Sees a. — Awe i ar rn prema pebangete soseocce3: os iN Tee sae me eatery ters ee rense She. —— AON ‘ it) Overy iy hy Ta tas’ Sie iia Renan NAAN: x a Pi re a We ie HAN : i (NY if Nan en ae ort! 2 ‘ Y ASSN) SVL ORO EL an AGS See tie iy Ayn " 7 atAt fh t : ja i > eas Ne Eyantves a iy) i) mY y f i My t ; | \ an iaaieiaptbigbulapieiabaieaddgabieninnsctomtemmetiece NS esa ee See Se PERE ial AY i a RNY i vel His I > me ii ‘) ty Ae x BA Na Kee a io ore See ii ee ree ee ws Sa Aye H Tota i PAY f iy ‘ Aaty Wey ay aN Chen y tl BATE Mi sete uuet bah oas Oo \) : i Nya y ‘ ex ree i, ‘ \} wth uy Ht OM vad So Ve : i ; ‘ 4 », ‘\ : prem ecs oats aN at EOFS AZ are i ia Rn eee si Sears instant i , IK wean ee ees Senta tar fr = a oes a. = yy anh PN , a mae may ES ah See ; 2 = eit * Lr ‘>: ae (i RUNG M BENG ONNC i sini shes Xa i NUN BATA pail My Segsee setae vite ¢ Sate vee SS Eon he Sete eee << erent meee one ease OT eres Poe rae Ate es HN uy ‘ i) ‘ ithe BH i ee 7 : ¢ ROR UaK TAS ON Ns : a Ne Oe: Med ‘ i ; NE ity i i “); Riis w) N Og ‘ > x ‘ yh) 4 + G ei OL s Ea \ Lis ’ . 1fn 4 t La ARR ' a TAVERDNT ERY ft) Cn * Se Mave, ATED DIN OO EL ~ ‘ y de uv V7} ‘i hy { ry Sy wes es Sos . vant - = aed See ass Bh fo ¥ fs Serene Ty ace ‘ OAT t : a ’ aR , Ae i shh Went <0 i % *y ent ax in Say 389 na ay te" WI GASSSYSY aw VVSSEEE Ay Ww PILLS, ss) ewe Suwwowwuwwwwr™ ee | | Ay YIUSeS SYYUUe See oe Soe Wyis wie vvnvvies erect ee eee es UI ASSIS SMS Say vv YYSSA GAYS wy : vv NW we y) oeweeeenee J JOSE UN DION uve’ Zuciy eS} cd AM | JIGS | SISSY ey BUSS) Wiilviievere heal lale/ly Vole etremmueete yy = = SS WAI Ygy CI VY yeyveveey vey anion ONAN Ae dvd VV vile Wve) DIGS ated |] oo v ibe ig -_ lhc, lL . —lCUr EO re lCe Dl i el oe, lll lO 0G el | NS \ had Vesreeeerey FE] Y Nd ed td ed ded HNN Soy, y SoS sre aavavevavel ee eae cane nn hy AAA j) hd ee i had Wed Neg Ne Lk KC@ Ce ¢ 7 ; fy apt : wh et A ney , | oo ; | ia 4% SMITHSONIAN MISCELLANEOUS COLLECTIONS (ox se sa \\ i xo igh wad “A & ae / ent INGTO “EVERY MAN IS A VALUABLE MEMBER OF SOCIETY WHO, BY HIS OBSERVATIONS, RESEARCHES, AND EXPERIMENTS, PROCURES KNOWLEDGE FOR MEN ””__sSMITHSON (PUBLICATION 3450) CITY OF WASHINGTON PUBLISHED BY THE SMITHSONIAN INSTITUTION 1937 The Lord Baltimore Press BALTIMORE, MD., U. S. A. ADVERTISEMENT The Smithsonian Miscellaneous Collections series contains, since the suspension in 1916 of the Smithsonian Contributions to Knowledge, all the publications of the Institution except the Annual Report, the annual pamphlet describing the Institution’s field-work, and occasional publications of a special nature. As the name of the series implies, its scope is not limited, and the volumes thus far issued relate to nearly every branch of science. Papers in the fields of biology, geology, anthropology, and astrophysics have predominated. GG. -ABeor, Secretary of the Smithsonian Institution. (iii) CONTENTS . Cotittins, Henry B., Jr. Archeology of St. Lawrence Island, Alaska. 431 pp., 84 pls., 26 figs., Aug. 9, 1937. (Publ. 3411.) . Fuint, Lewis H., and McAtitsTeEr, E. D. Wave lengths of radia- tion in the visible spectrum promoting the germination of light- sensitive lettuce seed. 8 pp., 1 pl., 2 figs., June 16, 1937. (Publ. 3414.) . Jounston, Eart S. Phototropic response and CO, assimilation of plants in polarized light. 7 pp., 1 fig., July 12, 1937. (Publ. 3440. ) . BusHNELL, Davin I., Jr. Indian sites below the falls of the Rappa- hannock, Virginia. 65 pp., 21 pls., 11 figs., Sept. 15, 1937. (Publ. 3441.) Snopcrass, R. E. The male genitalia of orthopteroid insects. Foy pp, 42 hgs., sept. 25,1937. (Publ. 3442.) . Jounston, Eart S. Growth of Avena coleoptile and first inter- node in different wave-length bands of the visible spectrum. 19 pp., 4 figs., Nov. 6, 1937. (Publ. 3444.) (v) Baa ae uh wt - ' 5 r r oe at a arenes -aecimenen eames aemeennmans aan eS ‘7 pixe A te \ GuAIuUaaRO rade ra th s inden pi f 4 Wt, sa ae pha ee ee GABMAAAS 2:08 - . "ate 2 {> aor 1M \ . . ~ . ey Kot ~ . a . i \ "On yi * ~*~ } : ~ a OF é [ a ior - - ay ji y asi > ¥ } / Ms ’ by : ~~, L : \ } i ‘ > ? 4 ‘ ¥ 3 rs = a J Sn ~ Yom Px a A’ / ae i ; ] a , ~ ws Pid ’ : d } J 7 ‘ ; \ \ J co p. fy ee ar. a ‘ Pa ™ ‘an = \ = 4 : a : . _—— y Ps . < fo SE \ j { f \ , r x \ : i a24gRwe ¥ . P \. \ * © i n ‘ — +. o™ 3 3 # m , } ‘ i : 4 ue “ : ‘ Q 2 a - “> ~ 2 > ‘" tet > : * ‘a eX ¢ e » . ta 8 o ‘ : ~ ' “ "A J bs a © = — : : . VJ cn en ee agement een nee me eR NN — one ‘ 18) 4) x or ote se ealz0d srl! fo asl 100° 40° 80° 60° 4 ? >! D: i 3} 110° iG: 0! Pp: oO: he — vy 4 : 120° 2 120° 60° 50° SMITHSONIAN MISCELLANEOUS COLLECTIONS VOLUME 96, NUMBER 1 ARCHEOLOGY OF ST. LAWRENCE ISLAND, ALASKA (WitH 84 PLATES) BY HENRY B. COLLINS, JR. Division of Ethnology, U. S. National Museum (PUBLICATION 3411) CITY OF WASHINGTON PUBLISHED BY THE SMITHSONIAN INSTITUTION AUGUST 9, 1937 The Lord Baltimore Press BALTIMORE, MD., U. & Ay PREFACE The present paper is a report of excavations conducted for the Smithsonian Institution at old Eskimo sites at Gambell, St. Lawrence Island, Alaska, during the summers of 1930 and 1931. The 1930 investigations were conducted by the writer, assisted by James A. Ford; those of 1931 were carried on by Moreau B. Chambers. The results of the writer’s previous investigations on St. Lawrence Island in 1928 and 1929 are also utilized to a limited extent. The first section of the paper describes the excavations and the materials found; the second, or comparative section, treats of the distribution of the ele- ments and the archeological problems involved. This paper, in essentially its present form, was completed in the fall of 1935, and under the title “Archeological Investigations on St. Lawrence Island, Alaska, and their bearing on the problem of the origin of Eskimo Culture” was submitted to the Royal Academy of Sciences and Letters of Denmark in response to a prize competition announced in March 1934 on the subject “ Quelles sont les origines de la plus ancienne civilisation des Esquimaux?” In February 1936 it was awarded the gold medal of the Academy, the judgment of the committee appearing in the official organ of the Academy, Oversigt over det Kongelige Danske Videnskabernes Selskabs Virksomhed, Juni 1935-Maj 1936, pages 160-164, and in French on pages 320-322. Since then the paper has been expanded somewhat, a number of illus- trations have been added, and numerous changes, principally of a minor character, have been made in the text. I wish to record here my obligation to James A. Ford, to whom in no small measure is due the successful outcome of the 1930 investiga- tions; to Moreau B. Chambers who carried on so well the work that we had begun; and to our loyal and intelligent Eskimo helpers, Paul Silook, Philip Maskin, and Moses Soonogoruk. I wish also to express my gratitude to Capt. E. D. Jones, Commander of the U. S. Coast Guard Cutter Northland, for transportation to and from St. Lawrence Island both years ; and finally to Dr. Ales’ Hrdli¢ka who was responsi- ble for first directing my interests to the Alaskan field. The arduous task of identifying the thousands of bird bones from the excavations has been performed by Dr. Herbert Friedmann. The mammal bones have been identified by Dr. Remington Kellogg; the ili iv PREFACE fish bones by E. D. Reid; mollusks by Harald A. Rehder ; and crusta- ceans by Dr. Waldo L. Schmitt and Dr. H. A. Pilsbry. The mineral identifications were made by E. P. Henderson. The drawings, maps and. charts are the work of E. G. Cassedy, Frank Imrey, and J. W. Wilson. Mr. Cassedy has made the frontis- piece, and text figures I-3, 7, 10-13, 18-22, 24-26; Mr. Imrey has made plates 14, 16, 18, and text figures 5, 8, 14; and Mr. Wilson text figures 4, 6, 9, 15-17, 23. The photographs of specimens were made in the photographic laboratory of the United States National Museum by G. I. Hightower. CONTENTS PAGE ERELACE menerarereleteeelove fers) srorckaverers SEE eR a Ale the ee arm ncn iota a ETO ROT OPE ili HETESCOCITIGC Ul OlinMret teen ereveictevevsie-axesevevorersKerenctalueatvesceaiiatoreen chereteraate chal ates cesta cuciotcieionaies I HE CORStroUT II Calis SCEEIIT EY rev ayayes ovey cnalonsvavet es once erent atiaeee rateen eetotane oimie custeneete le apclote’ « 13 IStOnvaands prevdOUSMEXPlOLAatiONG iin. ctus thiclstoto took aes arate ols -elelevoredelehele etehetsye « 15 Abandoned sites, and the excavations of 1928 and 1929...............-00: 26 Bexcavations at Gambelll, CSevtaolkolc)| (s,s (his acibie ave bee sete cletietes ole er aelereie ete 31 IAT Get SiWet arena SO aC it teen er ES ea eee I A rir en Oe een rg 36 Wecoratedsobyectsss Old) Berime: Sear. cee neice acer ane 40 Decorated objects from Little Diomede Island.................... 53 MP Urisas atch cEd MNES ome ay See PON Po Zs ch pee wash caster eESL Ge) «Lasel fe: cusednsas. Steelers pelea Ate DRE 56 Gitssin the morth western SCtHLOM'=.2:.4 sc cet ar. ctehevard.e dictutasdee opacdleue lores 58 Crtseinuthessoutheastentmsectiomers criierieteis ole rcs-1a ose. creuceieierercievereevale 61 IBIGBRES. Sab cdo ce boned 6.0 DUO EEIG TOR cae ae tion tice aeons conta occ 69 Wecoratedcapjects, Old) Bering Seas oar ce.aie:vidieia vie hele OD anielge elas 7 WD ECOratediOD}ECES easly; satIMtt sereys ctetsvercieieieterciexcl-serei eltekelstelielererepelel ele g2 fanpoonpneads,, Old Bering | Searvsists.. scievscas.0 siereses ieraeted Abele eee 07 Elarpoonmeadsy ear lyase ctoyeiescveyevcke ai ieterseteueeue miebeletetertbe D eleye' siete 115 iniplenienthtypesw Olde BerinawSeasacictic.-seciscemiinlcteaie cleans oe 12 JRE Vaal Neha lee Aber erocite ae cate croc eine hed clon eonciots. caper: 12 Bicdedartsteind visi cSMEAt si tania wiisseco dcvaecistel cereus tee aictaelo bare 131 PIR DEOMT INS DOALGS Mts. 5 fay cioresattbeyseateiaicber aucd.d:0 o,.a,crchchteeene viele ofeereee ele 177, BISONS tama ea cto ci SE Ve isl wis Totes Rio: Sone % aseiarate dgeloranee events Mptetereisiortorate 178 pimiplement types ene Meh UNUKe. 5)<:5\0'sic's aca <'ew aoaivynetaa rete ae ae 179 vi CONTENTS PAGE BEVOpIIVIGG maleic reise o/c iaietelayel-< etereiat= fom a) 0) 2= ape iaiensier sicuavs, syerctsiesnesegs ssoisistonte 181 1S 1b oS aR TDI IRI IOE ROID OM OMe BIbIG Gioo.ooKKaGs 183 Seklowaghyaget ccc. cece ebce acc + cee mminteln oilers ochre Sehr 186 PTO USO: Sizes ere sec chcvate SoS lg leuoh gb ei havreneie av dhe allel ICLP Ta Oe tee OR eRe een 187 Old ‘section: of ‘Gambelll 2.2.58. 2. i oie ee 209 Harpoon heads from old section of Gambell and houses 8, 9, and 10.. 212 Whaling: harpoon) Iead'sys -eyre terete ts)-to =) eletete deleted el etebatey ret dl ene\lehoh Naeem 215 Implement. types of the Punuk period. ..22 cs potey> 2 cl-tetenatel-lare eich eee 217 Skeletal) sremaims?) do. jc. cicsalasdeele Gaaooco aueloucie ecco ttoreacke Orem ROC ae eee 246 Animal remains, trom: the) tive: sites emir eter erento renee eee 247 FRESUINE sis ciaiess wns dlegietaetetere Si oka vee rteee Sichs lonstode eae noe ORC era 251 Comparative analysis. cio... sce 1 je avocssaleyoieceiove ole seexcionenerseueei\ey-8- 14 256 FLOUSES © coisssce oie ease vous dee teve teteusse is be be RST Re Pertti Oe Se eens 256 LC, a er Maar rE RE non oodhodechdutc UM oo 000s 286 EONS TAVIM COOLS, o:ccoaris iver fore eie-eie or cieds) oie okt beNe hones eke nesees keto ere eee 303 Ornaments: And. LOYS.i ciecvecaue casuals oislene Cleneraeerel ed beh nore eer th keener 305 Hlaripoor. Weads. cic ieve ect aseinjanere avis ney otthciete stavelene ane intel Te lated atte tee ee 306 FOr Sa bts ne jeiosgus wcove.dicceauredive re Sudeels ioteneue CRAM ONETCNE oI Oe OOO een 321 Slee) dat aah Lit oh en en AA eh anche on hoe b.cadoo0lCoNKC 321 PSM SET LESES .- vs ssiteea tee eras, eigen cotab neue chek Rot motes OMetoeT: ER RGR ICR Pte eee ea 321 TGO. OVC KS. on si soe ore inva cavatnve fovote varesnve si crane spivs vololeroisks ofeliote¥alelehotohol kote te vi ts iaaaame 321 TG]: CEEEPELS. cosa seve acs le 9 ledeccote insane otsketeke noe Fane a ae ee eee 322 Sealing: iscratchens. i torecciersmssr- -voucns shore teretenefe nee e tated eke eee ea ee 322 Throwing boards and bindsdartsececeeuier tir eieei eee 322 BOws' and!) ALLOWS \eicccveteccctees cts esore ere erolenereueieresenehtehey Nit Penne aon en eee 323 Biint=poimtede bird eaniOwsreeeie eee eee aii a ehels Poe 324 Wirist staan Se ..cccx6 le -snciare sore. ave evantcovs sate yenenouees chetenenene no Reka NOR foment RN Ree a eee 325 leit cdl )\c) it rE ERNE Ohno a aonocoos Monts doobocAcCoNS 325 Plate ABIMIOR: 6 sa. ce save cove esis sara rersce kayroye ents wile Vere ROO MOWoue Rodel SOI IC ROBE Ree RNC eee 325 Mle n’Sic LAI ES. seisievovereie-eccvecerenuce so 1000 irre lana fa raverwsens seeronokotetehekotenedehet ete Meteeteketenetels 333 INGZ MEAS) seicsnese.ce wre vevevene sonerdva vo 15s ro son Selene tee pers Net HeM OT TORR CNM CRT ae te ene 333 Sfome implements: .....c so levesdussewe eves aucrenelele ote Feretede Be ePeNeT arene’ WR NORE Reena ‘gee Sledges: and; tobog ais aj. .5:cscvcre-cranavcuole topenekesolopebetohene: oi hetenel ey lettre Acer eeenenete 338 amps and! (COOkIM GE POtSe ereicu-irescteusuetelenenenel en oNeReke aN Ne Nate ee ole Netet= t-te eee 341 Balleem, VeSSELS.. aiajeisc.aveusséecuscovesayonchenciereuereuemete Neue onewen on obere etolen toe Re one ene 350 Wroimeti’ se Tantves «a. s.creiescscetecenctereiaicenleroun eles oer tele eto hoor eters . 350 Eat \SCAPCTS. yéic.cie-bienaeexe esatareia oie evens asvonads ronepekersee olen potekea for ate kettle tea 351 Needle 1 CASOS: or .u sce seieseyensicons. a saxe rouse toitenc loca ne taseneleye korenet ete Keaefete rs CRT CTeR TRC enemies 351 JUMPING StONES. 566.5.6.c10s. sossis senators vore baverensialeetoroaates ROE EORTC tee wet 354: SUMMA Y. 6 e.e:.ciers so.c ih vers ail loka fa roucnuaveyore els veto low foe OOO aCe Oat 350 COMCIUSTON ..ovorerevarecaie iavevdte.iovaidenduejena worareaekevevensua tone sefodonelted tet ataneve nek ane Ren ne ane 361 18510) hoye5¢2\1\ he nee MPAA ES Fh BiG inidio go clonsanangaccocc 5 ate Beeplanation:of Plates, Zo ~~_ PD ee me ee Fre. to.—Outline of Miyowagh, showing excavations of 1930 and 1931. to the Eskimos as Miyowagh, the “ climbing-up place’, so called be- cause the trail from Gambell here begins to ascend the hillside. The midden lies at the head of the gravel plain within the first two—the oldest—beach lines which extend westward from the base of the pla- teau for a distance of three-quarters of a mile. It is situated about 100 yards from the north end of the lake and three-quarters of a mile and a half a mile back from the sea (text fig. 2). The midden is irregular in outline with an average diameter of a little over 100 yards (text fig. 10). It appears to be rather low, with a visible height NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 57 of less than 5 feet. However, this is due to the encroachment of the gravel, for the bottom of the midden was found to extend several feet below the level of the surrounding plain, at one place reaching a total depth of 8 feet. In spring and early summer the small stream which flows down among the rocks of the Hillside site passes along the southeastern edge of the Miyowagh midden and empties into the lake. House pits are represented by shallow depressions, more or less cir- cular in outline. Some 18 or 20 of these pits were visible, but the surface of the midden had been so disturbed by the Eskimos digging for ivory that it was not always possible to recognize a house pit with certainty. The surface of the midden was practically free of whale bones and timbers, such as are found at some of the more recently abandoned sites. No remains of roof or wall structures were visible, most of them no doubt having been utilized in the construc- tion of later houses. The Miyowagh midden, as well as the three other middens situated on the gravel plain, differed in several respects from that at the Hill- side site. Being on the level ground, it had not been subjected to erosion or to disturbances from the shifting or sliding of rocks as had the Hillside midden. As a result the Miyowagh midden represents a full record (although one which may have been altered here and there through sloughing or from the excavation of house pits), whereas at the Hillside site the present configuration of the surface indicates that an unknown amount of refuse has been washed down the slope. There was also a difference in the nature of the deposits: at the Hillside site, as in kitchen middens generally, the débris of human origin was mixed with a certain amount of natural soil, for the gently sloping lower part of the hillside is covered with a fairly deep soil layer. At Miyowagh, on the other hand, the first houses were built on a barren gravel plain, and the extensive pile of débris which slowly accumulated around, and eventually over them, was al- most entirely artificial, the greater part of it resulting from the decomposition of animal remains and other organic materials. The midden therefore consists of a heavy, rich, black matrix containing great quantities of animal bones and baleen in addition to artifacts and rejectage of stone, bone, ivory, wood, and baleen. The midden was found to be compact and, in general, unstratified. Occasionally a thin layer of mussel shells or a mass of bones, etc., would stand out distinctly, and along the western periphery lenses of midden material were sometimes separated by masses of gravel. As a rule, however, there was no stratification in the usual sense, the bones, artifacts, and rejectage being a relatively homogeneous mass, held together in a 58 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 solid matrix of permanently frozen, rich, black soil. This is a con- dition that would naturally result when a kitchen midden was formed as this one was—by the daily accumulation of rubbish thrown just outside the houses. In appearance such a midden differs markedly from the more familiar type found all along the coasts as far north as the Aleutian Islands. The principal point of difference is that these more typical middens to the southward contain more or less distinct strata which are due in large part to the presence of shells, ashes, sand, and other beach material, none of which is present to any appreciable extent in an Eskimo midden of the Miyowagh type. The first cuts at Miyowagh were made in the northwestern section of the midden which was slightly higher and had also suffered less from the Eskimos’ digging than had the southeastern section. Fur- thermore, it was relatively dry and well drained, whereas much of the southeastern part of the midden was in a soggy condition and so remained until midsummer when the snow from the hillside, which fed the small stream at the eastern edge of the midden, had almost disappeared. As stated above, the cuts were 12 feet square and were taken down in layers of from 2 to 6 inches in thickness, except for thicker surface layers of sod or sterile layers of gravel. In the follow- ing pages descriptions of individual cuts will be reduced to a mini- mum. Many of the cuts yielded between 200 and 400 specimens, in addition to quantities of animal bones, potsherds, broken slate blades, and other fragmentary artifacts. In the brief descriptions of indi- vidual cuts only those artifacts having definite chronological signifi- cance will be mentioned. Later, when describing the different types of objects, significant distributions according to cut and depth will be noted. Cuts 1-4, 6, and 17 (season of 1930) were sunk in the northwestern section of the midden, and cuts 5, 7-9, 9a, gb, 10, 10a, 11-16 in the southeastern section (text fig. 10). In 1931, Mr. Cham- bers continued the work on cut 17, and cuts 19, 20, 22, and 27 were sunk in the northwestern section of the midden, and cuts 18, 21, 23-26, 28, and 29 in the southeastern section. Three houses were excavated at Miyowagh in 1930, no. 5 in the northwestern section and nos. 3 and 4 in the southeastern. CUTS IN THE NORTHWESTERN SECTION Cut 1 (pl. 6, fig. 1, the excavation in foreground).—Cut 1 was 5 feet 3 inches deep; taken down in 10 levels. Surface covered with scattered grass, no sod. The upper 18 inches consisted of almost pure gravel, with sporadic patches of midden material; 8 inches deep was NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 59 a layer of mussel shells 1 to 2. inches thick. A large piece of walrus hide was found at a depth of 30 inches. Small pieces of wood began to appear at this depth, becoming more numerous in the deeper levels. Artifacts found numbered 198.° No examples of Old Bering Sea art or harpoon heads were found in cut 1. Two of the harpoon heads, from depths of 30 and 53 inches, belonged to open socket type III x; one of these is shown on plate 24, figure 22. The other open socket heads belong to the later types: type II (a) x, II (a) y (pl. 28, feee)el (e) (pl: 28, figs 4); IIT (a) x (pl28, fig. 13). The:closed socket heads were also of the small, simple forms, such as type V x (pl. 28, fig. 22) and V y (pl. 28, fig. 26). Nine of the harpoon heads, and two other objects bore Punuk ornamentation. Cut 2.—Ninety-two feet NE. of cut 1. This cut was 4 feet 2 inches deep; taken down in 10 levels. From the surface to 20 inches deep was almost pure gravel, with occasional artifacts; patches of midden material appeared at depths of from 20 to 24 inches. Stones and whalebones, apparently remains of an old cache, were encountered at 34 inches, and these continued to the base. The end of a large baleen toboggan was found at depth of 43 inches (pl. 6, fig. 3), the greater part of it extending into the south side of the cut, which had to be enlarged for its removal. Two pieces of wood at 28 inches; more in the deeper sections. In this cut were found 85 artifacts, in- cluding the toy trident (pl. 22, fig. 6) from depth of 20 inches. Har- poon heads very similar to those from cut 1; three are figured in plate 28, figures 7, 8, 25. Four were decorated in Punuk style; only one example of Old Bering Sea decoration from this cut, a fragment of a harpoon head from depth of 43 inches, which had been made over into a small knife handle. Cut 3.—Forty-two feet SW. of cut 1; 5 feet deep; taken down in 11 levels. The upper 16 inches was sarteh solid gravel, and con- siderable quantities of gravel continued to a depth of 3 feet. Arti- facts numbered 238. The ivory from the last three levels (50 to 60 inches deep) was in general noticeably lighter in color than that above. Two harpoon heads decorated in Punuk style, from depths of 20 inches (pl. 24, fig. 19) and 24 inches; two fragments bearing Old Bering Sea decoration, from 54 inches (pl. 15, fig. 5) and 60 inches deep. *The number of artifacts listed for this and all the other cuts includes only complete specimens or recognizable fragments; in addition to these there are in every case large numbers of potsherds, fragmentary stone blades, pieces of worked ivory, bone, wood, baleen, etc., which were collected but which are not included in the total count of specimens. 60 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 Cut 4—Adjoins cut 1 on the east; 5 feet 8 inches deep; taken down in 11 levels. Specimens numbered 195. Very similar to cut 1 as to gravel content, etc. Plate 6, figure 1, shows the beginning of work on cut 4, with cut I in the foreground; plate 6, figure 2, is a view of the same at a later stage of excavation. A piece of textile found at depth of 26 inches. The decorated needle case shown in plate 17, figure 9, came from a depth of 55 inches. Harpoon heads were were all Punuk types; four of these are illustrated (pl. 24, fie. 8; pl. 28, figs. 12, 14,05). Cut 6—Thirty-five feet south of cut 3; 4 feet 7 inches deep; taken down in eight levels. Specimens numbered 98. At a depth of 20 inches was found a mass of organic material, evidently disintegrated animal flesh, flaky, yellowish, blubber-soaked, together with fallen pieces of wood which had probably formed part of a cache. More of the cache—fallen whale ribs, scapulae, and stones—were found below this to a depth of 50 inches. There was relatively little gravel in cut 6, but more baleen and other organic material which emitted a sour, pungent odor. The ivory winged object with elaborate Old Bering Sea decoration (pl. 20, fig. 4) was found at a depth of 36 inches. The bone harpoon head (pl. 28, fig. 6) comes from a depth of 20 inches; another (pl. 26, fig. 6) from a depth of 26 inches; a fragment, from the latter depth, bears Old Bering Sea ornamentation. On the whole, the midden deposit at this particular spot appears to be older than that to the northward. At a depth of 50 inchés was found the skull of a bearded seal, with a baleen thong through the foramen in the palate; it had evidently been preserved prior to being disposed of in some special manner. Cut 17.—Seventy-five feet W.-SW. of cut 1; 40 feet from edge of midden; three levels, to a depth of 26 inches, taken down in 1930; three additional levels, to depth of 34 inches, in 1931; bottom not reached. Specimens numbered 153. The part of the midden where this cut and cut 20 were made seems to have been the latest section of Miyowagh. One of the harpoon heads, from a depth of 20 inches, is shown in plate 28, figure 3; the others were mostly of open socket type III (a) x. No trace of Old Bering Sea art or artifacts, or even of early Punuk. Cut 20.—Between cut I and house no. 5 Excavated by M. B. Chambers in 1931. Taken down in five levels to a depth of 3 feet 3 inches, bottom not reached. Artifacts numbered 115. Material prac- tically identical to that from cut 17. Fifteen harpoon heads and two other objects bearing Punuk decoration, one weathered fragment of harpoon head with Old Bering Sea decoration. Four of the harpoon NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 61 heads from cut 20 are shown in plate 28, figures 5, 19, 24, and 28, all from the surface layer. The decorated drill shown in plate 22, figure 8, is from the same level. Cut 19.—Excavated by M. B. Chambers in 1931. This cut, 18 feet long by 12 feet wide, was made at about the center of the midden, between cut 4 and the end of the entrance to house no. 3. It was the most productive of all the cuts, yielding 514 specimens besides the usual potsherds, baleen, fragments, etc. It was 7 feet 2 inches deep and was taken down in 22 levels. The midden here was very trashy, containing a large quantity of baleen, bones, and wood, which sloped downward from east to west. Plate 6, figure 4, shows the north side of the cut with baleen and other refuse projecting from the wall and some of the timbers of a razed structure lying at the base; plate 6, figure 5, is a view of the same cut showing the timbers at a later stage of the excavation. In plate 7, figures 1 and 2, the north and west walls of cut 19 are visible in the background (cut 27 in foreground) ; the strings along the north wall (pl. 7, fig. 2) indicate the slope of the refuse. The slope of the midden at this point may account for the fact that harpoon heads of Punuk type occurred at a somewhat deeper level here than elsewhere, for three were found at depths of 51 and 59 inches (pl. 26, fig. 8; pl. 28, fig. 27), in addition to a large number of others at lesser depths. In all 52 harpoon heads and fragments were found in cut 19; 29 of these, from the surface to 59 inches deep, bore Punuk ornamentation, mostly Punuk style 1, while only one fragment, from a depth of 78 inches, bore Old Bering Sea ornamenta- tion. However, there were several other objects decorated in the latter style, from the following depths: 5 inches (pl. 15, fig. 10) ; Ey iuches (pl. 05, fig. 9)'; 24 mehes; 51 inches’ (pl. 10, fig. 1) ; 56 inches ; 63 inches (pl. 15, fig. 6) ; 76 inches ; 79 inches (pl. 15, fig. 1). Cut 22.—Between cuts 2 and 27. Excavated by M. B. Chambers in 1931. Taken to depth of 3 feet 10 inches, in six levels; bottom not reached. Specimens numbered 91. ; CUTS IN SOUTHEASTERN SECTION _ Cut 5—Near SE. end of midden ; taken down in six levels to depth of 3 feet 10 inches, bottom not reached. Specimens numbered 68. At depth of 28 inches a mass of skin, vegetable matter (plants for eat- ing?), wood, and baleen was encountered, along with fallen whale ribs, jaws, and scapulae which had evidently formed the walls of a cache. The baleen consisted of 14 small-sized slabs, from a young whale. On removal, the vegetable matter, baleen, etc., was found to 62 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 rest on four thicknesses of split walrus hides, lying at a depth of 4I inches. Pieces of wood and bones, from the cache, continued to this depth and beyond. Cut 7.