, at Cockroach Key. The Englewood mound (So-1). Top: Before clearing; bottom: exca- vations showing old sod line near bottom of cut on the 30-foot profile line. 6. Englewood (So-1) burials. Top: Flexed; bottom: bundle. 7. Perico Island (Ma-6). Top, left: The burial mound before excavation; Io. Il. 12. 13. 14. top, right: typical flexed burial from the burial mound; center, left: excavating the burial mound; center, right: excavating the smaller midden; bottom, left: the cemetery site; bottom, right: three flexed burials in the cemetery. Cockroach Key (Hi-2). Top: A small area (3.5x6 feet) with the bones of 15 secondary burials massed together, burial mound; center: look- ing north from the highest shell accumulation on Cockroach Key; bottom: the burial mound after clearing, looking north. The northwest Gulf Coast. Top: Looking southward across the “trough” or depression between the dune ridges on Santa Rosa Island. The man in the center is standing in the midst of a typical midden site of the island; bottom: a view of Nine-Mile Point (Fr-9), Franklin County. The northwest Gulf Coast. Top: Tucker site (Fr-4), Franklin County, looking down the cleared “right-of-way” over the shell midden; bottom: the destroyed Hall site (Wa-4) midden, Wakulla County. The Pierce site (Fr-14), Franklin County. Top: Area of destroyed mid- den; bottom: cross section of flat-topped mound as seen from railroad cut. This mound may be Moore’s Pierce mound C. Deptford Period sherd types. All Deptford Linear Check Stamped from the northwest coast. Deptford Period sherd types. a-c, Deptford Bold Check Stamped; d-f, Deptford Simple Stamped; g, St. Simons Plain. From the northwest coast. Perico Island Period sherds. a, d, Perico Island Incised; b, c, e, Perico Island Linear Punctated; f, unclassified incised sherd on Biscayne-type 595 596 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I13 Plate paste; g, Deptford Bold Check Stamped podal support; 4, complicated stamped sherd; i, Glades Plain with unusual rim projection. All from Perico Island site. (Courtesy U.S. National Museum; Nos. 383990, 383996, 384005, 384000, 384004, 383080, 384027, 384032, 384030.) 15. Perico Island Period artifact types. a, flat-surface or Strombus shell celt; 16. 17. 18. 10. 20. 2I. 22. 23) 24. b, single-grooved columella pendant or plummet; c, d, inner and outer sides of concave-surface or Busycon celt; e, f, conch columella chisel and hammer; g, h, Busycon hammers; i, 1, hammerstones; 7, single- grooved stone plummet; k, grinding stone; m, bone dagger or point. All from Perico Island site except e-h, j, and 1, which are from Shaws Point, Manatee County. (Courtesy U.S. National Museum; Nos. 384019, 384058, 384050, 384057, 384057, 384060, 384060, 384016, 384062, 384018, 384063, 384021.) Perico Island Period artifact types. a, b, flat-surface or Strombus celts; c, concave-surface or Busycon celt; d, bipointed bone projectiles; e, shell bead; f, 7, chipped-stone projectiles; g, double-grooved columella pendant or plummet; /, bone awl; i, pierced turtle-carapace fragment, probably from a gorget. All from the Cockroach Key site. (Courtesy U.S. National Museum; Nos. 384297, ?, 384207, 384302, and 384204, 384209, 384301, 384305, °, 384300, 384306.) Santa Rosa-Swift Creek Period sherd types. All Alligator Bayou Stamped from the northwest coast. Santa Rosa-Swift Creek Period sherd types. a-c, Santa Rosa Stamped; d-h, Basin Bayou Incised; i-7, Santa Rosa Punctated. From the north- west coast. Santa Rosa-Swift Creek Period sherd types. a, b, Crooked River Com- plicated Stamped; c-e, Alligator Bayou Stamped. From the northwest coast. Santa Rosa-Swift Creek Period sherd types. All Swift Creek Complicated Stamped, Early Variety, from the northwest coast. Santa Rosa-Swift Creek Period sherd types. a-e, New River Complicated Stamped; f, g, Gulf Check