—Five feet north of entrance to house no. 3; 8 feet deep, taken down in 14 levels. Specimens numbered 235. This cut affords a good example of the extent to which patination may be affected by local conditions. The ivory from the first (upper) four levels was very dark, while beginning with layer 5 (37 in.) it became noticeably lighter in color, that from the last five layers (75 to 96 in.) being the lightest of all. This condition appears to be the result of water action. The midden here is low and in early summer is more or less saturated by water that flows down the hillside. This has discolored the ivory in the upper part of the midden, while that in the permanently frozen soil below has escaped saturation and remained lighter in color. Timbers and whale bones of a razed structure of some sort were found beginning at the 6-foot and extending to the 7-foot level. Two examples of Punuk art from this cut, a harpoon head with a light line and spur decoration, from a depth of 20 inches, and another, 62 inches deep, with the earlier simple line decoration (pl. 24, fig. 20). The other decorated objects were all in Old Bering Sea style; they were found at the following depths: 67 inches (2) (pl. 19, fig. 23 pl. 24, fig. 16); 75 inches (pl. 24, fig..1) ; 92 inchesn@piya fig. 11); 96 inches (pl. 24, fig. 13). The unfinished winged object shown in plate 20, figure 2, came from a depth of 80 inches ; another broken specimen of the same type from a depth of 51 inches. The needle case (pl. 17, fig. 5) came from a depth of 33 inches and the small ivory comb (pl. 58, fig. 10) from 67 inches. Cut 5—At east end of midden; 30 feet NE. of house no. 3; 6 feet deep, taken down in seven levels. Specimens numbered 68. This cut was in the lowest part of the midden, which in early summer is coy- ered with water. The bone and ivory, particularly the latter, is dark colored and poorly preserved, even that from the bottom. At the SW. corner of the cut was found part of a filled-in pit which ap- peared to have been oval and about 6 feet in diameter. The inner side of the pit was of dark, almost black, soil, very firmly packed and soaked with blubber. It was filled with softer and lighter colored soil. than the rest of the cut. Bordering the pit and extending NW. by SE. was a row of fallen whale bones, small timbers, and one walrus skull, apparently part of a razed wall. Similar material was found in the eastern part of the cut, beginning at a depth of about 2 feet, and in order to follow this out and determine if possible the nature of the structure it represented, four other cuts were begun (cuts 10, II, 12, NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 63 13) adjoining cut 8. Three decorated objects, all Old Bering Sea, were found in cut 8: the winged object (pl. 20, fig. 3) from a depth of 23 inches; fragmentary harpoon head, 32 inches ; another harpoon head (pl. 26, fig. 19), 40 inches. Cuts 10, II, 12, 13—Cuts 10 and II were two small cuts, each 6 by 4 feet, adjoining cut 8 on the SW. Cut 10, 45 feet deep, was taken down in seven levels; cut 11, 5 feet deep, in six levels. Cut 12, 6 by 6 feet, adjoining cut 8 on the north, was taken to a depth of only 27 inches; bottom not reached. Cut 13, adjoining cut 8 on the east, was triangular in shape, 12 by 10 by 6 feet; 5 feet deep, taken down in five levels. ‘Specimens to the number of 162 were found in these four cuts. The material was similar to that from cut 8; the ivory was dark and rather poorly preserved and there was a considerable quantity of baleen. Unfortunately, we were not able to get any clear idea as to the nature of the structure or structures revealed in these cuts. The pieces of whale bones and timbers found in the eastern part of cut 8 were shown to have been part of a cache, for they continued into cut 13, where they were found to enclose two layers of split walrus hides. Below this were found other timbers and whale bones, evidently part of a house, for they rested on the remnants of a stone floor found at a depth of 54 inches. On the floor were found a fragment of a harpoon head and a gut scraper (pl. 17, fig. 2), both with Old Bering Sea decoration, and a wooden object with one end carved to represent a human head (pl. 47, fig. 15). Just outside the stone floor was an upright timber 8 inches in diameter, broken off at the height of 24 feet. This seems to have been a corner post, for extending at right angles from it were remnants of two small horizontal timbers. Three other broken uprights were found 6 and 9g feet south of the first one, and undoubtedly the other timbers and pieces of whale jaws lying on and above the stone flooring had been part of the walls or roof of the house, but it had been so thoroughly demolished that no structural details could be made out. Six inches below the stone flooring was what appeared to have been an earlier floor of the same kind. On this lower floor was found a small harpoon head with Old Bering Sea decoration (pl. 26, fig. 13). Cut 9.—Thirty feet south of cut 8, on the low-lying eastern edge of the midden, was a large shallow depression, the pit of an old house. To uncover this house an excavation, cut 9, was begun. The sod here was thicker than at any other part of the midden, and below it barren gravel continued to a depth of 2 feet. The stone floor of the house was found at a depth of 4 feet (pl. 8, fig. 3). Specimens found numbered 294, and the material was segregated according to 64 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 three levels—i1 and 2 above the floor, and level 3 on the floor. The dimensions of the house (no. 3) were 21 feet 3 inches by 18 feet 6 inches, but our original excavation was larger, 26 by 30 feet. This larger excavation took in 5 feet additional space to the SW. and NE. and 114 feet to the NW. and SE. Only Punuk material was found on and directly above the house floor (some of the objects will be mentioned when describing the house). On the other hand, the deco- rated objects from outside the house, in the space included in the NE. and SE. extension of cut 9, were all Old Bering Sea. Included among these objects were the two needle cases shown on plate 17, figures 6, 7, and the two harpoon heads (pl. 24, figs. 5 and 14). Cut 9 a.—This area of approximately 10 by 12 feet, back of (SE. of) house no. 3, was at first included in cut 9, but was designated cut 9 aafter the walls of house no. 3 began to appear. Plate 8, figure 1, is a view of this cut (in background) at an early stage of the excava- tion, before either of the two houses, nos. 3 and 4, had been com- pletely uncovered. Plate 8, figure 2, is a later view of the same, show- ing James A. Ford excavating in cut ga (at the right) and Silook working on the floor of house no. 3; house no. 4 appears.in the right foreground. Cut 9 a seemed to have been the site of a razed structure of some sort, probably a house, for a small section of stone flooring was found at a depth of 5 feet, above which were loose stones, whale bones, and numerous pieces of logs. All was in such disorder, how- ever, that we were unable to form any idea as to what the structure may have been like. The cut was fairly rich in artifacts, there being 69 in addition to considerable quantities of potsherds, stone blades, baleen, and wood; the decorated objects, 10 in number, were all Old Bering Sea. Of greatest interest, however, was a burial, the only clearly intentional burial which we found in the Miyowagh midden. It lay 8 feet back of the rear (SE.) wall of house no. 3 at a depth of 4 feet. It was oriented toward the N.-NW. and was covered with several layers of matted skins which seemed to have been two parkas, one of bird skins and one of seal skins, probably P. hispida accord- ing to Dr. Remington Kellogg. The state of preservation of the body was remarkable; the permanently frozen ground had preserved the tissues in much the same way as dry sand may preserve burials in arid regions. The greater part of the skin, muscles, and tendons re- mained intact, appearing as a brown, parchmentlike covering, brittle and flaky, but maintaining its original relationship to the bones. The fleshy parts of the thighs and pelvic region, though much flattened, were especially well preserved. The body was flexed, with knees ex- tended to the left at the level of the hips and arms placed over chest. WO: 1 ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 65 The skull, badly crushed, rested on a large flat stone—not one of the floor stones mentioned above; there were several more stones and a mass of timbers overlying the body, probably having been so placed at the time of burial. The following decorated objects—all Old Bering Sea—were found above the burial at the following depths: decorated box handle, 12 inches deep (text fig. 14); two harpoon heads, 18 inches deep and one fragment, 20 inches (pl. 19, fig. 7) ; two needle cases, 20 and 24 inches deep (pl. 17, figs. 8 and 4) ; harpoon socket piece, 36 inches. Just beneath the stone on which lay the skull was found a much weathered harpoon head with traces of Old Bering Sea decoration. Other objects with similar ornamentation were found nearby but not in contact with the burial: an ivory gut scraper, 50 inches deep (pl. 17, fig. 1); an ivory winged object, 38 inches deep (pl. 20, fig. 5); and four harpoon heads, 36 inches (pl. 24, fig. 18, and pl. 26, fig. 1), 38 inches (pl. 26, fig. 18), and 48 inches (pl. 24, fig. 15). Cut 9 b.—This was a small, rather unproductive cut adjoining cut 9 on the SW. (pl. 9, fig. 1) ; the high spot where the two men are work- ing is cut 9b; house no. 3 appears below, to the right, and no. 4 in the foreground. Cut 9b yielded only 10 specimens, including a harpoon head with Old Bering Sea decoration at depth of 22 inches. The cut was begun in order to expose a cache, the whale skull walls of which became visible as cut 9 was being dug. Cut 9b was 6 feet long by 4 feet wide and was carried to a depth of 30 inches; it ex- posed only the NE. and NW. walls of the cache, which had been constructed of three whale skulls placed base upward, with several connecting timbers. Cut 14.—A small cut, 6 feet long and 4 feet wide, adjoining cut 9 on the NE.; taken down in six levels to a depth of 3 feet 5 inches. Only seven specimens, in addition to potsherds and stone blades, were found. Cut 15.—When a little preliminary digging had revealed the en- trance passage leading to house no. 4 an excavation was started in the area immediately in front (to the west) so as to uncover the passage in its entirety. This excavation was designated as cut 15. The cut was made in the area to the left of where the two men are shown digging, in plate 9, figure 1; plate 9, figure 3, shows cut 15 after the entrance floor (to house no. 4) had been reached. The cut was taken down in six levels to the floor, which averaged 5 feet below the surface. The first four levels, to a depth of 3 feet, yielded 40 artifacts, including the winged ivory object with Old Bering Sea decoration shown in plate 20, figure 1, found at a depth of 28 inches. 66 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 Level 5 (3 to 4 ft. deep) immediately above the roofing timbers, yielded 31 artifacts, including the decorated pail handle (pl. 19, fig. 3) from a depth of 44 inches; one harpoon head with Old Bering Sea decoration, 42 inches deep; and two other fragments from the same depth. Between the roof timbers and floor of the passage (level 6, 4 to 5 ft. deep) 15 artifacts were obtained, including one harpoon head decorated in Old Bering Sea style, found on the floor. Cut 16.—This was a narrow cut, 6 feet long by 3 feet wide, ad- joining cut ga on the SE. It was taken down in eight levels to the undisturbed gravel at a depth of 4 feet 4 inches. Fifty-five artifacts were found, including a broken drying rack of wood 18 inches deep and two ivory gut scrapers; a wooden drum handle with a large part of the rim (pl. 55, fig. 5) and a long oval cup or ladle of ivory (pl. 51, fig. 12) 22 inches deep; ivory scraper, Old Bering Sea decoration (pl. 16), 30 inches deep; ivory harpoon head, Old Bering Sea decoration, 37 inches deep. At a depth of 30 inches were found 12 seal noses, several of them strung together on a baleen thong. These had no doubt been preserved in accordance with religious practices connected with sealing, such as are still observed on St. Lawrence Island, Cuts 18, 23, 24, 25, and 20.—Excavated by M. B. Chambers in 1931. These five cuts, which were all connected, were situated just to the west of the entrance to house no. 3, in a section of the midden which yielded an abundance of artifacts and interesting evidence of old house structures at the bottom. Cut 18, which formed the NE. corner of the quadrangular space occupied by these five cuts was 6 feet 9 inches deep and was taken down in 22 levels. It yielded 363 artifacts, including many of those which are illustrated. The 28 harpoon heads and fragments were all of the older types; a fragment of a wrist guard from a depth of 12 inches has a “ modern ”’ decoration of spurred lines; three harpoon heads from depths of 48, 60, and 72 inches, although belonging to old types structurally, have the simple line decoration which I have called Punuk style 1; the others are either plain or bear an Old Bering Sea ornamentation. The following harpoon heads from cut 18 are illustrated: plate 24, figure 3 (78 in. deep), figure 17 (48 in. deep) ; plate 26, figure 5 (48 in.), figure 11 (48 in.), figure 12 (40 in.), figure 16 (66 in.) ; plate 72, figure 1 (60 in.). Other objects bearing Old Bering Sea decoration were found at the following depths: 18 inches (3) (pl. 15, fig. 4) ; 38 inches; 48 inches (2) (pl. 15, fig. 8, and pl. 19, fig. 5) ; 60 inches; 64 inches; 81 inches. Cut 23, adjoining cut 18 on the west, was 6 by 12 feet instead of the usual 12 by 12 feet. It was taken down to a depth of 6 feet 10 NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 67 inches, in 17 levels. It yielded 285 artifacts, very similar to those from cut 18. Three harpoon heads or fragments with early Punuk decoration from depths of 8 inches, and 14 inches (2) (pl. 24, fig. 21). The others are of older types and are either plain or decorated in Old Bering Sea style; those illustrated are: plate 24, figure 2 (65 in. deep), figure 4 (81 in.), figure 7 (14 in.), figure 9 (58 in.) ; plate 26, figure 2 (81 in.). Other Old Bering Sea objects are shown on plate 15, figure 2 (18 in.), figure 3 (72 in.) and plate 19, figure 4 (8o in.). Cut 24, adjoining cut 18 on the south, was also 6 by 12 feet. It reached a depth of 6 feet 9 inches and was taken down in 12 levels. It yielded 187 artifacts, very similar to above. One harpoon head with early Punuk decoration from depth of 18 inches, also the object on plate 22, figure 2. Two harpoon heads and two fragments with Old Bering Sea decoration from depths of 13 inches (pl. 26, fig. 20), 18 inches, 25 inches, 54 inches (pl. 26, fig. 3). The seal-headed toggle (pl. 15, fig. 12) comes from depth of 54 inches and two other objects with Old Bering Sea decoration from depths of 64 and 81 inches. Cut 25, which was 6 feet square, adjoins cut 24 on the west. It reached a depth of 8 feet and was taken down in 13 levels. Artifacts numbered 179. Two harpoon heads with Punuk decoration from depth of 18 inches; seven with Old Bering Sea decoration, from depths of 18 inches (2), 37 inches (pl. 26, fig. 15), 39 inches, 48 inches (pl. 19, fig. 8), 61 inches (pl. 26, fig. 17), 79 inches. Cut 29 was a narrow strip 24 feet wide to the west of cuts 23 and 25. It was taken to a depth of only 3 feet, in four levels. Fifty arti- facts were found, including the harpoon head shown on plate 25, figure 6, from a depth of 25 inches. The six photographs shown in plate 10, figures 3, 4, and plate 11, figures 1-4, will afford some idea of the nature of cuts 18, 23, 24, and 25, and of the house remains found at the bottom. Plate 10, figure 3, is a view looking SE. toward the hillside, showing the four contiguous cuts at an early stage of the excavation. The deeper excavation in the left foreground is cut 18. The Eskimo, Silook, is standing in cut 24, and on the ground in front of him is a large piece of walrus hide. Beneath this were found pieces of timbers, shown at the left in plate 10, figure 4. In plate 11 are shown four views of the excava- tion at a later stage. Plate 11, figure 1, shows the floor stones of the house, 5 feet deep, before being completely uncovered (at left center) and to the right of these, two rows of upright timbers mark- ing the position of the entrance passage. Plate 11, figure 2, shows a 68 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 different view of the same at a later stage; the remains of the floor consisted of a stone pavement roughly circular in outline and having an E.-W. diameter of 64 feet. The entrance passage, 2 feet wide, faced W.-SW., and seemed to have been formed of timbers laid horizontally behind upright logs; some of these remained in place behind the uprights on the south side. The floor of the passage, as indicated by a few stones and pieces of whale scapulae, was some 16 inches below the stone flooring of the house. A small whale scapula had been placed in a vertical position at the threshold and had been worn smooth in two places by the scraping of the feet, in just the same way as the whale jaw which formed the threshold of house no. 4 had been worn down. Plate 11, figure 3, is a view from another direction which shows in the foreground the stone flooring of the house and the passage leading off toward the W.-SW. Part of another floor, consisting of flat stones and scapulae may be seen to the right of the passage way; this may have been the floor of an annex to the passage. On the opposite side of the passage a small enclosure of fire-burned stones indicated a fireplace. Plate 11, figure 4, is a view of the stone house floor showing the pit near the center. It was lined with upright stones and probably covered over originally by the piece of whale scapula now fallen inward. The small winged object (pl. 59, fig. 22) is shown here in situ at the edge of the pit. Cuts 21 and 26.—Excavated by M. B. Chambers in 1931. These were two contiguous cuts near the south edge of the midden, to the southwest of cuts 18, 23, 24, 25, and 29. Cut 21 was 5 feet 6 inches deep, taken down in 13 levels. Cut 26 was 4 feet 7 inches deep and was taken down in seven levels. Neither of them was rich in material, cut 21 yielding 123 artifacts and cut 26 only 38. The four harpoon heads from cut 21 are Old Bering Sea types. One of them is the small specimen shown in plate 26, figure 14, found at a depth of 24 inches. A cache and part of the walls of another structure were found at the bottom of these two cuts. Plate 10, figure 1, shows the upright walls of the cache, at the left, and the horizontal timbers of the other wall, at the right, before either of them had been uncovered. Plate 10, figure 2, shows the same structures at a later stage, with the walls and floor of the cache, at the left, completely revealed. The cache walls had been formed of upright whale jaws and timbers and one whale skull. The floor was of flat stones covered with walrus hide, above which were several whale scapulae. The horizontal timbers at the right were probably part of a house wall. Three superimposed logs were visible, also several bone and wooden stakes that held them NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 69 in place and one large upright timber that had evidently been a roof support. Cut 27.—Excavated by M. B. Chambers in 1931. Near the center of the midden, overlapping the SE. corner of cut 19; 4 feet 4 inches deep ; taken down in 11 levels; bottom not reached. This cut revealed the remains of a rectangular structure, either a cache or a small house with four corner posts and walls of horizontal timbers. Plate 7, figure I, looking northwest, with cut 19 in the background, shows the pro- jecting ends of three of the corner posts, that in the upper right corner being a whale jaw, the others logs; at the center may be seen a hollow whale vertebra, probably a ventilator. Plate 7, figure 2, is a view from the same point taken later when the wall timbers were beginning to appear. Plate 7, figure 3, is a view of the cut looking in the opposite direction, southeast, with cut 19 in the foreground. The four corner posts are all visible, one of them, in the foreground, with a section of wall abutting (the post and wall timbers are lashed together to pre- vent them from falling as the earth was removed from around them). The timbers behind the man may also have been part of the wall. From this cut came 270 artifacts. In contrast to cut 19, there were more Old Bering Sea than Punuk pieces. The latter is represented by a fragmentary wrist guard from a depth of 9 inches (pl. 22, fig. 10) and a harpoon head, 16 inches deep (pl. 28, fig. 20) ; the Old Bering Sea pieces come from the following depths: 29 inches (pl. 24, fig. 12) ; 32 inches; 38 inches. HOUSES AT MIYOWAGH The house remains mentioned above, which were found at or near the base of the midden at a number of places, were too poorly pre- served to afford any clear idea as to their exact size, shape, and manner of construction. However, fragmentary though the evidence was, it indicated that the earliest houses at Miyowagh, like those at the Hill- side site, had stone floors, a narrow entrance passage at a lower level than the house floor, and walls built of horizontal, superimposed timbers. Fortunately, two other house ruins were found, which pro- vide a much better although by no means a wholly satisfactory basis for reconstructing the house types of the Old Bering Sea and early Punuk periods. These are the two houses, nos. 3 and 4, which, as mentioned above, were found in cut 9 and in the area immediately to the south. Although the excavation of cut 9 was begun on August I and continued into October, the major part of the work on the two houses, nos. 3 and 4, was conducted by J. A. Ford during my absence. 6 70 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 Mr. Ford’s careful excavations not only revealed the outlines and essential structural features of the houses themselves but also brought out the important point of their mutual relationship, as described below. House no. 3.—In plates 8 and g are shown a number of views of houses no. 3 and 4, both of which were originally included in the excavation which was begun as cut 9. They are also shown in text figure 11. The larger house, no. 3, faced NW. Plate 8, figure 3, look- ing NW., shows the exposed stone floor, the remnants of walls and the roof timbers of the entrance passage after the superimposed debris had been cleared away. The outer end of the passage had been en- larged to form a small antechamber 11 feet long by 63 feet wide. This antechamber was roughly oval in outline, with a stone floor, and walls made of whale skulls, stones, timbers, and smaller bones. The roof was formed of timbers placed transversely across two whale jaws which extended NW.-SE. The inner (SE.) wall consisted principally of a small whale skull which had been placed base down- ward on cross timbers beneath which the occupants had to crawl to enter the passageway leading to the house. The entrance passage proper was 5 feet wide and 16 feet 3 inches long and had a stone floor. The walls were made of stones and whale bones—principally jaws and scapulae—with an occasional walrus skull. Fhe roof was of timber, consisting first of three transverse beams, of about 10 inches diameter, which spanned the passage at 4-foot intervals; and over these a layer of smaller logs averaging 6 inches in diameter, laid lengthwise. There were two series of these small horizontal timbers, one extending for 12 feet from the antechamber toward the SE., with their ends resting on the third transverse beam ; the other only 4 feet long, with ends resting on the same beam and abutting the ends of the longer logs. The SE. ends of the 4-foot timbers rested on two cross beams which formed the inner end of the passage (pl. 8, fig. 4); just beyond were laid a few additional timbers and a whale jaw, with an opening between, through which the occupant would emerge as he left the passage and entered the house. Remains of two roof supports were found at the entrance, one the upright whale jaw visible in plate 8, figures 3 and 4, and the other a small log on the opposite side. In front of these, em- bedded in the floor, was a 5-foot section of a large whale jaw, which formed the threshold. The roof of the entrance passage had fallen almost to the floor, only 6 inches to 1 foot of heavy sand and gravel lying between the fallen timbers and the floor stones. From the positions of the broken off ends of the transverse roof beams it could be seen that originally the roof sloped downward toward the house, ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 71 I NO. ~~ SS a : LORS] G ———— _S——— we ee ee os ea ane va me oad | —————= ~ ~ eee een + — oe _——-—-— — 4, Miyowagh, showing remaining timbers. Fic. 11.—Ground plan of houses nos. 3 and 72 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 from a height of about 6 feet at the point of junction with the ante- chamber to 4 feet at the SE. end where it entered the house. Above the original level of the roof, except at the extreme NW. end, was a layer of soil 30 to 50 inches thick. The floor of the entrance passage was about 16 inches lower than the floor of the inner room. The dimensions of the house or inner room were 18 feet 6 inches NW.-SE. and 21 feet 2 inches NE.-SW. Unfortunately we were unable to obtain any evidence as to the nature of the roof which covered the inner room. Timbers and whale bones were found in the débris overlying the floor but in such disorder that it was im- possible to say whether or not they had formed part of the roof. The same was true of the upper part of the walls, although enough remained to show that at least the lower part had been formed of horizontal timbers laid one above the other, as in the case of house no. 2 at the Hillside site. The largest number of superimposed logs was four, at the north corner to the right of the entrance (see pl. 8, fig. 3). Five feet beyond the north corner was the remnant of an upright timber (in front of man standing, pl. 8, fig. 3), from which two horizontal timbers extended to the entrance, but it was not clear whether the small rectangular space thus formed had any real con- nection with the house. Of the NW. wall four logs remained, two to the left and two to the right of the entrance. At the west corner back of the two wall timbers, was the broken off section of a large upright wall support. Along the NE. side only scattered remnants of horizontal wall timbers were found, extending to a large upright timber, a corner roof support, at the east corner. The opposite, or SW. wall, was likewise much demolished, although enough single logs remained to show the outline. Across the south corner were two logs laid end to end, the purpose of which was not clear. Within the triangle thus formed were two small parallel logs, 30 inches apart, just below the floor level. Stuck into the ground along both sides of these two logs were a dozen or more small wooden stakes, no doubt for the purpose of holding in place additional logs which originally lay above them. These parallel logs might be interpreted as the rem- nants of an entrance to an earlier house which may have occupied the space back of (to the SE. of) house no. 3. In this area, which was included in cut 9a, we found a confused mass of fallen timbers, stones, and bones which seemed to be the remains of a house or struc- ture, the nature of which we were unable to determine. The remains of the rear or SE. wall of house no. 3 were somewhat better preserved than those of the two side walls (pl. 8, fig. 2, in front of the man at left). Beginning at the east corner the wall remnant consisted NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 73 of a whale jaw resting on one small log and overlaid by another ; just inside was a pile of four stones, the function of which is un- known. Continuing from these were two more logs, one above the other, followed by two similar logs extending to the south corner, braced on the inside by a wooden stake. The last mentioned logs are of particular interest in that they were laid directly over the fallen timbers of another and earlier house (no. 4), which is at the right in plate 8, figure 2, in the immediate foreground in plate 8, figure 3, and plate 9, figure I, and in the background in plate 9, figure 2. The central part of the floor of house no. 3 was covered with large flat stones, but along the walls the floor was of earth except for an occasional stone slab. Four feet from the entrance was found a small pit or cache between the floor stones. It was roughly triangular in outline, 2 feet long by 1 foot wide, and was covered by a piece of whale scapula. It was 34 feet deep and the sides were lined with stones and whale and walrus bones. The contents were loose earth and gravel with little refuse. Five artifacts were found in the cache: a strip of baleen 114 inches long by 1} inches wide; an ivory pick ; bone snow knife (pl. 79, fig. 5) ; bone awl; and a broken ivory box handle. The material found on and above the floor of house no. 3 was note- worthy in two respects: (1) the bone and ivory, particularly the lat- ter, was very dark in color, ranging from a deep brown to almost black; it was also brittle and poorly preserved; (2) every harpoon head and decorated object belonged to the early Punuk stage, not one example of Old Bering Sea art being found. This was in striking contrast to the adjoining areas where, except for a few Punuk pieces in the upper levels, only Old Bering Sea art was found. The follow- ing harpoon heads from cut 9 are illustrated: plate 24, figure 10 (36 in. deep) and figure 23 (24 in.); plate 26, figure 7 (48 in.) ; plate 28, figure 11 (12 in.) and figure 18 (24 in.). It is evident, therefore, that house no. 3 dates from the early Punuk period. It was probably one of the last houses occupied in the SE. section, which no doubt accounts for the fact that before excavation, it appeared as the deepest and most prominent pit of any on the midden proper. House no. 4.—As stated above, an interesting case of superposition was found at the back of house no. 3, the south corner of this house having been built over the fallen timbers of an earlier house (no. 4). The latter was a smaller, square structure, 9 feet 6 inches E.-W. by g feet 7 inches N.-S. (text fig. 11). It had a stone floor which was t foot lower than that of the larger house and a long curving entrance passage which faced toward the west. Plate 8, figures 2 and 3, and plate 9, figures 1 and 2, show the relationship of the two houses. 74 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 The remains of the rear or east wall of house no. 4 consisted of two whale jaws and two small logs laid horizontally behind two up- right corner roof supports (pl. 8, figs. 1, 2, at right center, and pl. 9, figs. 4, 5). The larger whale jaw was cracked so that it was some- what bowed inward; it rested on one of the logs, and both were propped up with a small wooden post and a walrus penis bone. The upper whale jaw had slipped from its original position and lay about a foot to the rear and just a little above the lower (pl. 9, fig. 4). Be- tween the two lay the second small log. Three feet back and about one foot above the whale jaws was the remnant of another wall section or perhaps of the same wall. It consisted of four pieces of logs and a fifth fragment superimposed, with a stake wedged in between. The south wall (pl. 9, fig. 2, in background) consisted of a tier of horizontal timbers and a piece of whale jaw, the ends extending 2 feet beyond the SE. corner post. The wall was braced on the inside by wooden stakes and a walrus penis bone. Plate 9, figure 4 (extreme lower right corner), shows the outside of the wall at the SE. corner, consisting of three superimposed logs, just behind the corner post. (The post is shown wrapped with string to keep it from disintegrat- ing.) At the SW. corner was another large upright corner post. The north wall consisted of three superimposed logs, the back ends of which are visible at the right foreground, plate 8, figure 3, behind the cord-wrapped NE. corner post. The end of one of the logs pro- jected 8 inches into house no. 3, at a level just below the floor. Above the wall timbers of house no. 4 and partly beneath those of the larger house, no. 3, lay a row of small timbers which had fallen inward from a vertical position. They could hardly have been part of a fallen roof, for the outer ends were resting directly on the uppermost of the three wall timbers, and the latter alone would not have been high enough to have constituted the entire wall. More likely they were supplementary wall timbers, which had been set into the ground, either upright or leaning slightly inward, just behind the horizontals which formed the base of the wall. Twelve of these fallen timbers remain, but there were evidently others, extending to the NW. corner, which had been broken into and removed in the construction of the later house, no. 3. Further and unmistakable evidence of the superposition at this point is afforded by the two stakes which had been driven into the ground to help support the rear wall of the later house. Plate 8, figure 2, shows the stake (at the left center) placed against the inside of the wall of the later house, and just opposite it the similar stake which had been driven into the gravel overlying the fallen walls of NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 75 the earlier house. In plate 9, figure 1, the same stake is visible from another angle, still resting on the pillar of gravel which was left in place so as to show this detail. In the disturbed NW. corner we were unable to find a roof support such as had been placed at the three other corners. Of the front or west wall nothing remained save the threshold, a whale jaw (pl. 9, fig. 3, in foreground) which extended from the NW. corner to within 20 inches of the post at the SW. corner. It was laid on edge and projected about 6 inches above the stone floor. Two deep, smoothed notches had been cut into the somewhat sharp upper edge of the bone, over which the occupants crawled as they entered the house. The entrance passage was traced out and excavated for a distance of 18 feet, beyond which point we were unable to go because of the lateness of the season and the frozen ground. The floor of the passage was of stone slabs (pl. 9, fig. 3), those just below the whale jaw at the entrance being 15 inches lower than the floor of the house; from this point outward the passage floor descended very gradually until about midway of its length it was 22 inches below the floor level of the house. The walls of the passage were formed of stones, most of them long slabs set upright ; some were braced with walrus penis bones and tusks. The roof of the passage, like that of house no. 3, had been constructed of logs, first a series of transverse beams spaced at inter- vals of 3 to 4 feet and over these a covering layer of longer timbers running lengthwise. Two wooden uprights placed in front of the whale jaw threshold served as roof supports for the inner end of the passage. There were three other upright timbers set into the walls of the passage which helped to carry the superstructure of the roof. Most of the transverse roof beams, however, rested on the up- right stone slabs except at one place, where the ends of two of them rested on two of the longer lengthwise pieces. At the inner end the entrance passage was 5 feet wide. Beyond this, at the center, it narrowed to a width of about 3 feet; then, at a distance of 12 feet, it turned rather sharply to the left, making a bulge or recess as it turned. Beyond this point the passage narrowed to a width of 27 inches. How much further it extended we were unable to determine, for the excavation did not reach this stage until October, and it was impossible to remove both the 5 feet of undisturbed frozen soil above the passage floor and the additional 4 feet or so of similar material, now also frozen, which we had thrown out. For the same reason the wall was not traced out entirely at two other places—the recess where the passage turned to the left and the enlargement at the inner end. 76 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 No artifacts of any kind were found on the floor of house no. 4 and few in the refuse immediately above. A badly damaged human skull and incomplete skeleton were found resting on the largest of the floor stones, at about the center of the house. Another skull, with the lower jaw and most of the face missing, was found on the floor of the entrance passage 46 inches from the SW. corner of the house. House no. 5.—Of the third house excavated at Miyowagh, nothing remained except some of the floor and wall stones and an imperfectly preserved passage way which faced W.-NW. (text fig. 12). The dimensions of this house were 22 feet 4 inches long (in the direction of the passage) by 19 feet 8 inches wide; the passage was 16 feet long by 3 feet wide, expanding at the front end to a width of 5 feet 8 inches. The walls of the passage had been formed of upright slabs, and the same was true of the house walls to judge from a few slabs which remained in place. Unlike the entrances just described, the floor of this one was at the same level as that of the house. No timbers or bones were found, nor any artifacts. Since the house was located at the north edge of the midden, which section was found to be later than the opposite side, this house was probably one of the latest to be occupied at Miyowagh. The floor was 2 to 3 feet below the surface. There was no evidence as to how the roof may have been built ; pos- sibly it was of skin. DECORATED OBJECTS, OLD BERING SEA, FROM MIYOWAGH The decorated objects from Miyowagh included no clear examples of what has been described above as Old Bering Sea style 1. The greater number of them fall into the more inclusive style 2, the prin- cipal design elements of which are shown in text figure 15; others fall into what I have tentatively assumed to be the still later Old Bering Sea style 3, the principal characteristic of which is the predominance of paired “ eyes.”’ Stratigraphic evidence of the sequence of the styles is less striking than at the Hillside site, where five of the eight objects decorated in style 1 came from beneath the floor stones of the two houses. Stylistic changes along the lines indicated undoubtedly oc- curred at Miyowagh but the evidence is less direct, as indeed might be expected in view of the fact that there is no sharp line of demarca- tion between styles 2 and 3, the latter being only a modification and in a sense a simplification of the more variable style 2. The distribu- tion of the various styles of decoration is shown in table I (p. 202). On some of the Miyowagh objects there appear new design elements and combinations, some of them apparently quite old, which add to Be INES 78 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 the richness and diversity of this old northern art. These objects, which are listed in table 1 under the heading “ Old Bering Sea art, style indeterminable ”, will be considered first. The flat ivory object shown in plate 15, figure 1, comes from cut 19, at a depth of 79 inches. The tapering ends, now broken, were originally perforated, suggesting that the object was worn suspended as an orna- ment of some kind. Decoration consists of pairs of curving lines forming bands with single or double rows of dots in the intervening spaces. At the center are two opposed arcs formed in the same manner, one attached to the upper border, one to the lower. There is also a vertical elliptical figure, rather crudely executed, containing at the center a square formed of dotted lines., Dotted triangles, some in- complete, occupy the triangular spaces above and below, and appropri- ately shaped dotted figures also fill in the ovoid and hourglass-shaped panels to the right and left respectively, of the center. This particular a NOOR mek 9 TEENS ES me NR ED Mere el Fic. 13.—Bone comb from Kurile Islands (after Baba, 1934). combination of pairs of curving lines bordered by rows of dots has not been observed previously in Old Bering Sea art, although the elements themselves are common enough. A rather striking parallel, considering that it comes from beyond the Eskimo territory, is pro- vided by a bone comb recently described from the Kurile Islands, (Baba, 1934, pl. 1, fig. 3), a sketch of which is shown in text figure 13. The broken adz handle shown in plate 17, figure 3, has an arrange- ment of curving bands and dotted lines very similar to that of plate 15, figure 1. This specimen was purchased from an Eskimo who had dug it up at Miyowagh. The framework of the design consists of con- verging rows of slightly curving parallel lines. The center line, which is deeply incised, is flanked on either side by a more lightly incised line and a row of closely spaced dots. Several small round pits oc- cupy some of the free spaces, and on the upper edge, in high relief, is a realistically carved human head. The combination of lines and dots is characteristically Old Bering Sea, but the small round pits are more typical of the early Punuk stage. NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 79 The function of the small perforated ivory object shown in plate 15, figure 9, is uncertain; the smaller end has been roughened, proba- bly for the purpose of holding a lashing. This object was found in the upper part of cut 19 (depth 17 inches) accompanied by harpoon heads of early Punuk type. The ornamentation, however, is definitely Old Bering Sea in conception even though the design is unique. It consists of a double line which runs along the edges and rises to two sharp prongs and a rounded projection at the upper end, an oval central figure, and two small detached crescents—all bordered by deli- cate dotted lines. The design on the broken brow band shown in plate 15, figure II (cut 27, depth 42 in.), is another that is new to the Old Bering Sea complex. Rising from a base line are three equidistant triangular figures with incurving sides and hatchured basal angles. In the curv- ing spaces between and beyond are four “ boat-shaped ’’ figures, the two outer ones with spurs attached to the upper line, the two inner ones with short cross lines between. In its conception the design as Fic. 14.—Ivory pail handle with Old Bering Sea decoration, Miyowagh. a whole is similar to that on another fragmentary brow band illus- trated in plate 58, figure 12. Plate 19, figure 1, is an ivory handle, probably for a baleen pail, from cut 19, depth 51 inches. The ornamentation, although simple, is unusual ; it is the same on both sides. The long row of slanting spurs, with one upright spur at the center is similar to the end designs on the scraper shown in plate 13, figure 8. Paralleling this spurred line is a dotted line which also follows the indentations made for finger grips. Along the upper, beveled edge is a long row of inward-pointing straight spurs, facing a similar row from the other side. Text figure 14, from cut ga, 18 inches deep, is a drawing of another pail handle, which is decorated in a somewhat unusual fashion. Down the center is a long narrow panel containing short cross lines arranged in pairs. To either side of this is a curvilinear ornamentation com- prising three elements: (1) at the ends four extremely attenuated petaloid figures composed of concentric lines; (2) continuing from these, four other figures, similarly formed but with concave outer ends, and (3) at the center two narrow rectilinear panels with con- cave ends. 8o SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 Plate 109, figure 5, is a small, flat ivory object of unknown use, evi- dently an ornament. It comes from cut 18, depth 48 inches. The upper, slightly convex surface bears a simple, graceful design, which, though typically Old Bering Sea in spirit, contains few of the elements usually associated with this style. A pleasing bilateral symmetry is produced through the harmonious arrangement of three closely spaced petaloid figures around a deep V-shaped depression, and at either end a curving panel enclosing two ovoid figures, each of which has at the center a pair of oblique, deeply incised lines bordered by more faintly incised lines. The small fragment, plate 15, figure 2 (cut 23, depth 18 in.), is decorated in a somewhat similar style: angular spaces outlined by double lines and bordered by faintly incised broken lines; at the end is a “ladder” design—a pair of horizontal lines between which are numerous closely spaced cross lines. The ivory fat scraper shown in plate 17, figure 2, from cut 10, depth 54 inches, provides an interesting example of the possibility of individual expression in the use of a conventional design—in this case of the bilobed figure, suggestive of the flattened-out head of a fish, which is often seen in Old Bering Sea art. One of these “heads ” is visible on the harpoon socket piece shown in plate 13, figure 1 (the second “ head” from the top, on the right side) ; other examples will be described below. In the present specimen the two bilobed figures lack the customary “eyes” and the slightly everted “ snout ” in low relief, which are usually responsible for the zoomorphic ap- pearance. Instead, they contain small crescents and curving lines and “Jadder ” figures, some of which are so arranged as to form an inner design similar to the enclosing one. These two opposed bilobed ele- ments are now brought together in a most skillful manner by means of a centrally placed quadrangular figure with incurving sides. The ends of this central figure, formed of a series of concentric curves, appear at first glance as continuations of the bilobed figures them- selves, the effect of which is the blending of the three independent elements into a graceful, compact, harmonious design. The ivory objects illustrated in plate 15, figures 3-6 and 10, are decorated in what I have called Old Bering Sea style 2. Comparable examples from the Hillside site have already been described (pl. 13, figs. 1-3, 6; pl. 23, figs. 1, 2, 4, 9, 10, 14). The stratigraphic evidence obtained at that site, although less conclusive than one might wish, seemed clearly to indicate that style 2 was later than style I, an in- terpretation which finds further corroboration in the fact that style 1 is absent at Miyowagh. As pointed out previously, there are marked NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 81 stylistic differences between the two, even though certain elements are possessed in common. Style 1 was characterized by a relative preponderance of straight, radiating lines, short detached lines and prominent spurs; concentric circles were also used, but these did little to relieve the general appearance of diffuseness. Style 2, on the other hand, makes effective use of curving lines, and these, together with circles, dotted lines, and independent design elements of various kinds, are combined to form graceful, harmonious patterns of rare beauty. The principal motives and design elements of style 2 are illustrated in text figure 15. Plate 15, figure 3, is a flat ivory object, evidently an ornament, from cut 23, depth 72 inches. Two slots on the under side show that it was attached to some flat surface. This is one of the few instances in which the geometric elements of Old Bering Sea art have been so arranged as to produce a clearly recognizable, almost realistic animal form. The exposed teeth, particularly the long, sharp canines, are in- dicative of a carnivore, an appearance which is further accentuated by having the nose carved in relief with a pit at the end for the nostril. The eye is represented by a nucleated, concentric circle enclosed be- tween two pairs of lines which come together just below a deep round pit set in an elevation representing the ear. From this point back, the ornamentation follows the usual course in being essentially geo- metric, although, as is so often the case, the two elevated “eyes” near the edges are so placed in relation to curving panels as to suggest the idea of a pair of animal heads. If viewed with the “ snout ” in- clined upward, these are suggestive of a fish or mammal; if down- ward, of a bird. The impression of a bird’s head is heightened by having the “beak” divided and by having the “eye” set cometlike between diverging lines which give it a decidedly avian appearance. An unusual detail is the backward continuation of the faint dotted lines that enclose the circles and their appended spurs. These two “heads ” are outlined by two deeply incised lines, widely separated at either end, close together at the center. At the front they enclose an oval panel formed of concentric lines, spurred along the outer periph- ery, and containing in the center an elevated nucleated circle with streaming lines attached. This filler design serves as an effective bal- ance between the two curving side panels, or “ heads ”’, and the realistic mammal head at the end. Similar oval panels, it will be recalled, were present on an elaborately decorated gorget found at the Hillside site (pl. 13, fig. 2). At the back the composition ends with a panel of triangular outline with rounded apex enclosing a small nucleated circle and a V-shaped figure set between two long oblique spurs. The 82 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 a Or @® 16 Fic. 15.—Principal decorative motives of Old Bering Sea style 2. NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 83 pits around which the circles are inscribed are slightly more than 3 mm deep, and each of them, except the one at the rear, is placed on a rounded elevation. The two larger holes extend through to the under side, emerging at the openings formed by the two slots. Plate 15, figure 4, is a small pendant or toggle with a slotted base from cut 18, depth 18 inches. Decoration is restricted to a series of lines around the lower border: first a pair of parallel lines with cross lines at regular intervals forming narrow rectangular spaces, each with a single straight line at the center ; rising from this base line at either end is a curving line, terminating in a sharp tip at the center ; a broken line follows the lower border of the opening at the top. Plate 15, figure 5, is the lower end of a harpoon head from cut 3, depth 54 inches. The broken edges are smoothed and worn as if by constant handling, suggesting that it was preserved, perhaps as an amulet, long after it had been broken and discarded. The ornamenta- tion was applied with more than ordinary skill. The design centers around two figures placed on either side of the sunken pit where the lashing slots emerge; one of these is ovoid in shape, with a small four pointed star at the slightly raised center; the other is an oval panel with a straight base from which rises a small triangular figure with incurving sides, a design similar to that observed on the haunch of the ivory polar bear from the Hillside site (pl. 13, fig. 3). The button-shaped ivory ornament shown on plate 15, figure 6, was found in cut 19, at a depth of 63 inches. The base is slotted in the manner of plate 15, figures 3 and 4, and a central perforation extends through to the slot from the upper side. By cutting out four triangular segments the remaining central portion was given the ap- pearance of a Maltese cross. Each of the four arms is ornamented by bordering lines, continuous and broken, inclosing at the center a nucleated concentric circle. Plate 15, figure 10, is a small hook which appears to have been re- adapted from some other implement. It comes from the surface layer of cut I9 in the Punuk section of the midden and should probably be regarded as a relic or an heirloom. At the lower end are two small circles with appended lines, placed at the center of oval panels, and on the upper end is a similar panel enclosing a circle (partly cut through by the hole) from which two pairs of appended lines descend obliquely to the right. The design is completed by the usual broken and continuous lines. In plate 17, figure 1, is shown a fragmentary fat scraper from cut Qa, depth 50 inches, which is decorated in a style difficult to classify exactly. Despite the prominence of the circles, which are suggestive 84 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 of style 3, the design as a whole should probably be classed as style 2. The execution is not particularly good and is carried out with more of a flourish than is customary. The design differs from others of comparable complexity in that the component elements are enclosed in a single field. The most prominent part of the design—the lower, as shown—takes the form of a large bilobed figure with the char- acteristic upcurving end which is usually responsible for the animal- like appearance. In the present instance, however, the filler elements seem to carry no further implication of zoomorphism. The pair of large concentric circles in the two lobes are placed too far back to resemble “eyes”, and the three other circles toward the end are like- wise inconsistent with such an interpretation. To the uppermost pair of circles is attached a long single spur, or V-shaped figure, pointing obliquely downward. The two lower circles are each formed of two lines with the space between filled with numerous cross lines; these circles have a pair of opposite spurs attached and they differ from the two circles above in the further particular of having no dot or “nucleus ”’ at the center. Between the two lower circles is another, which is nucleated and concentric, with two backward-pointing, hach- ured spurs attached. A series of five parallel lines, the center one more deeply incised than the others, begins at the lower end of one of the uppermost circles, and curving downward, continues up to join the opposite circle. At the lower end of the design is an oval figure flanked by hachured triangles, so opposed as to appear hourglass- shaped. The upper borders of the two lobes are formed of rather carelessly made double lines with cross lines or spurs in between. Where the cross lines are relatively uniform as to length and breadth, they produce a curving “ladder ’’ design; where they are shorter and thicker at one end, they produce only the effect of a “toothed” or spurred line. It is such a design—a double line, part of which is spurred and part ladderlike—that encloses an oval panel, the long tapering end of which extends down between the two lobes of the larger figure. At the center of this oval panel is another prominent concentric circle (the sixth) with two large V-shaped spurs attached. In plate 16 are shown three views, drawings, of one of the most splendid examples of Old Bering Sea art thus far found. It comes from cut 16, depth 30 inches. The function of the object is uncertain ; the hole at the end was no doubt for a suspension thong, and the two broad, smoothed notches on the sides seem clearly designed as finger grips; the hollow, bifurcated end is suggestive of a scraper of some kind—perhaps for removing the fat from small intestines— but if so, it is a type not previously known. The general conception NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 85 of the design, the harmonious relationship of carefully selected ele- ments, mark the work as that of an assured artist. The esthetic effect of a naturally pleasing composition is heightened by a virtuosity of execution which makes this object stand out as one of the finest ac- complishments of Old Bering Sea art. At the perforated end are two opposed pairs of “animal heads”’ practically identical with those on the ivory object previously described (pl. 15, fig. 3). The space be- tween the open “ jaws” is covered by a series of deeply incised lines, broken lines, circles, and spurs; at the center is a small elevated, nu- cleated circle with four spurs attached which give it the appearance of a four-pointed star. Between the constricted finger grips is another circle, with three spurs attached. From this point on the ornamentation is somewhat attenuated, consisting of bold curving panels which follow the outline of the curving bifurcated end. The under surface, although unetched, is divided into panels by a continuation of the deeply in- cised bordering lines from the upper side. Additional examples