Stamped; hf, West Florida Cord-marked, Early Variety; i, tetrapod base of a Franklin Plain vessel. From the northwest coast. Santa Rosa-Swift Creek Period vessels. a, Pierce Zoned Red; b, plain miniature vessel, unclassified but probably of this period; c, d, f, g, Swift Creek Complicated Stamped, Early Variety; e, Franklin Plain. (Courtesy R. S. Peabody Foundation, Andover, from the following sites in order with catalog numbers: Pierce mound A, 39301; Yent, 39175; Anderson’s Bayou, 38945; Green Point, 39248; Franklin County, 39238. Respective heights as follows: a, 14 cm.; b, 6 cm.; c, 19 cm.; d, 22 cm.; é, TO cm: fF; 13Vemis-g; Toem.) Santa Rosa-Swift Creek Period vessels. Crystal River Negative Painted. (a, courtesy R. S. Peabody Foundation, Green Point site, No. 39147; b, c, courtesy Heye Foundation, Crystal River site, Nos. 17/3523, 18/326. Respective rim diameters: a, 7 cm.; b, 9.7 cm.; ¢, 16.5 cm.) Santa Rosa-Swift Creek Period artifacts. a, double-grooved shell plummet or pendant; b, socketed bone projectile point; c, d, chipped-stone pro- jectile points; e, g, h, pottery pipes; f, shell gorget; 7, 7, copper ear WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—WILLEY 597 Plate 25. 26. 27. 28. 20. 30. 31. 32. 33- 34. 35. 36. 37. ornament plated with silver. (Artifacts a-d from northwest Florida middens. All others courtesy Heye Foundation with sites as follows: e, Huckleberry Landing, No. 17/1092; f, Crystal River; i, Crystal River, No. 17/62. Artifact a is 10 cm. long and b-d are to same scale; 4 and 7 are about 6 cm. in diameter; no scale for other objects.) Weeden Island Period sherd types. a-f, Weeden Island Incised; g, h, Weeden Island Punctated. From the northwest coast. Weeden Island Period sherd types. a-e, Weeden Island Incised; f-h, Weeden Island Zoned Red. From the Weeden Island site. (Courtesy U.S. National Museum; Nos. 360390, 325671, 325671, 325671, 325671. 360388, 369388, 360388.) Weeden Island Period sherd types. a, d, e, g, h, Weeden Island Punctated ; b, c, f, Papys Bayou Punctated. From the Weeden Island site. (Cour- tesy U.S. National Museum; Nos. 369390(a), 325671 (all others).) Weeden Island Period sherd types. b, Papys Bayou Punctated; all others Weeden Island Punctated. From Thomas mound. (Courtesy U.S. National Museum; Nos. 384210, 384207, 384267, 384187, 384201, 384156, 384151, 384233.) Weeden Island Period sherd types. a, Weeden Island Zoned Red; all others, Carrabelle Incised. From northwest Florida. Weeden Island Period sherd types. All Carrabelle Punctated. From north- west Florida. Weeden Island Period sherd types. a, b, Carrabelle Punctated; c-f, Indian Pass Incised. From northwest Florida. Weeden Island Period sherd types. a-d, Keith Incised; e-f, Tucker Ridge- pinched; g, h, St. Petersburg Incised. From northwest Florida. Weeden Island Period sherd types. a-e, Carrabelle Incised; f-i, Keith In- cised; j, k, St. Petersburg Incised. From the Weeden Island site. (Courtesy U.S. National Museum; Nos. 3609387, 325671, 325671, 325671, 325671, 325671, 325671, 330622, 325671, 325671, 325671.) Weeden Island Period sherd types. All Swift Creek Complicated Stamped, Late Variety. From northwest Florida. Weeden Island Period sherd types. a-f, Swift Creek Complicated Stamped, probably Late Variety; g-7, Tampa Complicated Stamped; k, Old Bay Complicated Stamped. From Thomas mound. (Courtesy U.S. National Museum; Nos. 384288, 384167, 384163, 384200, 384256, 384198, 384233, 384233, 384238, ?, 384172.) Weeden Island Period sherd types. a, Swift Creek Complicated Stamped, probably Late Variety; b, c, Old Bay Complicated Stamped; d, e, Sun City Complicated Stamped; f, unidentified stamped sherd; g, h, Little Manatee Zoned Stamped; i, Ruskin Dentate Stamped. From the Weeden Island site. (Courtesy U.S. National Museum; Nos. 325671, 324671, 369389, 325671, 325671, 325671, 369390, 360390, 325670.) Weeden Island Period sherd types. a, b, Ruskin Dentate Stamped; c, d, Thomas Simple Stamped; e, f, Little Manatee Zoned Stamped; g, Ruskin Linear Punctated. All from Thomas mound. (Courtesy U.S. National Museum; Nos. ?, ?, 384230, 384231, 384264, 384257, 384158.) 