of Old Bering Sea style 2 may be seen on a number of harpoon heads, such as plate 24, figures I, 5, 7, 11-13, 17, 18; and plate 26, figure 17. The ornamentation in plate 24, figure I (cut 7, depth 75 in.) centers around a pair of large concentric circles near the center and a pair of smaller circles between converging lines above and below the two line holes; all of the circles are flat and have short spurs attached. Plate 24, figures 5, 7, and 11 (from cut 9, depth 24 in.; cut 23, 14 in.; cut 7, 92 in.) have parallel rows of con- tinuous and broken lines, some of which form curving panels; circles are absent. Plate 24, figures 12 and 13 (cut 27, depth 29 in. and cut 7, 96 in.) are somewhat more elaborately ornamented. In both cases the surface is divided into well defined panels; in figure 12 effective use 1s made of very small circles and long lines or spurs bordered by lightly incised broken lines; figure 13 has two side panels resembling animal heads but lacking the pair of “ eyes” usually found on these ; between the line hole and blade slit is a rectangular panel containing cross-hatched lines, one of the rarer Old Bering Sea designs. Plate 24, figures 17 and 18 (cut 18, depth 48 in., and cut ga, depth 36 in.) might be considered as belonging to style 3, were it not that the op- posed “ eyes’ are subordinated to a rather profuse linear ornamenta- tion. It is of interest to observe that traces of red pigment are visible in the lines on both specimens. As explained above, “‘ Old Bering Sea style 3 ” is the somewhat arbi- trary designation given to that style of ornamentation which is based primarily on the use of elevated concentric circles or ellipses, usually arranged in such a manner as to suggest the eyes of an animal. A 7 86 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 typical example is the small ivory object from cut 18, depth 48 inches, shown in plate 15, figure 8. The ornamentation on both sides centers around a pair of raised, concentric ellipses, which combined with curv- ing lines at the rounded tip, produce a strong impression of an animal head. Small plugs of baleen with the centers hollowed out are placed in the “eyes”’ and in the smaller circle below. The harpoon head fragment shown in plate 19, figure 7 (cut 9, depth 20 in.) is decorated in a similar manner, although a certain variety is afforded by a judicious use of bordering lines, spurs, and a small graceful petaloid figure. The elaborately decorated pail handle shown in plate 19, figure 2, from cut 7, depth 67 inches, is a good example of what can be ac- complished by an effective grouping of concentric circles or ellipses. In this case the elevated concentric ellipses are so profusely applied as to prevent any close resemblance to an animal’s head, even though the individual panels enclosing the “eyes”? might well represent a series of such heads. Whatever may have been the intent of the artist, the composition as a whole is more suggestive of a feather or plant design than anything else. Although repetitive in character, the design is remarkable for its beautiful flowing symmetry ; repetition has served only to intensify the feeling of life and movement inherent in the graceful component elements of the design. The ivory pail handle shown in plate 19, figure 3 (cut 15, depth 44 in.) 1s unique in combining a typical geometric design with a per- fectly realistic animal carving. The head is that of a walrus, with one tusk broken off. It is carved in high relief and is remarkably life- like in every detail. The accompanying incised decoration is a scrolled design made up of concentric rows of continuous and broken lines, enclosing a pair of raised ellipses. This design is the nearest approach to a true scroll that has yet been found in Old Bering Sea art. The curving lines, however, instead of being free at the ends, all come together and continue as a single line to join the opposite curve. Another example of style 3 is the beautifully decorated harpoon socket piece (with lower end broken off) shown in plate 27, B. This specimen is owned by Capt. E. D. Jones, former Commander of the Coast Guard cutter Northland, who obtained it from an Eskimo at Sevunga. It probably came from the Kukuliak midden. In the two well-defined animal heads just above the center we have another of the rare examples where the artist chose to portray an actual life form by a varied arrangement of the usual geometric motives. With consummate skill the two animal heads are so blended into the com- position as to avoid any feeling of disharmony. Although they form NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 87 the dominant element in the decorative scheme, there is no undue emphasis or contrast to impair the effectiveness of the composition as a whole. This harmonious relationship has been consciously retained by having the eyes and inner border of the mouth made to conform exactly with the other elevated circles and curving lines. The ears are carved in low relief, the nose is higher. The opening of the oblique line hole lies at the inner angle of the jaws, and the open mouth with its long sharp teeth is shown holding the upper end of the socket piece. Behind the animal heads is a graceful bilobed figure, which with its two elevated concentric circles, has the appearance of a flat- tened-out fish head. Opposed to this is a somewhat similar figure enclosing circles which are set off by outward-pointing curving lines. Between the two figures are two small animal heads in low relief, featureless except for the eyes which are small drilled holes. Style 3 was peculiarly adapted to the rounded contours of harpoon heads, and many of the finest Old Bering Sea specimens are so dec- orated. Several have been described previously (Mathiassen, 1929, fig. 13, a; fig. 14, b: Collins, 1929, pl. 1). A single example was found at the Hillside site (pl. 23, fig. 13); others from Miyowagh are shown in plate 24, figures 14-16, and plate 26, figures 1-3, 15, 18-20. Although there may be considerable variation in details, the general pattern remains constant: pairs of elevated concentric circles or ellip- ses, usually with a small pit at the center, placed on either side of the line hole and often on the basal spur. As a rule the artist chose to emphasize the idea of an animal’s head by having the “ snout ” stand out in relief, thus forming small ornamental projections or barbs along the edges. The panels enclosing these “ heads’ are marked off by deeply incised lines within which are more lightly incised con- tinuous and broken lines. As decorative schemes, these “ animal heads ” are highly effective on both closed and open socket harpoon heads. Although they appear somewhat more standardized when ap- plied to the closed socket heads, they definitely add to the graceful symmetry of these forms (pl. 26, figs. 18-20). They likewise appear to excellent advantage on open socket heads, where the prominent divided spur provides opportunity for interesting variation (pl. 24, figs. 14-16; pl. 26, figs. 1-3). In plates 20 and 21 are shown five of the ivory winged objects which are so characteristic of the Old Bering Sea period. None of these Miyowagh specimens has recurved wings such as were seen on two comparable specimens from the Hillside site (pl. 12, figs. 1, 3). They are, however, very similar to the unfinished specimen from that site shown in plate 12, figure 2. 8&8 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 90 The beautifully decorated winged object shown in plate 20, figure 1, comes from cut 15, depth 28 inches. In shape and to a certain ex- tent in decoration, it is much like one described by Gordon (1916, fig. 100). On the front, or flat, side the wings extend out at an up- ward angle from the wedge-shaped center; on the reverse (pl. 21, fig. 1) the wings were cut down, leaving the central part in high relief. At the base is a square socket measuring 1.3 cm across and 1.5 cm in depth. Above this socket are two round lashing holes 7 mm in diameter, drilled obliquely from the bases of the wings and meeting at about the center. The ornamentation is to be classed as style 3, in that it is based primarily on the use of circles. Real artistry has been shown in the selection and execution of an appropriate design, and the harmonious relationship of the circles and the rows of parallel and curving lines produces a bold sweeping surface ornamentation which is perfectly adapted to the flowing symmetrical contours of the object itself. The ornamentation on the reverse (pl. 21, fig. 1), while pleasing, has not this same feeling of motion. It consists of three independent design elements : a very large circle surmounting a central boss, enclosing a smaller nucleated circle and radiating lines; and two panels formed of a succession of curving lines within which are two large flat circles with the usual streamers attached. In the free spaces below and to either side of the large central circle are four smaller circles, all of them nucleated and concentric. Plate 20, figure 2, is an unfinished object which is practically iden- tical in general outline with the one just described. It comes from cut 7, depth 80 inches. Plate 20, figure 4, is a poorly preserved specimen from cut 6, depth 36 inches. The raised central portion on the reverse (pl. 21, fig. 4) has small upwardly inclined wings like the two preceding specimens, but on the front or flat side the wings have the appearance of sloping downward. This impression is due in large part to the decorative treatment of the wings, the panels of which flow downward from the center in contrast to plate 20, figure 1, where the lines on the wings radiate upward, thus giving the wings the appearance of a more marked upward inclination than they actually have. The ornamenta- tion in plate 20, figure 4, belongs to style 2, consisting of relatively complex designs composed of closely spaced straight and curving lines—continuous and broken—small nucleated concentric circles and rows of short, parallel lines. The elaborately ornamented front sur- face is badly cracked and eroded. On the reverse the central circle and its appended spurs enclosing a V-shaped figure is practically a counterpart of the design at the back end of the flat ivory object INO ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 89 shown in plate 15, figure 3; furthermore, the “ bird heads ” with down- ward pointing beaks which are incised on the wings are practically identical in form and spirit with those shown in plate 15, figure 3. The basal socket on the present specimen is rectangular in shape and is 1.4 cm deep; the cylindrical lashing holes were drilled obliquely through the wings, beginning at their bases and emerging at the notches above. The small winged object shown in plate 20, figure 3 (cut 8, depth 23 in.) differs from the others in having a raised central section on the back side which is more or less rectangular in outline with very short wings. The basal socket is oval-rectangular, and the circular lashing hole runs transversely through the raised center. The decora- tion on the back side (pl. 21, fig. 3) consists of several carelessly incised lines and a very schematic human figure—the body represented by a straight line, arms by two short lines and head by an oval with eyes, eyebrows, nose, and mouth indicated by dots and lines—just such a drawing as might be made by a child. The crude and care- lessly incised figures on the front side also appear to be the work of a child or an amateur. Plate 20, figure 5, is a poorly preserved specimen from cut ga, depth 38 inches. The raised central section on the back side differs from the others in having the wings pointed downward. The socket in the base is 2.3 cm long, 1.7 cm wide, and 1.2 cm deep. The lashing holes were drilled up from the bottoms of the wings as in plate 21, figures I and 3, but they stop at the center of the raised section where they meet in a deep groove. Ornamentation on this side consists of nucleated concentric circles, long, pointed spurs, and straight and curving lines, skillfully arranged with reference to the deep groove at the center. On all of the above objects except the unfinished specimen, plate 21, figure 2, there is a shallow indentation or pit at the end of the pointed central projection. The possible significance of this feature with regard to the function of the winged objects as a group will be dis- cussed later when the comparable forms from the Punuk sites are described. In addition to the above there are three unfinished specimens (from cut 7, 51 in.; cut 23, 25 in.; cut 27, 16 in.) and a small decorated fragment of a wing from cut 18, depth 18 inches. There are also three toys, well made miniature replicas of the full size objects, from cut 18, 60 inches; cut 27, 46 inches; cut 5, 23 inches. These are illustrated with other toys in plate 59, figures 22, 23, 24. go SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 Six decorated ivory needle cases from Miyowagh are shown on plate 17, figures 4-9. The ornamentation on figures 4, 8 and 9 shows up more distinctly in the drawings on plate 18. Plate 17, figures 4 and 8 (cut ga, 24 and 20 in. deep, respectively) are examples of Old Bering Sea style 3, the ornamentation in both cases centering around pairs of “animal heads” with prominent eyes formed of elevated concentric circles. Plate 17, figure 9 (cut 4, depth 55 in.) is decorated in an unusually delicate, graceful style. Four looped figures formed of finely incised concentric lines occupy the flattened center and enclose a long, narrow, sharp-pointed figure and three detached dots; a fourth dot has a circle inscribed around it with two short spurs attached to the periphery. The graceful symmetry of this entire central design is further enhanced by the decorative treatment applied to the projecting flanges at the center and the top. The ivory needle case shown in plate 17, figure 6 (cut 9, depth unknown) bears an unusual decoration consisting of obliquely in- cised longitudinal bands each composed of three parallel lines, each pair of bands connected by a series of short, curving cross bands also formed of three lines. : Plate 17, figure 7, is a poorly preserved needle case from cut ga, depth 13 inches. Around one end is a band of three parallel lines, with spurs attached to the lowermost line. At the center is a similar band of four lines with spurs pointing both upward and downward, and four equidistant petaloid figures each enclosing a row of five dots and having at the top three sharp projecting spurs. Below these are four more petaloid figures, each with a short “arrow”’ design at the center. The simplest of the needle cases is that shown in plate 17, figure 5, from cut 7, depth 33 inches. It has a double flange along either edge and a simple straight line decoration. In plates 15 and 19 are shown several Old Bering Sea objects with simple ornamentation which does not fall into any of the three recog- nized “styles.” Plate 15, figure 7 (cut 19, depth 34 in.) is an ivory plug with an oval face and blunt cylindrical tang. At the center is a deep, four-pointed figure with concave sides, cut out around a central nucleated circle—or cylinder—and outside of it four small round pits. In the central excavation is the remnant of a wooden inset. Plate 15, figure 12 (cut 24, depth 54 in.) is probably a drill cord handle. It is similar to one from St. Michael figured by Nelson (pl. 37, fig. 15). It is formed of two seal heads facing in opposite directions; the eyes are formed of slightly projecting wooden plugs set in round pits. NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS gI Plate 19, figures 4 and 6, are two more animal figures. The first (from cut 23, depth 80 in.) is an ivory object with a deep cylindrical socket in the larger end and the smaller end carved to represent the head of a mammal with prominent teeth. Figure 6 is a poorly pre- served carving of a walrus from cut 9, depth unknown. The base is slotted in the same manner as the two small bird figures from the Hillside site (pl. 13, figs. 4, 6). From the foregoing analysis we have seen that there are three gen- eral styles of Old Bering Sea art. Style 1, which was found only at the Hillside site and under conditions which seemed to indicate that it was the earliest of the three, is also the simplest and most generalized (text figure 6 and pl. 12, figs. 11-14; pl. 13, figs. 5, 7, 8). One object from Little Diomede Island (pl. 14, fig. 5) is also in this style and three others are similar (pl. 14, figs. 3-4, 6, and text figure 8). It is essentially a linear style, being characterized by a profusion of radiat- ing lines, long, sharp spurs and broken lines. Concentric circles and slightly curving lines were also used to a certain extent. Style 2 is distinctly curvilinear, and stands out in rather striking contrast to style 1, even though it employs some of the same ele- ments, such as spurs attached to lines, broken lines, and circles. The characteristic motives of style 2 are illustrated in text figure 15 and also in plate 13, figures 1-3, 6; plate 23, figures I, 2, 4, 9, 10, 14, all from the Hillside site, and plate 15, figures 3-6, 10; plate 24, figures I, 5, 11-13, 17, 18, etc., from Miyowagh. In addition to those illustrated there are 15 other examples of Old Bering Sea style 2, mostly fragments, from various depths at Miyowagh, as follows: cut 2, 43 inches ; cut 3, 60 inches; cut 6, 26 inches; cut 9, depth unknown ; cut 9 b, 22 inches; cut 10, 54 inches; cut 18, 18 and 64 inches ; cut 19, 24 and 78 inches; cut 24, 18, 25, and 64 inches; cut 27, 32 and 38 inches. The distinction between styles 2 and 3 is less sharply defined and the chronological implications less clear; however, the fact that style 2 is well represented at the Hillside site and that most of the com- parable specimens from Miyowagh come from the lower levels of the midden, would seem to indicate its priority over style 3. The latter is stylistically nothing more than an adaptation of style 2 in which special emphasis is placed upon concentric circles. These circles are larger than those of style 2 and are usually, but not invariably, ar- ranged so as to suggest the eyes of an animal. In some cases the smaller circles of style 2 also carried out the “ animal head ”’ concept, but as a rule they were not so employed. Along with the increased emphasis on, and dominance of, circles in the decorative scheme of g2 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 ‘style 3 there is also a general tendency toward simplification or sub- ordination of the accompanying ornamentation with the result that parts of the surface are frequently left smooth and plain; this is in contrast to style 2, where the profusion of small circles, straight and curving lines, and spurs, together with the practice of filling in the intervening spaces with dotted or broken lines, is suggestive of some- thing like an horror vacui on the part of the artists. While the rich and diversified ornamentation of style 2 is effective in avoiding even the slightest feeling of monotony, the smooth, flowing, and somewhat more restrained ornamentation of style 3 (e. g., plate 20, figure 1; plate 27, B) is in its way even more pleasing. Style 3 is especially characteristic of St. Lawrence Island. A few examples are also known from Bering Strait and the Arctic coast (Jenness, 1928 a, pl. XIII, c; Mathiassen, 1929, figs. 13 a, 14 b, 15 b), so that it cannot be regarded as a strictly local development on the island. However, it is nowhere so prominent as on St. Lawrence, where it may have first appeared, as a late and specialized development from style 2. There are seven additional examples of style 3 which have not been illustrated, all of them harpoon heads, from the following locations: cut 8, 32 inches; cut 9, depth unknown; cut ga, 18 inches; cut 13, 54 inches ; cut 16, 37 inches; cut 25, 18 and 39 inches. Finally, there are five specimens on which it cannot be definitely determined whether the decoration is style 2 or 3: cut 15, 60 inches; cut 18, 38 inches ; cut 19, 56 inches ; cut 25, 81 inches ; cut 20, 12 inches. DECORATED OBJECTS, EARLY PUNUK, FROM MIYOWAGH We have now to consider the Punuk or intermediate style of art from Miyowagh. Punuk art is characterized by the use of straight or slightly curving lines, which in contrast to those of Old Bering Sea art, are deeply and evenly incised. It also employs perfectly round compass-made circles, and small circular pits or dots either free or placed at the ends of short lines. The resulting patterns, though graceful enough, appear rigid and mechanical as compared with the elaborate flowing designs of Old Bering Sea art. Furthermore, there is a striking difference in technique. Old Bering Sea circles are in- variably somewhat irregular, having been made freehand; these and the accompanying lines could well have been made with stone tools. Punuk ornamentation, on the other hand, is clearly the product of metal tools, as may be seen from even a cursory examination of such typical examples of the art as those from Ievoghiyoq shown on plates NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 93 65 to 69. First, however, we will consider the Punuk pieces from Miyowagh, most of which, significantly enough, are somewhat less straightened in appearance than the later and more typical examples from the pure Punuk site, Ievoghiyoq. The earliest Punuk style, which we will designate as Punuk 1, may be divided into two phases. Phase 1, which is restricted to harpoon heads, makes use of lightly incised lines, straight and cleanly cut, but not very deep. Phase 2 employs the same kind of lines but in addition has short, lightly in- cised spurs, dots, and (rarely) freehand circles. At Miyowagh, Punuk ornamentation was applied to harpoon heads much more frequently than to any other artifacts. Consequently, these provide the best basis for chronological determinations, not only be- cause of their large numbers, but also because of definite and consistent changes of form which are correlated with this particular style of art. These will be discussed in some detail in the section on harpoon heads. It will be sufficient here to point out that practically all of the harpoon heads with Punuk decoration came from the northwest- ern part of the midden or from the upper levels of the southeastern part. One of the oldest forms of harpoon heads with this decoration is that shown in plate 24, figures 19-21. These heads, belonging to open socket type III x, represent a stage in the process of simplification which took place within this type; they differ from the earlier form (represented by pl. 24, figs. 11-18) in that they are smaller, have a less prominent basal spur, and a decoration of simple lines instead of elaborate Old Bering Sea designs. However, they have retained the graceful shape of the earlier heads, as well as the small curving barbs along the edges. Furthermore, the small rounded elevations around the line hole are clearly survivals of the Old Bering Sea “eyes” which were formed of circles or ellipses surmounting elevations that occupied the same positions. It should also be noted that the lines in figures 19, 20, and 21 follow the same paths as those more deeply incised lines which, on the Old Bering Sea heads, served to divide the surface into decorative fields. This particular style of decoration, consisting of nothing more than lines, is designated as phase 1 of Punuk style 1. Plate 24, figures 22 and 23, represent the end result of simplification within this type of harpoon head; they have lost the smooth, rounded contour and the graceful barbs of the older heads; the vestigial “eyes” are still further reduced, and the orna- mentation (in fig. 23), which includes the use of light spurs, is that which is typical of phase 2 of Punuk style 1. Exactly the same condition is observed in the next group of harpoon heads—open socket type III] y—which differs from the foregoing Q4 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 only in having the blade slit at right angles to the line hole. The earlier examples of this type from Miyowagh (pl. 26, figs. 1-5) are large, and most of them have prominent divided basal spurs and an Old Bering Sea decoration; the later forms (pl. 26, figs. 6-10) are smaller, have basal spurs which have been reduced to a point where they are merely “irregular ’’, and have a typical Punuk 1 decoration consisting either of lines alone (phase 1—pl. 26, fig. 6) or of lines together with short spurs and dots (phase 2—pl. 26, figs. 7-9). Plate 28, figure 23, is one of the few closed socket heads with a phase 1, style 1, decoration. Phase 2 of Punuk style 1 comes into greater prominence with the appearance of the next group—open socket types III (a) x and III (a) y—which are the two dominant types of the early Punuk stage at Miyowagh. Examples are shown in plate 28, figures 8-17, as well as on the less common closed socket forms, such as plate 28, figures 19-22, 25-27. One of these (pl. 28, fig. 15) differs from the others in having rows of dotted lines down the sides, a feature reminiscent of Old Bering Sea art. A typical example of phase 2 of Punuk style 1 is that incised on the ivory ulu handle shown in plate 22, figure 1. This specimen was purchased from an Eskimo who had excavated it from a depth of 3 feet. A unique feature is the circular opening at the front end for the index finger; there is also an oval depression for the thumb and on the opposite side two grooves for the fingers. The ornamentation is rather simple, consisting of single lines and dots, the latter either detached or placed at the ends of short lines. At the front end, around the finger opening, is a pair of lightly incised parallel lines and sev- eral rows of dotted lines, both of which are Old Bering Sea elements. At the back end there was a small perforation which may originally have held an ornamental link. The blade slit is 4 mm wide and 10 mm deep. Another example of phase 2 of style I is the small ivory plaque or ornament shown in plate 22, figure 4, from cut 19, depth 12 inches. It is decorated with curving lines and dots and has at the center a nucleated circle which was made freehand. The lines lack the regular- ity and precision of those typical of later Punuk art, but they have the characteristic straight, flat bottom. In addition to the above described examples of Punuk style 1, there are others not illustrated, but which are listed according to depth and locality in table 1. Punuk style 2 is somewhat more diversified than style 1, and for that reason may be divided into four more or less distinct phases. NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 95 Phase 1 is a decoration restricted to harpoon heads, such as plate 28, figure 14. It is characterized by very long spurs attached to lines and is merely a later and more advanced stage of Punuk style 1, in which the spurs were shorter, as on the three harpoon heads shown in plate 28, figures 11-13. The lines and spurs are also lightly incised as in style 1. In phases 2, 3, and 4 of style 2 the incisions are deeper and more sharply defined. Phases 3 and 4 were not present at Miyowagh; they will be described later. Phase 2 employs lines, spurs, dots (pits), and nucleated circles, the latter made with a metal bit or compasses. This phase is represented by only 10 examples at Miyowagh, all but one of them from the upper levels of the northwestern section. The ivory hand drill or reamer shown in plate 22, figure 8 (from cut 20, depth 12 in.) is a typical example of phase 2. Another is the ivory object shown on plate 22, figure 2 (from cut 24, depth 18 in.). On both of these objects the arrangement of the shorter lines and dots is that of style 1, but the mechanically made circles place them in style 2. The only harpoon heads from Miyowagh decorated in phase 2 of Punuk style 2 are the two shown in plate 28, figures 24 and 28, both of them from the surface layer of cut 20. The other decorated objects shown on plate 22 do not fall exactly into any of the above categories. They are to be classed as “ inde- terminable ’’, but since all ‘of them, with the exception of figure 10, were purchased, they are not included in table 1, which, as previously stated, contains only those objects which we ourselves excavated. Plate 22, figures 3 and 5, are two small ivory objects which are slotted at the base, recalling the two bird figures from the Hillside site. Figure 5 has the head carved realistically to represent an eider duck; the surface ornamentation consists of a few lines with spurs attached. In figure 3 the incised lines are applied rather carelessly ; the head is grotesquely human, the body is conical in shape. Plate 22, figure 6, is a miniature winged object of Punuk type (cf. pls. 68, 69) from cut 2, depth 20 inches. It is made from the sternum of a bird. Plate 22, figure 7, is an ivory ornament with a flat base and a slot at the center; the upper surface is convex and bears a simple decora- tion of lines, spurs, and short cross lines. The ivory wrist guard shown in plate 22, figure 10 (from cut 27, depth g in.) has a decoration consisting of pairs of lines with short spurs attached, which should probably be included in style 2. The bold, deeply cut lines of plate 22, figure 9, are likewise to be classed as style 2. This object, which was purchased from an Eskimo, is 96 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 flat on the under side, slightly convex above, and has a slanting hole evidently made to receive a prong; the small slot at the upper, forked end is probably for suspension, suggesting that the object was used as a hook of some kind. If we examine the distribution of the objects bearing Punuk decora- tion in table 1, it will be seen that by far the greater number of them (89 out of 107) come from the northwestern section of the midden, and that most of those from the southeastern section came from the upper levels. As to the relative positions of styles I and 2, we have also seen that practically all of the few examples of style 2 which were found at Miyowagh came from the surface layers of the northwest- ern section, indicating that it is the most recent of the several art styles present. The distribution of Punuk art is exactly the reverse of Old Bering Sea; the latter, it will be recalled, was the only style present at the Hillside site, while at Miyowagh it predominated at the southeastern section of the midden and at the northwestern section was found for the most part in the lower levels of the midden. The conclusion to be drawn is that the southeastern part of Miyowagh is the older, having been either in part contemporaneous with the adjoin- ing Hillside site or established soon after its abandonment. After a considerable interval of time, the Old Bering Sea art fell into decline and was gradually replaced by the simpler Punuk art, which is now found superimposed upon it in all sections of the midden. The ques- tion now arises whether Punuk art is a direct outgrowth of the Old Bering Sea or an importation from without. If the former hypothesis is to be upheld, it will be necessary to point out intermediate stages, and in fact there are several examples which might be regarded as affording evidence of such a transition. One of these is the knife handle shown in plate 22, figure 1, on which a typical Punuk 1 decora- tion of lines and dots is combined with rows of faintly dotted lines, which are just as definitely Old Bering Sea. The same is true of the harpoon head shown in plate 28, figure 15, where rows of broken lines accompany a typical Punuk 1 decoration. The needle case shown in plate 17, figure 9, is another example, this time of an Old Bering Sea pattern to which have been added Punuk dots and spurs. The ivory object shown in plate 22, figure 4, has another unusual combina- tion, a freehand circle accompanying a Punuk 1 decoration. This might in a way be considered as bridging the gap between the freehand circle of the Old Bering Sea period and the compass-made circle of the Punuk. It appears, therefore, that there was a period during the occupancy of Miyowagh when Old Bering Sea and Punuk motives were both in vogue. Probably if we had more material from just this NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS Q7 period, the nature of the transition would be clearer, but the fact that there are so few specimens on which the two art styles are combined is in itself evidence that it was a period of no long duration. There is, of course, no reason to believe that Punuk art had its origin at this particular site. We know that it was by no means limited to St. Lawrence Island, even though it seems to have been more prominent there than elsewhere. The existence of even a few specimens such as those just described suggests the possibility that future excavations may reveal a more gradual transition between the two styles. In this connection there will be other evidence to consider, especially harpoon heads ; as we shall see presently, these were undergoing constant modi- fication during the Old Bering Sea and Punuk periods, and there is ample evidence that in this case the transition from one stage to the other was quite gradual. HARPOON HEADS, OLD BERING SEA Of the many ingenious devices which have enabled the Eskimo to meet the needs of their exacting environment, none is of more im- portance than the harpoon. Without such an effective means of cap- turing the sea mammals on which they so largely depend, the Eskimo could hardly have adapted themselves to a life on the Arctic coasts. Fortunately for our present purpose, the harpoon is a complex im- plement, which differs from one locality to another; furthermore, archeological investigations have shown that throughout the course of its history it has been undergoing constant modification. Regional variations and developmental changes may be observed in various parts of the harpoon—in the ice pick at the butt end, the bone or ivory finger rest, the socket piece, the foreshaft, but most particularly in the detachable bone or ivory head. The toggle harpoon head is the most dependable criterion of cultural change at our disposal, and as such it is destined to bear the main weight of the chronology that must be established if we are to have a clear understanding of the stages of development in Eskimo culture. As a “ time indicator ’’, the harpoon head occupies a position in Eskimo culture analogous to that of pottery in the Southwest. To a peculiar degree it possesses the characteristics which Kidder has described as essential to any class of material that is to be utilized for diagnostic purposes. Valid cultural histories . .. . can only be compiled by studying series of speci- mens whose age, relative to each other, is known. Only from chronologically seriable specimens can dependable conclusions be drawn as to the origin and spread of culture traits and as to the improvement or decline of techniques. ... . But the time sequence of archeological materials is usually very hard to establish. 98 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 They are found in all sorts of places and under all manner of conditions. Very often they come from widely separated sites. They are collected in enormous quantities and in most bewildering variety. And, being undocumented, they are at first contact entirely undatable. To bring these into sequence, there must be selected certain classes of specimens which, more clearly and more easily than others, can be used as indicators of the passage of time, and which can serve as preliminary criteria for determining the relative age of the less readily seriated remains with which they are associated. Any class of objects to be employed in this way must possess certain indispensable characteristics. The specimens which go to make it up must be imperishable, they must be abundant ’ and widely disseminated, and they must be of such a nature as very sensitively to register cultural change. [Kidder, 1931, pp. 3-4.] Harpoon heads fulfill the above requirements admirably. They are abundant (the Gambell excavations yielded 417 specimens that could be identified as to type, in addition to many fragments) ; they are widely distributed, in a number of different forms, from northeastern Asia to Greenland; they are of durable material—ivory or bone— which is usually well preserved in the frozen soil; and most im- portant of all, they are complicated in form, and in Alaska, often elaborately ornamented. The relative abundance of harpoon heads in the middens is explained by the fact that this is the detachable part of the harpoon, the part that strikes and enters the prey, and which being subject to considerable stress and strain, is often broken, and later discarded. The harpoon head is complex in structure, possessing at least six features that often exhibit a wide range of variability: (1) the shaft socket at the lower or proximal end for engaging the end of the foreshaft; (2) the lashing slots or other means whereby (in those with open sockets) the foreshaft is held in the socket; (3) the pointed lower end, or spur; (4) the central perforation, or line hole, by means of which the harpoon head is attached to the line; (5) the presence or absence of lateral barbs or inset stone blades; (6) the anterior end which may or may not be slit for the purpose of holding an end blade. In addition there are certain combinations of these features to be noted, and other characters such as size, pro- portion of parts, and technique, all of which may and often do have a diagnostic value. A further fortunate circumstance is that during the Old Bering Sea and Punuk periods harpoon heads were decorated more frequently than any other class of artifacts, and in such a way that there was often a very definite correlation between the form of the harpoon head and the ornamentation it received. For convenience of reference it will be necessary to arrange our harpoon heads according to some definite order. We have the choice of two methods: first, to establish an exact, comprehensive classifica- NOTE ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 99 tion by letters or numbers, such as Mathiassen has used in describing the harpoon heads of the Thule culture ; or second, merely to describe and give an appropriate name or number to each type. As stated above, Alaskan harpoon heads are complex, possessing at least six variable features, the inclusion of which alone would require cum- bersome formulae for the designation of the types. At that, the features included would be only those which would correspond in a biological sense with the more fundamental divisions—class, subclass, order, family, subfamily, and genus; a still further division would sometimes be necessary to include the specific, subspecific, and varietal features on which a demonstration of close genetic relationship would depend. In view of this condition, I have therefore, after fully testing both methods, chosen the second, that of merely describing and nam- ing the various types, with numerals and letters used as sparingly as possible. Eventually, when systematic excavations have revealed the sequence of harpoon types at more places and when homologies can be clearly distinguished from analogies, it should be possible to estab- lish a comprehensive classification in which the various regional forms can be arranged in the order of their development. At the present time this cannot be done for the entire Eskimo area, not only because we have too few series for which chronological relationships have been definitely established, but also because the original forms remain to be determined. Thus we have no knowledge whatsoever of the pre- sumably simpler forms antecedent to the complicated harpoon heads of the Old Bering Sea culture. The excavations at Gambell produced abundant evidence of a num- ber of stages of development through which harpoon heads passed, beginning with the complicated and numerous forms of the Old Bering Sea culture and ending with the one simple form now in use on St. Lawrence Island (text fig. 24, opposite p. 216). At the Hillside site 16 harpoon heads were found which were suffi- ciently complete for type identification; 14 of these are shown in plate 23. Among them were examples of the several Old Bering Sea types already known, in addition to other types previously unknown. From Miyowagh there are 164 harpoon heads, more or less complete, in addition to an even larger number of fragments. We have at our disposal, therefore, a large body of material, fully documented as to depth and locality, upon which to base our study of the harpoon heads of the Old Bering Sea and early Punuk periods. The harpoon heads are divided below into two groups, those with open sockets and those with closed sockets. The types are indicated by roman numerals; subtypes by letters (a, b, c, etc.) in parentheses ; 100 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 and the relation of line hole and blade by the letters x and y, the former indicating that the line hole and blade are parallel, the latter that they are at right angles. The types in each group that are de- scribed first are those which were found only at the Hillside site or in the southeastern section of the next site, Miyowagh; these are, therefore, the Old Bering Sea types. Following these are the few early Punuk types, found mainly at the later, northwestern section of Miyowagh. Later, as we describe the material from Jevoghiyogq, Seklowaghyaget, and the recently abandoned ruins adjoining the present village of Gambell, the sequence will be continued to include the various Punuk, protohistoric, and modern types. The distribution of the various types of harpoon heads according to depth and site is given in table 2 (p. 216). Text figure 24 (p. 216) is a graphic arrangement of the principal open socket harpoon types at Gambell, showing developmental stages from Old Bering Sea to modern times. Open Socket Harpoon HEaps Open socket type I «—Two lashing slots; at proximal end a sym- metrical median spur, trifurcated, with center prong longer than the two adjoining; two line holes, one above and one below the socket, connected on the side opposite the socket by a deep groove into which the lashing slots open; two deep, narrow pits or sockets, parallel with the line holes, for holding the stone side blades; no end blade; oval to almost round in cross-section; decorated (Old Bering Sea) or plains (Pl. 23) fies) 15/2) and text-fig.' 24.) Two harpoon heads of this type have been described previously, one by Jenness from Little Diomede Island and one by the writer from northern Alaska, exact locality unknown (Jenness, 1928 a, pl. 13, a; Collins, 1929, pl. 2, f—the blade slots and one of the line holes incomplete). Three examples were found at the Hillside site, two of which are illustrated in plate 23, figures 1, 2; both were found among the rocks at the south end. Plate 23, figure I, is complete except for the tip; the thin partition between the blade sockets has broken through. The two line holes are drilled at an angle, converging slightly as they enter the deep connecting groove on the back side. The lower end just below the socket has been hollowed out smoothly to a depth and width that exactly fits the end of one’s thumb. The lower end of plate 23, figure 2, is broken and the upper end has been reworked into a drill or reamer; the side blades of jasper, parallel with the line holes, are still firmly in place. The decoration, as in figure 1, is Old Bering Sea. NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS Io! The third, fragmentary, head of this type, from the upper level of the midden, is decorated in the same style. From Miyowagh there is a single fragmentary example, undecor- ated, from cut 18, depth 53 inches; and an unfinished specimen from cut 27, depth 44 inches. (From cut 7, 48 in. deep, there is the lower end of a harpoon head with Old Bering Sea decoration that is unique in having the line hole (originally two ?) in the plane of the spur. The latter is trifurcated and synimetrical, the socket wide and deep with two lashing slots. This is the only instance I have seen of a sym- metrical trifurcated spur which was not at right angles to the line hole or holes.) Open socket type I y—Same as above except that it is a flattened oval in cross-section, and has the side blades at right angles to the line holes; decorated (Old Bering Sea) or plain. (Pl. 23, fig. 3; pls. 24 and 25, figs. 1, 2; pl. 25 shows the socket side of the same harpoon heads figured on pl. 24.) Three harpoon heads of this type have been described previously, one by Wissler from Cape Smythe near Point Barrow, one by Jenness from Little Diomede Island, one by Mathiassen from northern Alaska (Wissler, 1916, p. 410; Jenness, 1928 b, fig. 3, c; Mathiassen, 19209, ne. 23, D.). No finished examples were found at the Hillside site—only the un- finished one shown in plate 23, figure 3, from the second level of the midden, and one other from among the rocks at the south end. The beginning of the groove into which the two line holes would have opened is visible on the side opposite the socket ; on the opposite side the lower end has been slightly hollowed and beveled in preparation for a trifurcated spur. From Miyowagh there are three specimens (pl. 24, figs. 1, 2) and an undecorated fragment, from cut 18, depth 72 inches. Plate 24, figure 1 (cut 7, depth 75 in.), is almost straight-sided with a rather wide, blunt end; the side opposite the socket is almost flat, whereas in the other known examples it is gracefully rounded. The side blades, of chipped slate, are also placed nearer the tip than is customary. Plate 24, figure 2 (cut 23, depth 65 in.), is undecorated and small—only 6.6 cm long; the blades are of chipped jasper; the groove at the lower end was for a lashing which was applied after a split occurred between the central spur and lower line hole. Open socket type I (a) y—Same as type I y except that there are two pairs of small barbs, at right angles to the line holes, in addition to the side blades. (PI. 24, fig. 3.) 8 102 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 This type is represented by the single specimen shown in plate 24, figure 3. It is from cut 18, depth 78 inches. The sockets for the side blades are placed between the two barbs. Open socket type I (b)—(in part hypothetical.) Two lashing slots; two line holes, same arrangement as above; no blades; pair of small barbs at tip, at right angles to line holes; form of spur un- known, probably like above. (PI. 24, fig. 4.) A single specimen is known, that shown in plate 24, figure 4. It is from cut 23, depth 81 inches; the spur is broken, so this feature is tentative. Open socket type II «—Two lashing slots; at proximal end an asymmetrical, lateral spur, trifurcated, with sharp slightly curving prongs all of the same length; single line hole, below which is usually a groove into which the lashing slots open; two side blades, parallel with the line hole; oval to almost round in cross-section; decorated (Old Bering Sea) or plain. (PI. 24, figs. 5-8.) None were found at the Hillside site. From Miyowagh there are six examples. Four are from the older, southeastern section, as fol- lows: cut 23, 14 inches deep (pl. 24, fig. 7); cut 9, depth 24 inches (pl. 24, fig. 5); cut 29, 25 inches (pl. 24, fig. 6) ; cut 15, 42 inches. Two are from the northwestern section: cut 4, depth 36 inches (pl. 24, fig. 8) and cut 19, 63 inches. A striking feature of these harpoon heads is the marked asymmetry of the spur, which is placed partly below and partly to one side of the socket. The spur is definitely trifurcated, with the center prong straight and the two outer ones curved ; the prongs all have sharp points. Plate 24, figure 5, has one of the side blades, of chipped jasper, remaining in place. The slot extends through to the other side as is also the case with plate 24, figure 6, and one other. In plate 24, figures 7, 8, and one other, the blade sockets do not quite meet, a thin partition remaining at the center. Plate 24, figure 8, which was found in cut 4, 36 inches deep in the later, northwestern section of the midden, differs from the others in having one of the prongs of the spur set off from the other two by a deep straight groove which extends beyond the line hole. Figures 5 and 7 also have small grooves between the prongs of the spur, but figure 6 has no suggestion of these, the upper part of the spur being perfectly smooth. In the latter specimen the groove which extends downward from the line hole takes a slanting path, whereas on the others it is straighter. Morphologically, harpoon heads of this type appear to be derived from type I x: the side blades, parallel with the line hole, are re- tained as is also the trifurcated spur, although it has become asym- NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 103 metrical. The stratigraphic evidence also indicates that it is actually a later form than either type I x or I y; for whereas the latter were found only at the Hillside site or at considerable depths in the old section of the Miyowagh midden, those of type II x all come from Miyowagh, and mostly from the upper levels. Open socket type II y—Same as above but with the side blades at right angles to the line hole; flattened oval in cross-section. (PI. 24, figs. 9, 10.) This type was not found at the Hillside site, and there are only four examples from Miyowagh. Two of these are shown in plate 24, figures 9 and 10. The other is the decorated fragment shown in plate 15, figure 10, from cut 3, in the later, northwestern, section of Miyowagh, but at a depth of 54 inches. The spur, which is damaged, is definitely trifurcated, and there are two lashing slots. Although the upper end is broken, parts of two sockets for side blades, at right angles to the line hole, are still clearly visible. Plate 24, figures 9 and 10, from the southeastern section of Miyowagh (cut 23, depth 58 in., and cut 9, depth 36 in., respectively) and the one not illustrated (from cut 5, 41 in. deep) are of particular interest because they rep- resent the first appearance at Gambell of harpoon heads that embody the essential features of the Birnirk type, predominant at the older sites around Point Barrow. They all have the trifurcated, asym- metrical spur, the two side blades at right angles to the line hole, and the general shape, long and narrow, flattened oval in cross-section— features which are characteristic of the Birnirk heads, although as a rule the latter have only one side blade, with a sharp, prominent barb opposite, instead of two side blades. As we shall see later, harpoon heads of this generalized Birnirk type occur with somewhat greater frequency in the Punuk stage, and resemble the Birnirk heads of the Arctic coast in the further particular of being made of bone instead of ivory. Added evidence of the relationship of plate 24, figure 10, to the Birnirk group is shown in the decoration, which consists of two straight lines down the sides, one of them forking at the line hole, and a V-shaped depression just above the line hole. Bueemesocke: types ll (a) 4; JI, (a) y; ID (bo) 42; dT (0). 4; II (c) «; II (c) y—Under the above headings come a number of harpoon heads all of which have bifurcated spurs or side blades, or both. They are all derivatives of types II x and II y, but since they do not appear until fairly late—well into the Punuk period—their description will be deferred until we come to deal with the Punuk material. 104 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 Open socket type III «—Two lashing slots; a prominent, asym- metrical spur which is trifurcated, bifurcated, or otherwise irregular ; single line hole, below which there is usually a groove into which the lashing slots open; end blade, parallel with the line hole; usually small barbs along the edges; oval in cross-section; decorated (Old Bering Sea or early Punuk) or plain. (PI. 23, fig. 4; pl. 24, figs. 11-23.) Four examples described previously, one by Mathiassen from Bering Strait, and three by the writer, two from northern Alaska, exact locality unknown, and one from Kukuliak, St. Lawrence Island (Mathiassen, 1929, fig. 20; Collins, 1920, pl. 2, a-e). Three heads of this type were found at the Hillside site, one of which is shown in plate 23, figure 4. Half of the tip is broken and the remaining half reworked into a reamer. The spur is formed of one long, pointed prong to which is attached another, shorter and squared off at the end. Plate 23, figure 7, from house no. 1, exhibits some of the features of type III x but is not assigned thereto because of its small size (6.1 cm long) and its flat, squarish sides. Neither the line hole nor the socket has been completed. Type III x is one of the most characteristic forms at Miyowagh, and one which is of particular importance because of the evidence it affords of the transition from the Old Bering Sea to the Punuk. It is represented by 25 examples. Twenty-one are from the older, south- eastern section of the midden, distributed as follows: Cut 7, 62 inches (pl. 24, fig. 20); 67 inches (pl. 24, fig. 16); 92 inches (ply2z4emiee II); 96 inches (pl. 24, fig. 13). Cut 9, 12 inches; 24 inches (pl. 24, fig. 23). Cut 9a, 18 inches (2) ; 26 inches (pl. 24, fig. 14) ; 36 inches (pl. 24, fig. 18)-; 48 inches (pl. 24, fig. 15). Cut 18; 18 amches;e29 inches ; 38 inches; 48 inches (2) (pl. 24, fig. 17). Cut 23, 14 inches (pl. 24, fig. 21); 22 inches. Cut 24, 62 inches. Cut 271@ memes 29 inches (pl. 24, fig. 12). Four are from the northwestern section of the midden, from the following cuts and depth: Cut 1, 30 inches (pl. 24, fig. 22) ; 53 inches. Cut 3, 20 inches (pl. 24, fig: 19) Gur 55 inches. The harpoon heads of this type display considerable variability, ranging from those of large size, with a very prominent divided spur and an elaborate Old Bering Sea ornamentation (pl. 24, figs. 11-16) to those of smaller size in which the spur is simpler, though still ir- regular, and in which the ornamentation is reduced to the simple lines of the early Punuk (pl. 24, figs. 19-21, 23). There is clear strati- graphic evidence that these smaller and simpler heads are later than the others, but since there is no sharp dividing line between them typologically, they are here all grouped together. NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 105 The bone harpoon head shown on plate 24, figure 11, differs from others of the group in having the spur divided into three sharp prongs of equal length, a feature which is characteristic of types II x and II y. In plate 24, figure 12, the spur is also trifurcated, but in an entirely different manner. It is essentially a lateral spur, for its most important part—the long center prong—continues straight down the side, at right angles to the line hole, even though the smaller prongs curve inward toward the socket. One of these is shorter than the other, thereby producing an asymmetry of the spur; otherwise it is very similar to a type of spur found on closed socket harpoon heads of the Old Bering Sea period from the Diomede Islands, in which the two smaller prongs are of equal length (pl. 27, figs. 5-7). The spur of plate 24, figure 13, although broken, seems to have been exactly similar to that of figure 12. In figure 15 the spur is of the same general shape, although there is only one projection from the main prong. Figures 14 and 16 are examples of the most prevalent form of spur associated with harpoon heads of this type. The spur is flatter than those just described, extending beyond the mid line and to some extent beneath the socket; it is bifurcated, with the inner prong very prominent. Figures 17 and 18 exhibit a variation of this form of spur, in which the inner prong is reduced in size. In figures 21-23, through a process of simplification, the elaborately divided spurs have been reduced to a point where they are more appropriately de- scribed as merely “irregular”. That this typological distinction has also a chronological significance is seen in the fact that these three specimens and the three similar ones not illustrated are all from the upper levels of the midden at depths of 14, 16, 22 (2), 30 and 38 inches. In figure 20 a rounded median elevation produces a thicken- ing of the spur which is unusual. The orientation of the spur, whether to the right or left of the socket, seems to have no significance, about the same number of speci- mens having one arrangement as the other. Very often harpoon heads of type III x have small pointed pro- jections or barbs along the edges. In most cases these are small and distinctly ornamental (e.g. pl. 24, figs. 11-13) ; sometimes, however, they are large enough to have served as real barbs (e. g., figs. 14, 17, 18) even though they still form an essential part of the decora- tive scheme. The position of these small barbs is invariably along the edges at right angles to the line hole, and they are present only on those harpoon heads in which the line hole and blade slit are parallel. Most of the harpoon heads of this type which are relatively large, with an elaborately shaped spur, are decorated in Old Bering Sea 106 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 style (pl. 24, figs. 11-16), and these same heads, it will be recalled, came mostly from the deeper parts of the midden; figures 17 and 18, also ornamented in Old Bering Sea style, had less prominent spurs. Those of smaller size, with the spur simplified but still irregular (pl. 24, figs. 20-23), and which according to stratigraphic evidence are definitely later, are either plain or decorated in early Punuk style. Here, instead of an elaborate all over decoration, we have only simple lines, which, however, were placed in exactly the same positions as those lines which on the older specimens had served to divide the surface decoration into definite panels or fields. In both cases a pair of lines, beginning at the base of the blade slit, takes a downward, diverging course, passing around the line hole and terminating on both sides at a small barb above the spur. Similar lines, beginning nearer the tip, emerge just above the other pair of barbs, which are thus accentuated and set apart. In the Old Bering Sea heads, the areas between these dividing lines are filled in with more or less elaborate designs ; in the early Punuk heads they are left plain. The homology may also be traced in another feature, namely the pairs of elevated circles or ellipses, suggestive of “eyes”, which form so prominent a part of the Old Bering Sea ornamentation. In the early Punuk heads these elevations are still present, although less prominent, and they occupy the same position, on either side of the line hole. AlI- though as a group these smaller and simpler heads are later than the more elaborate forms, they appeared at a rather early stage, for one of them, plate 24, figure 20, comes from a depth of 62 inches, and two others, not illustrated, which have the same simple ornamentation (although they are larger and have prominent bifurcated spurs) were found at depths of 48 and 62 inches. Plate 24, figures 22 and 23 (cut I, 30 in. deep, and cut 9, 24 in. deep, respectively) represent the latest of the several forms included in type III x. In fact, it is only by virtue of their irregular spurs that they are included in this group. They have lost the smooth, flowing contour and the graceful, curving barbs which characterize such intermediate forms as those shown in plate 24, figures 19-21. Furthermore, the decoration in figure 23 is that of phase 2 of Punuk style 1 (lightly incised lines with small spurs attached—in this case with the spurs connected) a style which becomes quite common at a somewhat later period. Open socket type III y.