598 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I13 Plate 38. Weeden Island Period sherd types. a, Little Manatee Shell Stamped; b-g, Hillsborough Shell Stamped; h, unclassified fabric-marked sherd. From Thomas Mound. (Courtesy U.S. National Museum; Nos. 384144, 384266, 384226, ?, ?, 384195, ?, 384265.) 39. Weeden Island Period sherd types. a, b, Wakulla Check Stamped; c-f, Weeden Island Plain. From northwest Florida. 40. Weeden Island Period sherd types. a-f, Wakulla Check Stamped; g, Bis- cayne Check Stamped. From Thomas mound. (Courtesy U.S. National Museum; Nos. 38429, 384203, 384206, 384238, 384253, 384216, ?.) 41. Vessels of various periods. a, Oklawaha Plain; b, human-effigy vessel of the type Weeden Island Plain; c, human-figurine vessel in the Weeden Island style; d, Deptford Simple Stamped; e, compartment tray of the type Weeden Island Plain; f, Basin Bayou Incised. From the following sites: a, El Dorado Lake, Lake county; b, Warrior River, mound A; c, Aucilla River; d, Carrabelle midden and cemetery ; f, Warrior River, mound B; g, Pearl Bayou mound. (Courtesy Heye Foundation; Nos. 17/4501, 17/3947, 8/4154, 8/4153, 17/3948, 17/4038.) 42. Artifacts of various periods. a-f, Weeden Island Period chipped-stone projectiles; g, bird-efigy pipe fragment of pottery belonging to the Weeden Island Period; h, Santa Rosa-Swift Creek Period stone celt; i, Weeden Island Period stone celt; 7, k, Fort Walton Period stone celts. Celt form 7 also occurs in the Weeden Island Period. Artifacts a-g, from midden sites in northwest Florida; h, Jackson mound; 7, Calhoun County; j, Fort Walton; k, Fort Walton. (Celts are shown courtesy R. S. Peabody Foundation; Nos. 40382, 40376, 18586, no number.) 43. Fort Walton Period sherd types. All Fort Walton Incised. From north- west Florida. 44. Fort Walton Period sherd types. All Lake Jackson Plain. From north- west Florida. 45. Fort Walton Period sherd types. a-e, Point Washington Incised; f, g, Fort Walton Incised. From northwest Florida. 46. Englewood Period and associated vessel types. a, b, Englewood Incised; c, Sarasota Incised; d, e, Englewood Plain; f, St. Petersburg Incised, although fluting on body is unusual; g, Biscayne Roughened. From the Englewood mound. (Courtesy U.S. National Museum; Nos. 383166, 383168, 383171, ?, 383160, 383167, 383165.) 47. Englewood Period sherd types. Ail Englewood Incised. From Englewood mound. (Courtesy U.S. National Museum; Nos. 383170, 383187, 383178, 383189, 383185, 383189, 383183.) 48. Englewood Period and associated sherd types. a, b, unclassified incised with sherd b showing evidences of a former strap-handle attachment; c-f, Lemon Bay Incised; g, h, Sarasota Incised. From Englewood mound. (Courtesy U.S. National Museum; Nos. 383181, 383184, 383181, 383185, 383180, 383183, 383185, 383182.) 49. Safety Harbor Period sherd types. All Safety Harbor Incised. From Safety Harbor burial mound. (Courtesy U.S. National Museum, No. 351520.) 50. Safety Harbor Period and associated sherd types: a-d, f, Pinellas Plain; e, Biscayne Check Stamped. From Safety Harbor burial mound. (Courtesy U.S. National Museum; Nos. 351520, 351520, 351522, 351520, 35121, 351522.) WHOLE VOL. ARCHEOLOGY OF FLORIDA GULF COAST—WILLEY 599 Plate 51. Safety Harbor Period sherd types. All Pinellas Incised. From Safety Har- bor burial mound. (Courtesy U.S. National Museum; No. 35120.) 52. Safety Harbor Period sherd types. a, Pinellas Incised; b, Pinellas Plain; c-h, Safety Harbor Incised. Intrusive sherds from the Thomas mound. (Courtesy U.S. National Museum; Nos. 384286, 384283, 384263, 384173, 38410, 384238, ?, 384241.) 