—In general, similar to type III x except that the blade slit is at right angles to the line hole. The spur is bi- furcated or irregular, and the barbs along the sides inconspicuous or absent ; decorated (Old Bering Sea or early Punuk) or plain. (PI. 23, figs. 5, 6; pl. 26, figs. 1-2, 4-10.) WO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 107 Two heads of this type were found at the Hillside site (pl. 23, figs. 5, 6). Figure 5, from house no. 1, is unusual in having a series of four sharp barbs bordering the blade slit on one side. The un- finished head, figure 6, from the upper level of the midden, was clearly designed to have had the blade slit at right angles to the line hole. The latter was made in an unusual manner: the hole first drilled was not well centered, and evidently to correct this, the hole was filled with a cylindrical ivory plug (not a drill core) and another hole drilled adjoining it at the center. The broken specimen, figure 8, from house no. 2, is difficult to classify. The spur, while not symmetrical, lacks the marked asym- metry of type III. Instead of two lashing slots, it has a single slot with an opposite groove. Decoration consists of short detached lines, carelessly incised. On the whole it seems best to defer classification of this particular specimen until additional examples have come to light. From Miyowagh we have 12 examples of type III y. Eight of these are from the older, southeastern section, distributed as follows: Cut 7, depth 16 inches (pl. 26, fig. 4). Cut 9, 48 inches (pl. 26, fig. 7). Cut 9a, 36 inches (pl. 26, fig. 1). Cut 18, 48 inches (pl. 26, fig. 5). Cut 23, 58 inches ; 81 inches (pl. 26, fig. 2). Cut 24, 13 inches (pl. 26, fig. 10) ; 54 inches (pl. 26, fig. 3). Four come from the later, northwestern section, as follows: Cut 6, 26 inches (pl. 26, fig. 6). Cut 19, 24 inches (pl. 26, fig. g) ; 51 inches (pl. 26, fig. 8). Cut 20, 22 inches. The range of variation parallels that of type III x, and includes heads of both Old Bering Sea and early Punuk age. The earlier ex- amples are large, with prominent divided spurs and typical Old Bering Sea ornamentation (pl. 26, figs. 1-3). The later forms are reduced in size, have simplified spurs which, however, are still irregular, and ornamentation when present is early Punuk (pl. 26, figs. 6-10). The Old Bering Sea heads had prominent “ eyes” formed of concentric circles which occupied rounded elevations on either side of the line hole, on the spur, and sometimes near the tip. In the early Punuk heads these, along with the other elements of Old Bering Sea art, are absent, although the idea survives in the slight elevations around the line hole which are sometimes accentuated by having a small circular pit at the center. In figure 8 these pits are filled with small cylindrical plugs of baleen. A characteristic feature of the harpoon heads of type III y and of most others which have the blade slit at right angles to the line hole is the prominent median ridge which extends downward from the tip to the line hole. 108 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 Plate 26, figure 3, has a secondary blade slit, parallel with the line hole, made after the upper end had broken; originally the blade slit was at right angles to the line hole. Figure 5 is unfinished and was probably discarded because the line hole was not well centered. Types III x and III y are with a few minor exceptions the last of the open socket harpoon heads to be decorated in Old Bering Sea style; on the other hand, some of the specimens of these types are already decorated in early Punuk style. The latter (pl. 24, figs. 20-22; pl. 26, figs. 6-10) are the forerunners of types III (a) x and III (a) y, which above all others are characteristic of the Punuk stage. These will be described later with the other Punuk material. Mean- while, we will consider the closed socket types, beginning with those which were found at the Hillside site and at the older section of Miyowagh. CLOSED SocKET HArpoon HEApS Closed socket type I x«—At proximal end a plain, symmetrical, median spur; two line holes, close together, connected on the back side by a deep groove; end blade parallel with line holes (pl. 26, fig. II). This type is represented by the single specimen from Miyowagh (cut 18, depth 48 in.) shown in plate 26, figure 11; it has the appear- ance of being unfinished. The combination of a symmetrical median spur and double line holes connected by a deep groove is in itself suggestive of age, in view of the fact that these features are found only on the oldest of the open socket harpoon heads. Hence the desig- nation of this as one of the earliest closed socket forms even though it was evidently an unimportant one. Closed socket type II «—A lateral trifurcated or irregular spur; single line hole; end blade, parallel with line hole; small, ornamental barbs on edges; very thin in cross section; decorated (Old Bering Sea) or plain. (Pl. 23, figs. 11, 12.) There are only two examples of this type, one from the Hillside site, the other, exact provenience unknown (pl. 23, fig. 12). Plate 23, figure 11, from house no. 1, has an irregular spur similar to that of figure 6. A row of small barbs, four on one side, two on the other, gives the edges a serrated appearance. The most striking feature of this type of harpoon head is its extreme thinness, which is unusual for a closed socket type. Figure 12 was obtained from the Eskimos at Gambell by Dr. Riley D. Moore in 1914, and while its exact proveni- ence is uncertain it is included here since it is the only other known example of the type. Originally there were small barbs along the NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 109 edges as in figure 11, but these were destroyed when the specimen was filed down and polished by the Eskimos from whom it was pur- chased. The spur is trifurcated and seems to have been symmetrical. The rectangular slots near the middle were for the purpose of lashing the two sides together after a split had appeared. Vestiges of Old Bering Sea ornamentation still remain. Closed socket type III x+.—Similar to the above, but thicker in cross- section ; the lateral spur is trifurcated and symmetrical, with the center prong longer than the other two. (PI. 27, fig. 5.) Closed socket type III y—Same as type III x except that it has the blade slit at right angles to the line hole. (PI. 27, figs. 6, 7.) These two types are based on four harpoon heads from Little Diomede Island. The fact that none were found at Gambell may be due to accident, for several specimens from the Hillside site and Miyowagh have spurs which approach this type, e. g., plate 23, figure 12, and plate 25, figure 12. Furthermore, the symmetrical trifurcated spurs of these Diomede heads, although placed to one side of the socket, are essentially similar to the median spurs of our open socket types I x and I y. In view of these circumstances it seems permissible to reserve for these two types the designations here given, on the chance that they will eventually be found on St. Lawrence Island. Closed socket type IV y.—Straight, symmetrical spur; round line hole, beveled below ; end blade at right angles to line hole; upper end tapering and conical, without the longitudinal ridge which is usually present when the blade slit is at right angles to the line hole; almost round in cross-section. (Pl. 23, fig. 14; pl. 26, fig. 12.) There are only two examples of this type—plate 23, figure 14, from among the rocks at the south end of the Hillside site, and plate 26, figure 12, from Miyowagh, cut 18, 40 inches deep. Although in its main features this type conforms to type V y, it possesses certain minor but significant characters which set it apart. First of all, its shape is unusual—perfectly straight and smooth from tip to spur— whereas in all of the types previously considered there were surface elevations and depressions and usually some degree of a flare either along the edges or at the spur. Harpoon heads like plate 26, figures 16 and 17 (type V y), which most closely approach the present type, differ in having a more or less pronounced longitudinal ridge between the line hole and tip in addition to well-defined elevations, usually surmounted by “ eyes’, around the line hole. None of the other closed socket types have the smoothly rounded contour, and particularly the tapering conical upper end of the present type; these features are found elsewhere only in the open socket type I x. Finally, in these I1o SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 two examples of type IV y, the application of the design differs from that on all of the other harpoon heads. In both cases the decoration extends obliquely around both sides, enclosing the line hole. In plate 23, figure 14, the ornamentation is Old Bering Sea style 2 and is applied as a band which begins at about the middle of the blade slit on the side opposite the spur and extends downward obliquely to the tip of the spur. The ornamentation in plate 26, figure 12, has the appearance of being unfinished; it consists of only a few lines, but again applied obliquely, this time beginning near the upper end on the same side as the spur and extending downward to terminate be- neath the line hole. The narrow rectangular slot opening into the shaft socket beneath the line hole is a feature frequently found on closed socket harpoon heads; it may have been for the purpose of facilitating the drilling of the socket. Closed socket type V «.—A single, symmetrical spur ; line hole usu- ally round, and beveled below; end blade parallel with line hole; flattened oval in cross-section; decorated (Old Bering Sea or early Punuk) or plain. (Pl. 23, fig. 13; pl. 26, figs. 13-15.) This form of the closed socket harpoon head was not used very extensively during the Old Bering Sea period. Two examples, with Old Bering Sea ornamentation, have been described previously (Col- lins, 1929, pl. 9, c, d, possibly b). From the Hillside site we have only two examples, plate 23, figure 13, and one other undecorated head from among the rocks at the north end. Plate 23, figure 13, from among the rocks at the south end, is decorated in Old Bering Sea style 3, that is, the ornamentation centers around two pairs of raised ‘eyes’ placed on either side of the line hole. The decoration on the spur has been obliterated by secondary cutting. In the groove beneath the line hole is a small slot which opens into the socket, similar to those in plate 26, figures 12 and 19. From Miyowagh there are 22 harpoon heads of this type, the greater number of which are clearly of Punuk age as they are mostly from the later, northwestern section and are decorated for the most part in early Punuk style. There are only four with Old Bering Sea orna- mentation (pl. 26, figs. 13-15, and one other poorly preserved speci- men from cut 6, 26 in. deep). Figures 13 and 14, from cut 13, depth 60 inches and cut 21, depth 24 inches, respectively, are remarkable for their small size, which as we shall see later, is also a distinguish- ing characteristic of the Punuk heads of this same type. The decora- tion in figure 14 is rather simple, and it might be questioned whether itis Old Bering Sea, consisting as it does of nothing more than straight lines bordering the blade slit and a single transverse connecting line. NO: I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS III However, we have already seen that a design similar to this—pairs of short cross lines between straight lines which extend downward from tip to line hole—was found at the Hillside site (pl. 12, fig. 8). The ornamentation in plate 26, figure 13, is typically Old Bering Sea, with finely incised straight lines, broken lines and “eyes” formed of pairs of concentric, nucleated circles surmounting rounded eleva- tions on either side of the line hole. On the spur are two independent design elements, small concentric oval figures, placed transversely. On both of these harpoon heads the groove beneath the line hole takes an oblique path. Plate 26, figure 15 (cut 25, 39 in. deep) has a piece broken off the spur, and the upper end is broken and reworked into a reamer ; the blade slit has been completely obliterated, but the con- figuration of the remaining upper end shows that it must have been parallel with the line hole. The decoration is the simplified Old Bering Sea style 3, consisting mainly of three (originally four?) pairs of opposed “eyes” placed on rounded elevations. The Punuk heads of type V x will be described later. Closed socket type V y—Same as type V x except that the blade slit is at right angles to the line hole. (PI. 26, figs. 16-20.) This form of type V harpoon head was used more frequently during the Old Bering Sea period than was the preceding type, which had the blade slit parallel with the line hole. Six heads of this type, decorated in Old Bering Sea style, have been described previously (Mathiassen, 1929, fig. 13, a; Collins, 1929, pl. 1, a-b, c, d, e-f ; Mason, 1930, pl. 5, fig. 1). None was found at the Hillside site. From Miyowagh there are 21 examples, 10 of them decorated in Old Bering Sea style, 4 in early Punuk, 1 in late Punuk, and 6 un- decorated. The 10 Old Bering Sea heads all come from the south- eastern part of the midden as follows: cut 7, 33 inches; cut 8, 40 inches (pl. 26, fig. 19) ; cut 9 a, 38 inches (pl. 26, fig. 18) ; cut 15, 60 inches ; cut 16, 37 inches; cut 18, 60 inches; 66 inches (pl. 26, fig. 16) ; 72 inches ; cut 24, 13 inches (pl. 26, fig. 20) ; cut 25, 61 inches (pl. 26, fig. 17). In addition to these there are two undecorated heads from cut 18, 60 inches deep, and cut 21, 46 inches deep, which are probably of Old Bering Sea age. One of the most striking features of these harpoon heads, and one which often adds greatly to the symmetry of form and decoration, is the longitudinal ridge between the line hole and tip. On the small specimen, plate 26, figure 17, this ridge is less noticeable than on any of the others; in figure 19 it is most pronounced. On the latter, and frequently on others, the ridge is outlined by rather deeply incised EZ SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 bordering lines, and being left undecorated, stands out in effective relief between the decorated panels which flank it on either side. The type V y heads from Miyowagh which belong to the Punuk stage will be considered later. This brings to an end our study of the harpoon heads of the Old Bering Sea period. We have seen that the oldest heads, those from the Hillside site and from the lowest levels of the Miyowagh midden, are more complicated in form and more elaborately ornamented than others which came into use at Miyowagh at a later period. Thus, of the most complicated form of all (open socket types I x and I y, with trifurcated spurs, double line holes, and side blades) 10 ex- amples were found, 5 of them from the Hillside site and 5 from the lower part of the older, or southeastern, section of Miyowagh. Al- though the double line hole was discontinued at a very early period, divided spurs and side blades were retained for a much longer time. The first important derivative forms seem to have been types II x and II y, which have only one line hole but which continue the old associa- tion of side blades and trifurcated spurs. The spur, however, has become asymmetrical and more or less lateral in position instead of symmetrical and median as previously. Types II x and II y, in turn, passed out of use at a relatively early period, although the tradition of side blades and divided spurs (but bifurcated instead of trifurcated) was carried on into the Punuk stage, as we shall have occasion to observe later. The rest of the open socket harpoon heads from the Hillside site and the older part of Miyowagh fall into two groups—types III x and III y. Both of these types display a wide range of variability, which, however, is expressed in developmental changes within defined limits, the basic features remaining constant. The stratigraphic evi- dence indicates clearly enough that the observed variations are truly developmental in nature, reflecting definite stylistic trends over a con- siderable period of time. Thus the types III x and III y heads that are stratigraphically the oldest (pl. 24, figs. 11-18; pl. 26, figs. 1-3) are characterized by relatively large size, prominent divided spurs, and an elaborate Old Bering Sea ornamentation. Through a process of simplification these features became gradually reduced, with the result that the later harpoon heads of these types are of smaller size, have spurs which are much simpler although still irregular, and are either undecorated or bear the simple ornamentation of the early Punuk (pl. 24, figs. 19-21, 23; pl. 26, figs. 6-10). Through still further changes in the same direction the latter in turn developed into types III (a) x and III (a) y, the predominant types of the Punuk period, to be described later. NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS II3 Closed socket heads were much less common during the Old Bering Sea period than those with open sockets. Types I-IV, represented by only a few specimens, played an unimportant part as compared with type V. However, this may not have been the case everywhere, for type III (which was not found at Gambell) seems to have been fairly common on the Diomedes. The predominant closed socket types at Gambell are types V x and V y. The former, with the line hole and blade slit parallel, was of less importance than the latter. The most striking feature of this type is its small size. On the other hand, type V y, with the line hole and blade slit at right angles, displays the same range of variation in regard to size and ornamentation as was shown by the open socket types III x and III y. Thus the older specimens, with Old Bering Sea decoration, are for the most part large; while the younger forms (to be described later), which are either plain or decorated in early Punuk style, are considerably smaller. One important distinction between the open and closed socket heads of the Old Bering Sea period is in the form of the spur. In the great majority of cases the open socket types have elaborate, divided spurs, which have been described above as trifurcated, bifurcated, or irregular ; in fact, there are only three examples that have single spurs, plate 23, figures 9, 10, and one other from Miyowagh, cut 25, depth 18 inches. Heads of this type, with single rather than divided spurs, will be described later under the designations of types III (a) x and III (a) y. When we come to the closed socket heads, however, we see that the single, lateral spur is the prevailing form. Types V x and V y, comprising more than 90 percent of all the closed socket heads, have only this type of spur; and the same is true of the less common types, IV y and I x, although in the latter the spur is placed medially instead of laterally. The only closed socket types possessing divided spurs are types II x, III x and III y, the last two being known as yet only from the Little Diomede Island. In conformity with the extensive use of ivory during the Old Bering Sea period is the fact that the harpoon heads were made almost ex- clusively from this material. Only four harpoon heads in the entire series are made of bone, one of these being the specimen shown in plate 25, figure 11. In the Punuk period, as we shall see later, bone harpoon heads become somewhat more common, although the use of this material is restricted almost entirely to forms which are related in one way or another to the Birnirk type. Almost invariably the harpoon heads were made from the tips of very small, young walrus tusks. This is clearly indicated by the fact that in practically every specimen a very small core of dentin, less II4 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 than I cm in diameter, is visible at the center, surrounded by enamel, whereas the average adult walrus tusk has a dentin layer three or four times this size. Most of the harpoon heads were apparently made of young tusks from about 12 to 20 cm long. Further evidence that “ baby ” walrus tusks were used in this way is shown by the fact that on many of the harpoon heads the inner side of the basal spur still shows the original hollow of the proximal end of the tusk, with an inner diameter of less than I cm as compared with the very much larger hollow base of an adult tusk. With regard to technique, several points should be noted. The line holes on all of the harpoon heads which we have been considering were made by drilling and are perfectly round. The edges were usually trimmed down and smoothed, and there is usually a rounded depres- sion or groove extending downward from the line hole; in closed socket heads the groove is usually present on both sides, in open socket heads only on the back side at the place where the lashing slots emerge. In the closed socket forms the socket is invariably made by drill- ing and consequently is cylindrical. An unexplained feature is the occasional presence of a narrow rectangular slot cut through to the socket from one or both sides (pl. 23, fig. 13; pl. 26, figs. 12, 19), possibly for the purpose of facilitating the drilling. On open socket heads the sockets are rectangular and were made by cutting. In the older examples there is a marked tendency for the socket to be relatively shallow with a somewhat concave bottom and slightly flaring sides. These features may to some extent be correlated with a relatively increased size, particularly breadth, but they are also found on the smaller Old Bering Sea heads, such as plate 24, figures 2, 5-7, 11. On the other hand, the smaller early Punuk types have sockets which are distinctly narrower and deeper, with flat bottoms and vertical sides. In making the socket one of the first steps was the cutting of a small transverse slot just below the line hole. This formed the upper end of the socket and was often deeper than the floor of the socket. One of the most interesting features of the open socket heads from a technological standpoint is the lashing arrangement—the narrow rectangular slots on either side of the socket through which passed the strands of baleen or sinew which held the end of the foreshaft in place. On the older heads that have lateral spurs the two slots are usually of unequal length, the one on the side of the spur averaging Io to 13 mm, the one opposite the spur 8 to 10 mm; on the later heads the slots are usually of equal length, averaging around 10 mm. On those heads which have median spurs the lashing slots are invari- NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 115 ably equal in length. Most remarkable, however, is the extreme nar- rowness of these slots, usually a little less than 2 mm; the vertical length, or height, averages around 8 mm. The cutting of such a long, narrow slot must have been difficult, and it would be of interest to know just what kind of tool was employed. In several instances closed socket heads were found which, after breaking, had been provided with a groove and used in the manner of the open socket type. The position of the lateral spur on open socket heads, whether to the right or to the left, seems to have been a matter of choice, some- what more than half of them having the spur on the right side. It was somewhat surprising that among the great number of har- poon heads from Gambell which had been provided with end blades of stone, only one had the blade still in place. Side blades, on the other hand, being firmly wedged in deep slots, were often found in place. The slit for the end blade was rather wide, averaging on the whole around 2 mm, some being as wide as 3 mm. This width of the blade slit, together with the great number of complete or frag- mentary slate blades which were found, shows clearly that stone was the only material used for harpoon blades. If iron blades had been used, some traces of the metal would have remained, either as cor- roded particles or as stains, but nothing of this nature was found. From a number of unfinished specimens and from the appearance of many of the blade slits themselves, it can be seen that in many cases the slits were not made by cutting directly from the tip, but that instead, a slot of uniform breadth was first cut through from one side to the other, leaving the upper ends connected. This connection at the tip was then cut through, but by means of a narrower slit. When completed, therefore, the blade slit was wider at the bottom than at the top, with the result that the tips would have to be sprung back slightly to receive the blade which would then be held firmly in place. It should be noted that in harpoon heads with an end blade, those having the blade parallel with the line hole are usually thin in cross- section, whereas those in which the blade and line hole are at right angles are thicker. The reverse is true of harpoon heads with side blades: those having the blades parallel with the line hole (or holes) are relatively thick in cross-section, whereas those with blades and line hole (or holes) at right angles are thin. HARPOON HEADS, EARLY PUNUK, FROM MIYOWAGH In plate 28 are shown a number of open and closed socket harpoon heads of Punuk type from Miyowagh, most of them from the later, northwestern section of the midden. 116 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 Open SockeT Harpoon HEADS Open socket type II (a) «—Two lashing slots or one slot and an opposite groove ; bifurcated, asymmetrical spur; round line hole; two side blades, parallel with line hole ; long and slender, squarish in cross- section ; usually made of bone. (PI. 28, fig. 1.) Two examples, both of bone: cut 1, depth 12 inches; cut 19, depth 24 inches (pl. 28, fig. 1). This is a derivative of the older type II x, from which it differs in having two instead of three prongs on the spur; it is also long and slender in shape, whereas type II x was shorter and more or less rounded in cross-section. A further difference is that both specimens of the present type are of bone, and that one of them (pl. 28, fig. 1) has a single lashing slot, with an opposite groove instead of two slots. Open socket type II (a) y—Same as above but with the side blades at right angles to the line hole; flattened oval in cross section. (Pl. 28, fig. 2.) Three examples, all of bone: cut 1, depth 12 inches (pl. 28, fig. 2) ; cut 19, depth 17 inches; cut 20, depth 22 inches. This form is derived from type II y, differing from it in having two instead of three prongs on the spur, and in sometimes having one lashing slot with an opposite groove instead of two slots. The re- maining fragmentary side-blade of plate 28, figure 2, is made of a piece of shell. Open socket types II (b) x, II (b) y, II (c)_%, and II (c) y are variants of the type II group which were not found at Miyowagh. I-xamples from the next site, levoghiyoq, will be described later. Open socket type II (d).—Two lashing slots; a bifurcated asym- metrical spur; round line hole; neither end nor side blades; long and slender ; made of bone. (Pl. 28, fig. 3.) This type, which apparently was little used at Gambell, is based on the single specimen shown in plate 28, figure 3, from the north- western section of the midden, cut 17, depth 20 inches. It is a late variant of the type II head, and is of particular interest as repre- senting the first appearance at Gambell of a harpoon head which has neither blade nor barbs. Open socket type II (e).—One lashing slot and an opposite groove ; a bifurcated, asymmetrical spur ; triangular to rectangular line hole; two small lateral barbs, at right angles to line hole; no blades; long and slender. (PI. 28, fig. 4.) Another uncommon type, represented by the single specimen, of ivory, shown in plate 28, figure 4, from cut 1, in the northwestern section of the midden at a depth of 24 inches. NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS WA Open socket type II (f) x—Two lashing slots; bifurcated, sym- metrical (median) spur; triangular line hole; end blade, parallel with line hole; ornamental, vestigial slots for side blades, at right angles to line hole. (PI. 28, fig. 5.) This type is based on the single specimen, of bone, shown on plate 28, figure 5, from cut 20, depth 12 inches. The bifurcated, sym- metrical spur, placed medially below the socket, the triangular line hole, and the two vestigial sockets for side blades are features that show relationship to type II (b) y; a type not found at Miyowagh but represented by two specimens from the Punuk site, levoghiyoq. The slit for an end blade relates it to the following type, type II (g) x. Decoration consists of a pair of straight lines diverging downward ; the shallow grooves along the edges at the constricted center, although now only ornamental, undoubtedly represent vestigial sockets for side blades. Open socket type II (g) «Two lashing slots or one slot with an opposite groove; a bifurcated, asymmetrical spur; round line hole; end blade, parallel with line hole ; made of bone ; simple line decoration. (Pl 28; figs. 6, 7.) This is one of the instances where a determination of relationship based strictly on definition would be misleading. Thus, plate 28, figure 7, possesses the general features that would place it in type III x: two lashing slots, a bifurcated, asymmetrical spur, an end blade parallel with a round line hole. But both this specimen and plate 28, figure 6, lack the graceful, flowing contour and the small barbs of type III x. Furthermore, the nature of the bifurcated spurs, the decora- tion of straight lines, and the fact that these heads are made of bone rather than ivory show that their true relationship is with the type II group. Plate 28, figure 6, from cut 6, depth 20 inches, is rather crudely made: it has one lashing slot and an opposite groove. Plate 28, figure 7, from cut 2, depth 28 inches, has two lashing slots; in addition to the simple line decoration, it has a deep V-shaped depression above the line hole. The harpoon heads above described [types II (a) x and y; II (d); mee) 11 (i) x; Il (g) x] are collectively the most recent of all the Miyowagh forms. They all come from the northwestern section of the midden, one of them (pl. 28, fig. 7) from a depth of 28 inches, the nine others from even shallower depths. They were relatively unimportant at Miyowagh, being overshadowed by the more domi- nant forms of type III (a) x and y. One of them—type II (a) x—is a degenerate form of type II x (pl. 24, figs. 5-8), a type which goes 9 118 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 back to the Old Bering Sea period. The others are all related to the somewhat later type II (a) y (pl. 24, figs. 9, 10), which, as was pointed out previously, represents the first appearance in the Gambell sequence of a harpoon head which embodies the essential character- istics of the Birnirk (Point Barrow) type. In these later forms of the type II head the Birnirk relationship may still be seen, even though each of the subtypes has in one feature or another departed from the full Birnirk. Thus they all have bifurcated spurs, usually lateral and asymmetrical ; two of the forms have functional side blades, and one has ornamental grooves representing vestigial side blades; three of them lack end blades; and most of them have a single lashing slot with an opposite groove as frequently as two lashing slots. It is also significant that of these 10 specimens, g are of bone (in addition to 4 unfinished heads and 1 fragment which are not included in table 2 because of uncertainty as to which particular subtype of type II they would have belonged). On the other hand, bone was very seldom used in making other types of harpoon heads at Miyowagh ; among the hundreds of specimens, fragments and all, there are, in addition to those enumerated above, only nine which are made of bone. The fact that more than half of all the bone harpoon heads from Miyowagh are of these Birnirk-like forms would make it appear either that the latter were importations on St. Lawrence Island; or if of local manu- facture, that they were made by emigrants from some region where bone was the predominant material used, and the Birnirk head the predominant type. Further evidence in the same direction is afforded by the decoration on these Birnirk-like specimens. Wherever decora- tion is present it conforms rather closely to that found on the Birnirk heads from around Barrow—a pair of straight lines (one of which may be forked) down the sides, and frequently a V-shaped depression just above the line hole. Plate 28, figures 1, 2, 4-7, all show this style in some degree, although the decoration is rather crudely applied. The four harpoon heads shown in plate 27, figures 1-4, from Punuk Island and Cape Kialegak, are better examples of the Birnirk type, both in form and ornamentation, than those which we happened to find at Gambell. Open socket type III (a) x.—Two lashing slots; a single, lateral spur; round line hole, below which is usually a shallow groove into which the lashing slots open ; end blade, parallel with line hole ; decora- tion Punuk (rarely Old Bering Sea), or plain. (Pl. 28, figs. 8-14.) This type includes very few of the Old Bering Sea heads, mainly for the reason that during this period the spur was usually divided rather than single. There are, however, three specimens with Old NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 119 Bering Sea decoration which must somewhat arbitrarily be included in the group because they possess single spurs and the other features characteristic of the type. One of these is from the Hillside site (pl. 23, fig. 10), the other two from the southeastern section of Miyowagh, cut 25, 18 inches deep, and cut ga, 52 inches deep. The last mentioned specimen is so badly weathered that most of the super- ficial features are obliterated ; the other two, however, both have small ornamental barbs along the edges, such as are characteristic of type III x. These barbs, together with the ornamentation, indicate rela- tionship to type III x, even though for the purpose of classification it is necessary, because of the single spur, to include them in type III (a) x. Type III (a) x is an important one at Miyowagh, being represented by 38 specimens in addition to the 2 just mentioned. Of these 38, only one is undecorated, the others all bear a Punuk ornamentation. Three are from the older, southeastern section, as follows: cut 7, depth 20 inches; cut 9, 12 inches (pl. 28, fig. 11) ; 36 inches. Thirty-five are from the later, northwestern section: Cut I, 12 inches ; 24 inches (2) (pl. 28, fig. 13); 48 inches. Cut 2, 12 inches; 24 inches (2) (pl. 28, fig. 8). Cut 3, 24 inches. Cut 4, 15 inches; 20 inches; 25 inches (pl. 28, fig. 14) ; 31 inches (pl. 28, fig. 12) ; 39 inches. Cut 17, 26 inches; 27 inches (2) ; 34 inches. Cut 19, 5 inches fee 2 inches (3) (pl. 28; fig. 10); 17 inches (6) (pl: 28, fig. 9) ; 24 inches ; 48 inches. Cut 20, 12 inches; 22 inches; 26 inches (2). In describing the harpoon heads of the Old Bering Sea period, it was pointed out (pp. 104-106) that open socket type III x was a variable form that had undergone considerable modification during the period in which Miyowagh was occupied. The older specimens were mostly large, elaborately ornamented, and possessed prominent, divided spurs; the later specimens were of smaller size, were decorated in the simplified style of the early Punuk, and had spurs which, although considerably reduced, were still to be classed as “ irregular.” The type III (a) x heads represent a continuation of this process of simplification. In general shape and style of ornamentation they are similar to the latest of the type III x heads, such as plate 24, figures 22 and 23; the principal difference being that the spur has lost all trace of its former irregularity and is now perfectly straight and plain. Furthermore, while the earlier forms of the type III x heads were oval in cross-section, with a more or less rounded contour, a few of the later examples, like plate 24, figures 22 and 23, were somewhat squarer in outline, with plane median surfaces and beveled edges. In the type III (a) x heads the same tendency is continued : 120 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 the edges descend at a rather sharp angle from the center, which is either flat or slightly concave. The small rounded elevations to either side of the line hole—survivals of the elevated Old Bering Sea ‘eye ’’—are to be seen on these type III (a) x heads just as they were on the early Punuk forms of type III x. They are lacking on plate 28, figure 14, a head which, to judge from the decoration, is one of the latest of its type at Miyowagh. In short, the harpoon heads of type III (a) x have become thor- oughly standardized ; they exhibit hardly a trace of the rich diversity of form that characterized the ancestral Old Bering Sea heads of type III x. The ornamentation has likewise arrived at an almost dead level of uniformity. The same basic design, with but slight variation, is repeated with monotonous regularity on each decorated harpoon head of this type. This design consists of two parts: a pair of lines beginning at a single point at the base of the blade slit (on the side opposite the socket), then diverging to pass around either side of the line hole, and becoming forked at the lower ends; paralleling these are two outer lines, beginning nearer the tip and extending downward along the edges, where at a point below the level of the line hole they each meet a similar line from the opposite side. Between these outer lines on the opposite (or socket) side there is again a pair of lines passing around the line hole, but because of the space taken up by the socket, these lines do not fork at the ends but continue straight. Such is the skeleton of the design on each of these harpoon heads. Within these lines are smaller decorative elements—light spurs and dots—which fortunately are less stereotyped than the lines them- selves. At first the spurs are small and faint, and are attached to the lines more or less at a right angle (pl. 28, figs. 8-11). They may occur either singly or in pairs; when they occupy a space between two converging lines there is a tendency for them to join and for a dot, or two dots, to be placed just above or below the point of junction. Later, the spurs became longer (pl. 28, figs. 12, 13) and of necessity more acute in order that they might occupy the narrow space between the converging lines. Plate 28, figure 14, is the end result of this elongation of the spurs, which, having joined, now appear as bold V-shaped figures between the forks and angles of the lines. This and two others (cut 17, 34 in., and cut 20, 26 in.) are the only harpoon heads from Miyowagh with spurs that have reached this exaggerated stage, which as we shall see later is a characteristic style of decoration at the Punuk site, levoghiyoq. Open socket type III (a) y—Same as type III (a) x except that the blade slit is at right angles to the line hole; decoration Punuk, or plain (rarely Old Bering Sea). (PI. 23, fig. 9; pl. 28, figs. 15-18.) NOS, of ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS I2I One head of this type was found at the Hillside site (pl. 23, fig. 9), decorated in Old Bering Sea style 2. This is one of the few specimens from the Hillside site on which the “ animal head ” is clearly depicted. The opposed “ eyes ”, to the left of the line hole, are formed of small concentric circles placed on rounded, elevations, two pairs of short converging lines extending downward from the periphery of the outer circle. This element is enclosed within an ovoid panel; the “nose” is indicated by a shallow notch cut in the edge. The groove leading down from the line hole is rather deep, and the spur projects outward to an unusual degree. Type III (a) y is represented at Miyowagh by 13 examples. Seven of these are from the southeastern section, as follows: Cut 9, 24 inches (3) (pl. 28, fig. 18) ; 36 inches. Cut 21, 29 inches. Cut 23, 64 inches. Cut 25, 18 inches. The other six are from the northwestern section: cut 4, 43 inches (pl. 28, fig. 15). Cut 19, 17 inches; 38 inches (pl. 28, fig. 16) ; 45 inches (pl. 28, fig. 17) ; 51 inches ; 59 inches. As in the case of other open socket heads, this type with blade slit at right angles to the line hole was less frequently used than the type with blade slit and line hole parallel. However, the distribution indi- cates that it was a somewhat earlier form. It is a modification of open socket type III y (pl. 26, figs. 1-10), and it closely resembles the later forms of that type, differing only in having a single instead of a divided spur. It has retained the sharp, prominent ridge between line hole and tip, a feature that contributes greatly to the symmetry and beauty of this and other types of harpoon heads which have the line hole and blade slit at right angles. Plate 28, figure 16, was pro- vided with a secondary blade slit, parallel with the line hole, after part of the upper end had broken. Figure 18 is the only harpoon head excavated at Miyowagh which had the lashing thong still in place. It consisted of a narrow strip of baleen passing through the slots and was tied together in a neat, tight knot just over the socket. Seven of the 13 specimens from Miyowagh are decorated in Punuk style, and six are undecorated. Plate 28, figure 15, is of particular interest because in addition to the usual Punuk ornamentation of lines and spurs it has two curving transverse lines at the upper end and longi- tudinal rows of very short lines (broken lines) extending down the sides from the tip to below the line hole. Broken lines of this kind are common in Old Bering Sea art, but this is the first instance in which they have been found accompanying a typical Punuk orna- mentation. The decoration on the other specimens (pl. 28, figs. 16, 17, and the four not illustrated ) is practically identical to that occurring on the type III (a) x heads (pl. 28, figs. 8-13). Plate 28, figure 16, 122 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 and one other have plugs of baleen set in the small round pits, and most of them have the rounded elevations beside the line hole that are so characteristic of the early Punuk heads in general. CLosED SockeT Harpoon Heaps We will now consider the closed socket harpoon heads of Punuk type from Miyowagh, most of them from the northwestern section of the midden. No examples of the older types—I-I V—were found in this section. Closed socket type V x (pl. 28, figs. 19-22).—Definition of the type was given on page 110, where the six specimens with Old Bering Sea ornamentation were described, two of them from the Hillside site and four from Miyowagh. The remaining 18 specimens from Miyowagh are either undecorated or bear a Punuk ornamentation. Six of them come from the southeastern part of the midden, as fol- lows: Cut 9, 24 inches (2); 25 inches. Cut 23, 8 inches) (@uteaas 18 inches. Cut 27, 16 inches (pl. 28, fig. 20). Twelve are from the northwestern section: Cut 1, 30 inches (pl. 28, fig. 22). Cut 2, 12 inches: Cut 3, 24 inches. Cut 4, 43 inches. Cut 6,6 inches» @ateag 12 inches ; 17 inches; 24 inches; 42 inches; 45 inches; 48 inches (pl. 28, fig. 21). Cut 20, 12 inches (pl. 28, fig. 19). Most of the harpoon heads of this type are small, from 6 to 8 cm long. Three of the heads have line holes that depart from the usual round shape. That of plate 28, figure 20, is somewhat squarish, figure 21, oval-rectangular, and that of another from cut 24, depth 18 inches, is triangular. The small specimen, plate 28, figure 22, was provided with a lashing slot and opposite groove, and used in the manner of an open socket head, after a piece had broken away from the socket. The Punuk ornamentation on most of these closed socket heads differs from that on the open socket forms. On the latter, the lines that descend from the blade slit pass around the line hole and continue downward, each line becoming forked at the end. On the closed socket heads the lines begin in the same way, but instead of continuing to diverge, they come together again below the line hole, which is thereby enclosed in a lozenge-shaped figure. Another line, parallel with the edge, extends down both sides, and dots or spurs are applied at intervals. Closed socket type V (a) x.—Single, symmetrical spur; round line hole, beveled below; end blade parallel with line hole; small barbs along edges; oval in cross-section; decoration Punuk. (Pl. 28, figs. 23,242) NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 123 This is a relatively unimportant type, represented by the two speci- mens shown in plate 28, figures 23 and 24, from cut 27, depth 22 inches, and cut 20, depth 12 inches respectively. It is a variant of type V x, from which it is distinguished by the presence of pairs of small lateral barbs. Although these appear to be primarily decora- tive, they may also have been functional to a certain extent. I*igure 23 is of bone; its graceful, flowing contour and the marked elevations suggestive of “eyes ”’ around the line hole and on the spur bring this specimen into close relationship to the Old Bering Sea forms. The line decoration, however, is rather simple. After part of the upper end had broken, a blade slit was cut in the remaining tip at right angles to the line hole and the original blade slit. The second harpoon head of this type, plate 28, figure 24, is clearly younger, coming as it does from the surface layer of the latest section of Miyowagh, and bearing a decoration in the more developed Punuk style. The design above the line hole—a V or Y-shaped figure—is closest to that of plate 28, figure 14. The decoration on the spurs consists of deep round pits at the center of rather lightly incised mechanically made circles, each circle connected to a line that descends obliquely from a line near the center. Closed socket type V y (pl. 28, figs. 25-28).—The type was de- fined on page 111. In all, it is represented at Miyowagh by 21 ex- amples. Twelve of them came from the older, southeastern section and have been described with the Old Bering Sea forms; ten of them were decorated in Old Bering Sea style, two were undecorated. The remaining nine are from the northwestern section of the midden, five of them decorated in Punuk style, four undecorated. They are distributed as follows: Cut 1, 30 inches (pl. 28, fig. 26). Cut 2, @eeiiches (pl. 28, fig. 25); 40 inches. Cut 19, 12 inches (2); 30 inches; 59 inches (2) (pl. 28, fig. 27). Cut 20, 12 inches (pl. 28, fig. 28). The harpoon heads of this type which are decorated in Punuk style are as a group considerably smaller than the Old Bering Sea heads of the same type. The smallest of these is plate 28, figure 27, only 5.7 cm long. Like plate 28, figure 17, and plate 26, figure 8, it has plugs of baleen set in the small round pits, a feature which appears to be characteristic of a rather early stage of the Punuk. Plate 28, figure 28, is the only harpoon head found at Miyowagh having the large nucleated, mechanically made circles of the fully developed Punuk (Punuk style 2), although two other heads (pl. 28, figs. 10 and 24) have smaller, more faintly incised circles of the same kind. The two latter specimens should probably be regarded as intermediate I24 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 stylistically between those having simple dots—or more accurately small, deep, circular pits—and the large bold circles surrounding similar pits as on figure 28. IMPLEMENT TYPES, OLD BERING SEA Harpoon Parts Foreshafts —Considering the importance of the foreshaft, and the many that must have been used, it is somewhat surprising that only Ii examples, and those mostly fragmentary, were found at the Hill- side site and only 45 at Miyowagh. The relative scarcity and the poor condition of the foreshafts is probably to be explained by the fact that this is perhaps the least stable part of the harpoon, a slender rod of ivory fastened into the heavy socket piece, which is very likely to break on the impact, either because of a poor cast or because the free end fails to become disengaged from the socket of the harpoon head with the first vigorous movements of the quarry. Many fore- shafts must have been broken in this way and discarded immediately, since the harpoon would be useless until another had been fitted in place. The foreshafts from the Hillside site, though few in numbers, varied considerably in form. One of them is the decorated specimen shown in text figure 5, which was found beneath the floor stones of house no. 1. It is triangular in cross-section. The lower end, which rests in the upper end of the socket piece, is tapering and was probably somewhat pointed originally; the upper end, which fits into the socket of the harpoon head, is broken. Just below the center is the line hole, a narrow slit 11 mm long. Plate 29, figure 2, is quad- rangular in cross-section with one face much narrower than the other three. The line hole is a narrow slit as in text figure 5 but is placed nearer the middle. The lower end, broken, is tapering; the upper end is also tapering, and its rounded shape shows that it was intended for a harpoon head with a closed socket. Plate 29, figure 3, is similar in cross-section but the line hole is higher up. The base is conical and sharpened, the upper end broken. Two other fragments resemble it in having sharply pointed tangs and narrow rectangular line holes g and 12 mm long respectively. One other fragment has a triangular and another a round line hole. Plate 29, figure 1, differs from the others, having an enlarged and somewhat bulbous lower end, just above which is the narrow rectangular line hole; it is only 6.7 cm long; the tang is blunt and the tip cylindrical. Plate 29, figure 4, is a complete specimen 9.9 cm long with pointed tang and tip and a narrow line hole. NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 125 The foreshafts from Miyowagh are also for the most part frag- mentary and it is difficult to describe the types very exactly. There seem to have been two main general types: (1) rather heavy and long, squarish to rounded triangular in cross-section, with narrow rectangular line hole between center and tang; the tang conical and rather blunt; length 12 to 25 cm; (2) a smaller, more slender form; cross-section round or as above; line hole narrow and rectangular or short and triangular; tang conical but sharper and more tapering ; length 6.5 to 12 cm. The first type, of which there are 25 examples, is illustrated in plate 31, figures 1-5. The first two, from cut 9, are the only two fairly complete specimens of this type; the curved tang of plate 31, figure I, is probably due to warping. Plate 31, figure 7, from cut 7, depth 75 inches, is a small, sturdy foreshaft which structurally re- sembles figures 1 and 2; it has a blunt tang and cylindrical upper end for use with a closed socket head. Plate 31, figure 3, from cut 27, depth 26 inches, is broken off above the line hole, and has a squared- off tip, being intended for use with an open socket harpoon head. Plate 31, figure 4, the tip of a similar foreshaft, is from cut 18, 27 inches deep. Plate 31, figure 5, the tip only, is cylindrical, for use with a closed socket ; it is from cut 4, 40 inches deep. The second type of foreshaft, represented by 20 examples, is illus- trated on plate 31, figures 8-10. Plate 31, figures 8 and 9, cut 9, 24 inches deep and cut 24, 13 inches deep, have the tang roughened and the line hole about midway. Plate 31, figure 10, from cut 7, 51 inches deep, illustrates the triangular type of line hole. Plate 31, figure 6, from cut 19, 51 inches deep, differs from both types in having had a large round line hole. It is larger than type 2, and probably fell within the length range of type 1. Its rounded tip shows it to have been used with a closed socket harpoon head. It is decorated in Old Bering Sea style, with the line hole bordered by double, spurred circles to which were attached streamers—pairs of straight, rather deep lines paralleled by very finely incised lines. Provenience of the specimens not illustrated need not be given, as the distribution reveals nothing of chronological significance. Comparison with the foreshafts from the Hillside site (pl. 29) shows essentially the same types; figures 2 and 3 of that plate correspond in general with plate 30, figures 1-5, whereas figure 4 and other fragments not illustrated fall within the range of the type illustrated in plate 31, figures 8-10. In addition to the unquestioned foreshafts, there are a few perfor- ated objects and a very large number of plain, slender ivory and bone rods from both the Hillside site and Miyowagh which conceivably 126 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 might have been foreshafts. In view of the many possible uses which such simple objects might have had, it seems better not to attribute any particular function to them as a group. The two fragments shown in plate 33, figures 26 and 27, from cut 19, 12 and 24 inches deep, respectively, seem too small to have been socket pieces for harpoons; they were probably for darts, although there is no evidence as to the nature of the heads used with them. The first has a rounded knob at the end and a small, round socket 13 mm deep and 7 mm in diameter. Encircling the socket are two closely spaced lines with spurs attached, and below these a third line, also spurred. Plate 33, figure 27, has a raised rim and a small round socket Ig mm deep and 6 mm in diameter. The exact function of the two small socket pieces shown in plate 33, figures 28 and 2g, is likewise uncertain; they are from cut 7, 37 inches, and cut 10, 40 inches deep. Both have rounded sockets and triangular holes near the base; the tang of the first is conical, that of the second wedge- shaped with the hole passing through it. Shafts——Plate 46, figure 15 (cut 18, depth 44 in.) is probably a piece of a harpoon shaft. It is oval in cross-section and has a slanting, scarfed surface where another section, cut in the same way, was joined on. At the upper end is a socket 5.7 cm deep in which the tang of the socket piece rested. Lances.