53. Safety Harbor Period and associated vessel types. a, Pinellas Incised ; b, d, e, Glades Plain; c, Lamar-like Complicated Stamped. Vessel D from Parrish mound 3; all others from Parrish mound 1. (Courtesy U.S. National Museum; Nos. 383190, 383228, 383192, 383191, 383193.) 54A. Safety Harbor Period and associated vessel types. a, b, Opposite sides of a Safety Harbor Incised vessel (see fig. 64); c, Lamar-like Compli- cated Stamped; d, Pasco Red; e, Safety Harbor Incised. All from Par- rish mound 3. (Courtesy U.S. National Museum; Nos. 383225, 383224, 383226, 383227.) 54B. Safety Harbor Period chipped-stone projectiles and blades. From Par- rish mound 3. (Courtesy U.S. National Museum; Nos. 383231 (0), d), 383230 (all others).) 55. Safety Harbor Period artifacts. Chipped-stone projectile points and blades. From the Safety Harbor village site. (Courtesy U.S. National Museum; Nos. 362383(b), 362385 (c), 351530 (all others).) 56. Safety Harbor Period artifacts. Chipped-stone projectile points. From Parrish mound 1. (Courtesy U.S. National Museum; No. 383201.) 57. Safety Harbor Period artifacts. a-j, m-o, chipped-stone projectiles and blades; k-I, p-u, chipped-stone scrapers; v, green-glaze European pipe; w, x, European trade pipes. From Safety Harbor village site and burial mound. (Courtesy U.S. National Museum; Nos. 351530 (a-i) ; others 351515, 351520, 362384, 362383, 362383, 362383, 351520, 362384, 351529, 351520, 362384, 351530, ?, ?, 362386.) 58. Safety Harbor Period artifacts. a, b, European combs of bone or tortoise shell; c, twisted cone of sheet gold; d, beads of sheet silver; e, green glass pentagonal bead; f, green glass pendant; g, copper ear ornament with hollow bulbar center. From Parrish mound 1. (Courtesy U.S. National Museum; Nos. 383199, 383198, 383206, 383204, 383205, 383203.) 59. Safety Harbor Period artifacts. a, European comb of bone or tortoise shell ; b, European brass pendant; c, single-grooved stone plummet; d, e, chipped-stone projectiles; f, ground-stone tubular bead; g, h, shell beads; 7-1, charred fragments of carved wood; m-q, charred fragments of hair cordage which have outer woven sheathing covering inner braided cord. From Parrish mound 2. (Courtesy U.S. National Museum; Nos. 383215, 383218, 383216, 383220, 383217, 383213, 383213, 383214 (all wood), 383221 (all cordage).) 60. Leon-Jefferson Period sherds and artifacts. a-c, Leon Check Stamped; d-f, complicated stamped ware, probably of the Jefferson Series; g, European crockery; h, iron spike; 7, gun flint. From Jefferson County. 4i (4 t ewer rh alt ‘ ’ = Ped . i! ‘ j ; * { i o 4 m i i a f if mt ay ‘ * j | i i j ‘ Vea re ‘ wer a Geary a " fay or ‘0 he cia “Siete ee aia ie ete ae i | i nf Senygaeye itpebonittet vr He in Auiiealt eh huhuliee lek Pent e cr ou! ; ; ‘ t fh Tt eh et aMits i) Poon eae Sie Path! BG Fete at ainsi | de Ped i ener ye aye igh SONS a, ‘ | 1 ft A Tg Bel 0. 4 chon pen 4 an : v Pap F wainge | RAa Vee Pe laa % pee tr airs a f ft my i} My 7 me Sa) NCP aa ne oe b leetar ge Latengat ay A ae ee . tial uf Wa ‘ iT aa? Dae Teer ' : weak Pye rlew Stent Pa 1a? die ' L s | , " | i (PT ry Maiyt te ar i iat ae (i i, iy ” | 9 TMC te vr ; ' t (hy ee 4 ti ; Ashi A Di oett ' f ‘ v3 F tear | ) 5 : ‘ ne) 1 ae t ¥ iste J ’ Mf h \ 4 ' i ihe i { ae Li , \ P ley} \ bathe paper By Li hi i ’ , , ' * t) ‘ J dt J Noe j SMITHSONIAN MISCELLANEOUS COLLECTIONS VOR 3, PES PIT EXCAVATION AT CARRABELLE (FR-2), FRANKLIN COUNTY (For explanation, see p. 595.) 40 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL: 133 Rhee SHELL MIDDENS ON THE NORTHWEST GULF COAST (For explanation, see p. 595.) SMITHSONIAN MISCELLANEOUS COLLECTIONS NAHE. alatey, Jaa) LAKE JACKSON SITE (LE-1), LEON COUNTY (For explanation, s SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL SLL3 Pie BURIAL TYPES FROM THE THOMAS (HI-1) AND COCKROACH KEY (HI-2) BURIAL MOUNDS (For explanation, see p. 595.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VO Eda SPS THE ENGLEWOOD MOUND (So-1) (For explanation, see p. 505.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. Lis Pies6 ENGLEWOOD (SO-1) BURIALS (For explanation, see p. 595.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOES aes RLS 7. PERICO ISLAND (MA-6) (For explanation, see p. 595.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 213; Res COCKROACH KEY (HI-2) (For explanation, see p. 595.) SMITHSONIAN MISCELLANEOUS COLLECTIONS NAgES aiilc, [eee Sl So RE IE OE a se epee THE NORTHWEST GULF COAST (For explanation, see p. 505.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 13 PE THE NORTHWEST GULF COAST (For explanation, see p. 505.) SMITHSONIAN MISCELLANEOUS COLLECTIONS Wfojts alates, ek. plat & ee oy m THE PIERCE SITE (FR-14), FRANKLIN COUNTY (For explanation, see p. 505.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 123, PERS CA. DEPTFORD PERIOD SHERD TYPES (For explanation, see p. 505.) Pais VOL. 113, SMITHSONIAN MISCELLANEOUS COLLECTIONS DEPTFORD PERIOD SHERD TYPES 5.) (For explanation, see p. 59 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOLE. 2s Rees PERICO ISLAND PERIOD SHERDS (For explanation, see pp. 595-506.) VOLE sis Pleas SMITHSONIAN MISCELLANEOUS COLLECTIONS PERICO ISLAND PERIOD ARTIFACT TYPES 5090. ) see p. For explanation, ( SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 3 Bale APPRX SCALE IN CAN PERICO ISLAND PERIOD ARTIFACT TYPES (For explanation, see p. 506.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL is; PEs 27 Ch7. SANTA ROSA-SWIFT CREEK PERIOD SHERD TYPES (For explanation, see p. 506.) 41 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 113, PL. 18 CA. SANTA ROSA-SWIFT CREEK PERIOD SHERD TYPES (For explanation, see p. 596.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 113; PL. SANTA ROSA-SWIFT CREEK PERIOD SHERD TYPES (For explanation, see p. 596.) 19 VIAN MISCELLANEOUS COLLECTIONS VOL. 113, PE. 26 CAF: SANTA ROSA-SWIFT CREEK PERIOD SHERD TYPES (For explanation, see p. 506.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL) 113, PL 22 CA. SANTA ROSA-SWIFT CREEK PERIOD SHERD TYPES (For explanation, see p. 596.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOE> Lis Piee22 SANTA ROSA-SWIFT CREEK PERIOD VESSELS (For explanation, see p. 596.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOLE. 2123, PL. 23 & SANTA ROSA-SWIFT CREEK PERIOD VESSELS (For explanation, see p. 506.) SANTA ROSA-SWIFT CREEK PERIOD ARTIFACTS (For explanation, see pp. 596-597.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOLES Pi 25 WEEDEN ISLAND PERIOD SHERD TYPES (For explanation, see p. 597.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 213; IPE 26 WEEDEN ISLAND PERIOD SHERD TYPES (For explanation, see p. 597.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOE els P27, WEEDEN ISLAND PERIOD SHERD TYPES (For explanation, see p. 597.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 113, PES 28 CAT. WEEDEN ISLAND PERIOD SHERD TYPES (For explanation, see p. 597.) SMITHSONIAN MISCELLANEOUS COLLECTIONS Wels sins, (Heo ky CAT. WEEDEN ISLAND PERIOD SHERD TYPES (For explanation, see p. 597.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 1S Pease CA. WEEDEN ISLAND PERIOD SHERD TYPES (For explanation, see p. 507.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOLES dis PES si CA. WEEDEN ISLAND PERIOD SHERD TYPES (For explanation, see p. 597.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOLO 13) Piase CAT. WEEDEN ISLAND PERIOD SHERD TYPES (For explanation, see p. 507.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 113, PL: 33 WEEDEN ISLAND PERIOD SHERD TYPES (For explanation, see p. 597.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOES 213 Reese WEEDEN ISLAND PERIOD SHERD TYPES (For explanation, see p. 597.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOR sapien 35 WEEDEN ISLAND PERIOD SHERD TYPES (For explanation, see p. 597.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOLT 113 Pease WEEDEN ISLAND PERIOD SHERD TYPES (For explanation, see p. 597.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOER LIST Ps 37 CA. WEEDEN ISLAND PERIOD SHERD TYPES (For explanation, see p. 597.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 173, PE2 38 WEEDEN ISLAND PERIOD SHERD TYPES (For explanation, see p. 508.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOLES vis RES s9 WEEDEN ISLAND PERIOD SHERD TYPES (For explanation, see p. 508.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 113, PL. 40 WEEDEN ISLAND PERIOD SHERD TYPES (For explanation, see p. 508.) SMITHSONIAN MISCELLANEOUS COLLECTIONS Wolke ahtey 1k 41 VESSELS OF VARIOUS PERIODS (For explanation, see p. 508.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 213; PER ‘Sasa A CA. ARTIFACTS OF VARIOUS PERIODS (For explanation, see p. 508.) SMITHSONIAN MISCELLANEOUS COLLECTIONS 4 PAYA AD 1, o set fA 9 uk x e hy fa Meee CAT. FORT WALTON PERIOD SHERD TYPES (For explanation, see p. 508.) VOLE. 2137 PE: 43 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOLS dais iP FORT WALTON PERIOD SHERD TYPES (For explanation, see p. 508.) 44 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 213; PL. 45 FORT WALTON PERIOD SHERD TYPES (For explanation, see p. 508.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL is PES ENGLEWOOD PERIOD AND ASSOCIATED VESSEL TYPES (For explanation, see p. 508.) 46 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOES MIS PE 47 CA. ENGLEWOOD PERIOD SHERD TYPES (For explanation, see p. 508.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 113, PL. 48 ENGLEWOOD PERIOD AND ASSOCIATED SHERD TYPES (For explanation, see p. 508.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 113, PL. 49 SAFETY HARBOR PERIOD SHERD TYPES (For explanation, see p. 508.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOE: 113, PES So CA. SAFETY HARBOR PERIOD AND ASSOCIATED SHERD TYPES (For explanation, see p. 508.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOLES las PIE SAFETY HARBOR PERIOD SHERD TYPES (For explanation, see p. 590.) 51 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 2137 (PES2 SAFETY HARBOR PERIOD SHERD TYPES (For explanation, see p. 590.) SMITHSONIAN MISCELLANEOUS COLLECTIONS WOLD saliisy Tei € SAFETY HARBOR PERIOD AND ASSOCIATED VESSEL TYPES (For explanation, see p. 590.) 53 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 223 SPE S4: é i CA. B, SAFETY HARBOR CHIPPED-STONE PROJECTILES AND BLADES (For explanation, see p. 590.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL 13, RES 55 ; 25/53 ‘ ig fe) s CAN. SAFETY HARBOR PERIOD ARTIFACTS (For explanation, see p. 590.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 113, PL. 56 SAFETY HARBOR PERIOD ARTIFACTS (For explanation, see p. 599.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOLES 103 PES 57, v SAFETY HARBOR PERIOD ARTIFACTS (For explanation, see p. 599.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 113, PIE 58 CA. SAFETY HARBOR PERIOD ARTIFACTS (For explanation, see p. 590.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 113, PL. 59 CA. SAFETY HARBOR PERIOD ARTIFACTS (For explanation, see p. 599.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 113, PL. 60 LEON-JEFFERSON PERIOD SHERDS AND ARTIFACTS (For explanation, see p. 5090.) ee hs ya tee Ue ee ek Pie . Were oe es m4 ae, aa i Aas fae Pee a - * Pr a : = ee aa _ , , a q ' 1 a 7y Wd Wit Sfo99 Pile TY Ged) One AR 7 ; : Y hn, i ‘ a vey ir. na i : ; if \ a Ay by ; 1 ‘ i Aaa | ay ity ‘ 1 i Pre | Y ee ie an h ‘ ( i ‘ i i 7 RAgy et J FO. be Yebap ui a iy iy