—Plate 57, figure 1, is a detachable wooden foreshaft for a lance, from cut 23, depth a inches. It is a flattened oval in cross- section and is 32.3 cm long, with a blunt, rounded tang and a blade slit 3 cm long. Nelson figures several lance foreshafts of this same general type from western Alaska (pl. 57) and describes their use as follows: These lances are used when the seal or walrus has been disabled, so that it can not keep out of reach of its pursuers, when the hunter paddles up close alongside and strikes the animal, driving the detachable head in its entire length. The head remains in the animal, and the hunter immediately fits another point into the shaft and repeats the blow, thus inserting as many of the barbed heads as possible, until the animal is killed or the supply of points exhausted. [Nelson, p. 147.] Socket pieces—Two harpoon socket pieces were found at the Hillside site; one of these is the beautifully decorated but badly weathered specimen shown in plate 13, figure 1, the ornamentation of which has been described on page 48. It is 15 cm long and is broken off at the line hole. It is oval to round in cross-section, and was probably more than 20 cm long originally. In the upper end is a pit for the foreshaft 1.5 cm deep and 1.4 cm in diameter, which NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 127 appears to have been widened somewhat through weathering. Most socket pieces of this type (others to be described later) had a small opening near the upper end through which passed the line that held the foreshaft in place, and below it another and larger hole for the line that connected the socket piece with the wooden harpoon shaft. On this specimen, however, there is no hole near the upper end, from which it would appear that the foreshaft was attached in some other manner, probably by means of a line fastened to the lashing which held shaft and socket piece together. An unfinished socket piece was found in house no. 1. It is 21 cm long, oval in cross-section with a conical tang 4.5 cm long. Socket pieces are also scantily represented at Miyowagh, with two complete and seven fragmentary specimens. Plate 31, figure 18, is a badly weathered socket piece from cut 13, depth 14 inches. Only a few traces of the Old Bering Sea ornamentation remain, although the deep circular pits, or “eyes”, are quite distinct. In the upper end is an oval socket 18 mm deep. Five cm from the top is a circu- lar line hole from which, on either side, descends a deep narrow channel. This upper line hole would normally have received the line attached to the foreshaft, but in this case the downward-pointing grooves would seem to indicate a line connecting with the wooden shaft. If so, one of the two lower holes could have been for the fore- shaft line. The tang was conical and was set in the end of the wooden shaft, as was the case with an Old Bering Sea socket piece from Little Diomede Island figured by Jenness (1928 a, pl. 13, c). The few other socket pieces of this period that have been described have bifurcated tangs.’ One half of such a tang is shown in plate 31, figure 15 (cut 24, 54 in.). It is decorated in typical Old Bering Sea style, having two ovoid panels, both spurred, enclosing small, deep circles between converging lines, suggestive of “eyes’’; and between and beyond these a number of deep and lightly incised lines together with broken lines. Through the lower end is a small rectangular slot for the line which passed around the inserted end of the wooden shaft, the same arrangement as on the two specimens figured by Mathiassen, 1929, figure 15. Plate 31, figure r1, is a poorly preserved socket piece, probably un- finished, from cut 16, 26 inches deep. It had a conical tang and a triangular line hole; the upper end is broken. Plate 31, 17 and 19, * Mathiassen, 1929, fig. 15, a, b, from Point Hope and Kotzebue Sound. Collins, 1929, pl. 7, c, from Miyowagh; (pl. 8, c, is incomplete). Collins, 1935, pl. 5, fig. 1, from Kukuliak, St. Lawrence Island. 128 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 are the lower ends of two similar socket pieces, one from cut 9, depth 24 inches, and one from cut 23, 25 inches deep. The first has a simple decoration consisting of four pairs of lines, one to each side, extending upward from the level of the line hole. The second was ornamented in Old Bering Sea style, with the line hole bordered by continuous and broken lines and with panels (only the tips of which are left) made in a similar fashion. Plate 31, figure 16, is a small socket piece from cut 17, depth 27 inches. It is squarish in cross- section and appears to be unfinished; the socket is circular, 14 mm in diameter and 16 mm deep. Three fragments are shown in plate 31, figures 12-14. Plate 31, figure 14 (cut 27, depth g in.) has a socket 17 mm deep and II mm in diameter; it is decorated in the Punuk style with lines and spurs and detached dots. Plate 31, figure 13, from cut 19, 79 inches deep, is made from a walrus penis bone and has been cut off ; the socket is 9 mm deep and 12 mm in diameter. The remain- ing decoration consists of a pair of lines connected by a short curving line, and on the reverse a Punuk design composed of two straight lines, two lines forming a V, and two detached dots. Plate 31, figure 12, from cut 27, 52 inches deep, is likewise only the upper end. It has a socket 20 mm deep and 14 mm in diameter and is decorated with two curving lines. Ice picks—Twelve ice picks for the lower end of the harpoon shaft and eight fragments were found at the Hillside site, three of which are illustrated on plate 29, figures 13-15. They are all of ivory, almost always the distal end of a small walrus tusk. They are shorter than most Eskimo ice picks, the average length being around 15 cm. The upper part, or tang, which is roughened for inserting in the lower end of the shaft, is relatively long, sometimes being half of the total length. The lower end is either pointed or spatulate. The Miyowagh collection includes 103 complete or broken ice picks and a far greater number of broken tangs. Two of the ice picks are of bone, the rest of ivory. Four types can be recognized: Type 1 (pl. 32, figs. 1, 2)—Heavy; under side more or less flat, upper side convex ; roughened conical tang ; no abrupt shoulder ; aver- age diameter 3 cm. Fifty-four examples, from all parts of the midden, top to bottom. It is an old type (found also at the Hillside site, pl. 29, fig. 15) used during the entire period Miyowagh was occupied. Type 2 (pl. 32, figs. 3, 4).—Same as type I, except that it has a well-defined shoulder between tang and body. Thirty-four examples, distributed in same manner as type I and apparently of same age. For examples from Hillside site, see plate 29, NO I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 129 figures 13, 14. On the small broken specimen, plate 32, figure 4, are the outlines of two small human faces, with the eyes, nose, and mouth indicated by dots. Type 3 (pl. 32, fig. 6)—Slender; rounded triangular to squarish in cross-section ; roughened conical tang; no shoulder; average diam- eter 1.5 to 2 cm. Ten examples, distributed as follows: Cut 10, 22 inches deep. Cut 13, 54 inches (pl. 32, fig. 6). Cut 15, 24 inches. Cut 18, 24 inches; 49 inches. Cut 19, 56 inches. Cut 23, 52 inches; 58 inches ; 64 inches. Cut 24, 64 inches. With 7 of the Io examples coming from the lower levels of the older (southeastern) section of the midden, and none at all from the later (northeastern) section, this is shown to be an old type, and one that apparently passed out of use before the abandon- ment of the site. Three examples were found at the Hillside site (U.S. N. M. nos. 352622, 352755, 352853). Type 4 (pl. 32, fig. 5).—Flat ; roughened conical tang ; well-defined shoulder. Only four examples: Cut 1, 63 inches (pl. 32, fig. 5, bone). Cut 4, 15 inches. Cut 5, 23 inches. Cut 9, depth unknown. The distribution shows this to be a more recent type than the other three, and comparatively unimportant. It is of interest to note that both of the bone ice picks found at Miyowagh belong to this type; they are made of walrus penis bones. Finger rests—No finger rests for the harpoon shaft were found at the Hillside site, but this was no doubt due to accident. From Miyowagh there are seven examples, all of ivory. They are divided into two general types: Type 1.—Long, with slightly concave base and sloping top; a round hole, a slot or two slots for the lashing (pl. 32, figs. 9-12). Plate 32, figure g (cut 7, depth 62 in.), has an almost flat base, a deep notch at the rear where the finger rests, and a single drilled hole for the lashing. Plate 32, figure 10 (cut 27, depth 38 in.) has a concave base, a slightly projecting back end and two slots for the lashing. Plate 32, figure 11 (cut 7, depth 37 in.) has the base concave, the back end sloping forward, the front end truncated, and a single long lashing slot. Plate 32, figure 12 (cut 27, depth 16 in.) has a concave base, straight back end, and single small lashing slot. Type 2—High, with a flat base; sides slightly incurving ; rounded tip; large circular hole for the lashing (pl. 32, figs. 7, 8). Three examples: The two figured (cut 22, 46 in., and cut 18, 32 in.) and a third from cut 27, depth 38 inches. 130 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 Float mouthpieces—It is of interest to observe that no tubular ivory mouthpieces such as the modern Eskimos use for inflating the bladder floats on light harpoons were found at the Hillside site or the Old Bering Sea sections of Miyowagh, although several were found in the later sections of Miyowagh. The Old Bering Sea type of mouth- piece was a round or oval wooden plug, evidently for a sealskin float used with a heavier harpoon. Twenty-one of these wooden mouth- pieces, with a hole through the center and a deep groove around the edge were found at Miyowagh. The three examples shown in plate 32, figures 13-15, illustrate the range of shapes: oval to circular, and with the top surface flat, convex, or conical. They were found as follows: Cut 4, 43 inches; 55 inches. Cut 9, depth unknown. Cut ga, 60 inches. Cut 18, 36 inches; 44 inches (2) (pl. 32, fig. 13); 48 inches (pl. 32, fig. 14) ; 68 inches. Cut 19, 34 inches; 59 inches (2) ; 63 inches. Cut 23, 39 inches; 45 inches; 58 inches. Cut 24, 54 inches ; 64 inches; 72 inches. Cut 25, 55 inches (pl. 32, fig. 15). Cut 29, 36 inches. This distribution is rather striking; most of the 21 specimens come from the older section of the midden and mostly from the lower levels, and the two from the later section (cut 4) are also fairly deep. It would appear, therefore, that this is an old type which began to fall into disuse during the period Miyowagh was occupied. For closing the openings of such mouthpieces, a wooden stopper was used (pl. 32, figs. 21, 22). There are 19 of these, distributed in much the same way as the mouthpieces: Cut 18, 28 inches; 34 inches ; 36 inches ; 44 inches (pl. 32, fig. 21) ; 48 inches; 49 inches; 59 inches; 60 inches; 63 inches; 66 inches; 72 inches. Cut 19, 56 inches (pl. 32, fig. 22). Cut 22, 36 inches. Cut 24, 54 inches; 72 inches Gum 25, 72 inches. Cut 27, 26 inches; 38 inches. Cut 28, 31 inches. No float mouthpieces were found at the Hillside site, but the pres- ence of two wooden stoppers like those illustrated from Miyowagh shows that they were used. Float plugs——Wooden or ivory plugs were used for closing holes (either the natural orifices or wound tears) in the sealskin floats. Three of these, of wood, are very similar to the mouthpieces, lacking only the perforation. One is shown in plate 32, figure 16 (cut 9, 48 in.) ; it is oval and has a very deep, wide groove. The other two are from cut I, 42 inches, and cut 18, 44 inches. There are also four ivory plugs: plate 32, figure 17 (cut 2, 12 in.) is long and narrow, ~ with a deep, wide groove; it has a flat base and a convex top, on which a decoration, much eroded, is still visible. Plate 32, figure 18 (cut 22, 41 in.) is the same, but shorter and wider; it may be unfinished as the groove is very shallow. Plate 32, figure 19 (cut 25, 18 in.) is NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS HSE convex at both ends and has a groove 13 mm wide and 4 mm deep. The fourth specimen comes from cut 20, depth 34 inches. Float bars.—For attaching a line to the float, the modern Seward Peninsula and St. Lawrence Eskimos fasten an ivory bar or handle to one end (Nelson, p. 145). The only object in the present collection which has the sunken center and knobbed ends characteristic of the modern type is the one shown in plate 32, figure 20 (cut 4, depth 25 in.). In all probability, however, some of the “ handles ” illustrated on plate 47 were float bars. Birp Darts AND FISH SPEARS Side prongs for bird darts—lIvory prongs, three of which were lashed onto the shaft of the bird dart at a point usually a little below the middle. Twenty-five complete and thirty-two fragmentary prongs were found at Miyowagh, indicating that the bird dart was an imple- ment of some importance ; however, it differed in several respects from the modern types. On the basis of the 25 complete specimens, two general types, with one subtype each, are recognized: Type 1 (pl. 33, figs. 1, 2).—Short (7.5 to 8.5 cm long) ; oval to round in cross-section ; one outside, two inside barbs ; narrow rectangu- lar lashing slot, with outer opposite edge usually smooth; tang spatu- late (rarely pointed) and beveled on inside. Eleven examples: Cut 4, 15 inches; 20 inches. Cut 7, 33 inches. Cut 18, 53 inches (2). Cut 19, 58 inches. Cut 23, 8 inches (pl. 33, fie. 1); 14 inches (2). Cut 24, 39 inches. Cut 27, 29 inches (pl. 33, fig. 2). The prong from cut 7, 33 inches deep, was the only other one having a projection just above the end of the tang, like that shown in plate eenuere I. Type ra (pl. 33, figs. 3, 4).—Same as type I, except that it has two outside and three inside barbs. Two examples: Cut 7, 75 inches (pl. 33, fig. 3). Cut 18, 66 inches (pl. 33, fig. 4). Type 2 (pl. 33, figs. 5, 6).—Longer (10 to 14.5 cm long) ; usually a longitudinal ridge down one or both sides; two outside, three inside barbs; narrow rectangular lashing slot, sometimes with two notched elevations on opposite edge; tang spatulate and beveled on inside. Five examples: Cut 2, 24 inches (2) (pl. 33, fig. 6). Cut 18, 27 inches. Cut 20, 12 inches (pl. 33, fig. 5). Cut 22, 46 inches. Type 2a (pl. 32, fig. 7)—Same as type 2 except for lashing arrangement, a groove replacing the slot. 132 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 Three examples: Cut 2, 24 inches (2) (pl. 33, fig. 7). Cut 9, 24 inches. Four specimens do not belong to any of the above groups. One of them, plate 33, figure 8 (excavated by an Eskimo), has four barbs on both inside and outside; otherwise it conforms to type 2. Plate 33, figure 9 (cut 19, 24 in.) is unique in being regularly triangular in cross-section ; it has a single notched elevation opposite the lashing slot. From their distribution, the small and relatively simple types 1 and Ia appear older than types 2 and 2a. The latter two types were contemporary, for two type 2 and two type 2a prongs were found all in a group together in cut 2. The notched elevations, one to three in number, opposite the lashing slot (pl. 33, figs. 5, 6), belong to the later period (Early Punuk) rather than to the Old Bering Sea, for they are found on a majority of the complete specimens of type 2 and also on most of the broken specimens (of uncertain type) from the later sections of the midden. The use of small circular holes, sometimes plugged with baleen, as decorations around the lashing slot is another feature associated with the later type. Two badly weathered fragmentary side prongs were found at the Hillside site. They were small and apparently belonged to type 1. End points for bird darts—There is no very good evidence as to the nature of the single points which were set in the end of the bird darts. The only specimen from Miyowagh which is at all comparable to the long heavy points used on modern bird darts (see Nelson, pl. 59, figs. 7-10) is that shown in plate 33, figure 11, from cut 9, 24 inches deep. It is roughly triangular in cross-section with two pairs of opposite barbs and a plain conical tang. Plate 33, figure 10 (cut 19, 34 in.) seems also to have been an end point; the barbs are placed alternately instead of opposite, and there is a longitudinal ridge down each side, giving it a lozenge shape in cross-section ; the tang is rough- ened. Plate 33, figure 12 (cut 20, 30 in.) may have been an end point ; it has a long sharp tang and three barbs. Two other doubtful examples are shown in plate 33, figures 13 and 14 (cut 18, 18 in., and cut 9, 24 in.) ; the first has two barbs (partly removed by secondary cutting ) and the second, one barb; both have plain tangs. Plate 33, figure 15 (cut 19, 38 in.) is the only example of an end point which was lashed in place; for this purpose it was provided with a scarfed sur- face on the inside, with the opposite edge notched. The only end point found at the Hillside site was the badly weath- ered specimen shown in plate 29, figure 5. It has a conical tang and a line hole made by connecting two round drilled holes. There are NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 133 seven barbs, four on the right edge and three on the left, each placed midway between the two opposite barbs. Points for fish spears.—In plate 33, figures 16-22, are shown several slender ivory and bone points from Miyowagh, the use of which is not certain, although they may have served as points for light fish spears or darts. Plate 33, figures 16 and 17, from cut 19, 12 and 51 inches deep, have a single row of barbs and long pointed tangs; the latter specimen is lashed with baleen to a wooden rod or shaft, so light and small that it would seem to have been a toy rather than a real projectile. Plate 33, figure 18 (cut I, 30 in.) is similar but longer. Plate 33, figure 19 (cut 20, 22 in.) is the longest of these points; it is 20.8 cm long, has 11 widely spaced barbs, a pointed tang, and two small protuberances similar to the notched elevations on the side prongs of bird darts; it is of bone. Plate 33, figures 20 and 21 (cut 17, 26 in., and cut I, 18 in.) are short, relatively thick, and are equipped with three and two barbs near the tip. Plate 33, figure 22 (cut 27, 32 in.) is flat with three bilateral semicircular “ barbs.” THROWING Boarps Four pieces of throwing boards were found, all at Miyowagh. Plate 37, figure I (cut 28, 37 in.) shows the general shape, except for the upper end. The index finger rested in a deep hole on the back, so deep that it has worn through to the front; just below, along the left edge there are three pits for the other fingers; the thumb rested in a broad notch on the opposite edge. The smaller fragment, plate 37, figure 2 (cut 18, 53 in.), is from the lower end, which was rather narrow and straight. Two other fragments similar to this come from cut 18, at depths of 32 and 4o inches. The throwing board is no longer used on St. Lawrence Island, and these old examples differ in certain respects from all other known types. In the style of grip, they resemble the old Point Barrow form more than any other. The two small ivory objects shown in plate 29, figures 8 and 9, from the Hillside site, appear to be pegs which were placed at the end of the throwing board for engaging the butt end of the dart. The shank end of plate 29, figure 8, was set into the end of the throw- ing board at an oblique angle; plate 29, figure 9, was probably lashed on, with the flat lower side downward, leaving the point to engage in the end of the dart shaft. Two similar specimens are shown in plate 33, figures 23-24 (from cut 24, 39 in., and cut 1, 24 in.). In the first the tang extends almost straight back, having been set in an oblique hole. In the second the 10 134 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 tang is more nearly at right angles to the upper part. In both cases the free end or tip is pointed or nearly so, and the top surface flat. Plate 33, figure 25 (cut 27, 16 in.), is an ivory peg with a small rounded socket in the upper end. This may have been set in the butt end of a dart shaft to engage the tip of a spur such as just de- scribed; the socket, however, seems rather small for this purpose. Ick CREEPERS Like the modern Eskimos of northern Alaska, those of the Old Bering Sea period used “ice creepers ’—flat pieces of bone or baleen with inset ivory spikes, which were attached to the soles of the boots to keep the feet from slipping when walking on ice (see Nelson, p. 215). The collection from Miyowagh includes six, all different from those used at the present time. Plate 37, figure 3 (cut 18, 81 in.) is of whale bone, with three inset ivory pegs or spikes. It was fastened to the foot by means of thongs which passed through a pair of holes at each end; the holes are connected by sunken grooves to prevent the lashing from being worn down. Plate 37, figures 4, 5 are of baleen; the first (cut 19, 51 in.) has the ivory pegs in place, the other (cut 23, 58 in.) has the pegs missing. Another baleen specimen similar to this but shorter was found at a depth of 44 inches in cut 27. The two other ice creepers from Miyowagh are of bone and ivory, but since they were found in the upper levels of the later section of the midden, they are described with Punuk material (see pl. 75, figs. 23 and 25). Bows AND ARROWS There is no clear evidence as to the exact nature of the Old Bering Sea bow. From a number of fragments—and toys—it appears that the middle was constricted and thickened and the ends tapering. The baleen bow shown in plate 55, figure 1, was no doubt a toy, as was also the smaller one in plate 56, figure 15, and the two of wood shown in plate 59, figures 26, 27. In this connection it should be noted that flat bone braces such as are used on some of the modern Alaskan sinew-backed bows and the sinew twisters which are always associated with the latter are both characteristic of the Punuk stage but do not appear to have been used during the Old Bering Sea period. From this it would seem fair to infer that the sinew-backed bow was un- known to the Old Bering Sea Eskimos. The only clear example of a backed bow in the present collection is a toy, with rather elaborate reinforcements of baleen and wood from cut 19, depth 42 inches, in the early Punuk level at Miyowagh (pl. 60, fig. 8). NO: I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 135 Arrowheads.—Three arrowheads were found at the Hillside site, two of which are shown on plate 29, figures 6, 7. The first is of bone with smooth, sharp, conical tang just above which are two barbs. These are made in a somewhat unusual manner; beginning a little above the center four deep incisions, two on the front and two on the back, are carried downward becoming deeper and deeper, until, just above the tang, they cut through and meet, leaving two sharp barbs which lie close to the shank and point straight downward. The third specimen from the Hillside site is of the same type, but has a blade slit; on it a decorative effect is provided on one side by two shallow lines which extend from the ends of the deeply cut lines to the blade slit, and by a faint median line which runs from tang to a little above center; the latter feature is also present on the reverse of plate 29, figure 6. The latter had a pointed tip with no inserted blade. Plate 29, figure 7, is an ivory arrowhead of an unusual type, 15.8 cm long, with smooth conical tang and a blade slit 3 mm wide. It is square in cross-section, but the four sides instead of being straight were cut spirally, so that each side makes between a quarter and a half turn as it runs from tang to tip. The Miyowagh collection contains 25 arrowheads and 2 recogniza- ble fragments. It is of interest to note that 23 of these are of bone and only 4 of ivory. This is one of the few instances where bone 1s employed in preference to ivory. Six types of arrowheads can be distinguished : Type 1 (pl. 34, fig. 1).—Rounded in cross-section; end blade; no barbs. One example, of bone, from cut 27, depth 26 inches. This was the only arrowhead found with blade in place; the latter is of chipped chert, with a straight base. Type 2 (pl. 34, figs. 2, 3)—Round to oval in cross-section; end blade; a single barb near the tang. Three examples, of bone, from cut 18, 36 inches (pl. 34, fig. 2), 42 inches, and 44 inches (pl. 34, fig. 3). The barbs lie close to the body and are made in the same manner as those on the specimen from the Hillside site (pl. 29, fig. 6). In figure 2 a line extends upward from the barb for a considerable distance, meeting another line which continues downward to the tang. The long triangular space thus formed is divided in the middle by two short transverse lines, with a long, deep line occupying both halves. This decoration is present on one side only. The second example of this type (pl. 34, fig. 3) is larger—18 cm long—and the barb has an outward flare. The line extending from the barb is opposed by a similar line ending with a 136 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 slight shoulder or notch; there are two more of these shoulders or ornamental barbs at the center; at the top, two lines converge to the blade slit, and on the lower end a decorative effect is provided by two oblique lines. Type 2a (pl. 34, fig. 4).—Rounded to oval in cross-section; end blade ; two opposite barbs near the tang. Two examples, of bone, from cut 18, 60 inches (pl. 34, fig. 4) and cut 23, 72 inches. In both of these the barbs are close to the body, and a deep line continues upward as in type 2. In plate 34, figure 4, decoration is restricted to a long, deep line between the barbs and another down one edge. Type 3 (pl. 34, fig.5).—No end blade; a single barb near the tang. Two examples: Cut 21, 38 inches. Cut 27, 26 inches (pl. 34, fig.5). The first is of ivory, with a longitudinal ridge down both sides, giving it a squarish cross-section. The second, of bone (pl. 34, fig. 5), is roughly triangular in cross-section, apparently unfinished; the single barb (with tip broken) lies very close to the body, and as a continua- tion of it, a deeply cut straight line extends upward, almost meeting a similar line opposite, which, however, does not end in a barb; between the two lines are three short, oblique lines, with a similar one lower down. In the form of the barb this type of arrowhead shows a close relationship to types 2 and 2a. Type 4 (pl. 34, fig. 6).—Rounded in cross-section; no end blade ; two opposite, prominent barbs at about center. Three examples, all of bone: Cut 19, 17 inches; 24 inches; 34 inches (pl. 34, fig. 6). The other two of this type have deeply cut lines above the barbs like plate 34, figures 2-5. Type 5 (pl. 34, figs. 7-9).—Flat, or somewhat rounded in cross- section; no blade; a single, prominent barb at about center. Eight examples, seven of bone, one of ivory: Cut 4, 25 inches. Cut 9, 36 inches. Cut 17, 12 inches; 20 inches (pl. 34, fics 7)p em 18, 25 inches. Cut 19, 38 inches; 42 inches (pl. 34, fig. 8). Cut 20, 12 inches (pli 345 fie.) Type 6 (pl. 34, fig. 10).—F lat; neither blade or barbs. Five examples, four of bone, one of ivory: Cut 4, depth 68 inches. Cut 9, 24 inches} Cut 19, 12 inches (pl. 34; fig. 10)))@upezmae inches (2). There remain two arrowheads which do not fall into any of the above types. Plate 34, figure 11 (cut 25, 61 in.), is of bone and 1s 21.1 cm long. There are three prominent barbs on one side, which were made in the same manner as those of plate 34, figures 2-6, with an opposite line in place of another barb as in plate 34, figures 2, 3, WO, I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS LS, 5. The decoration, consisting of two short and two longer lines, separated by two groups of short transverse lines, is very similar to that in plate 34, figure 2; it is present on one side only. Plate 34, figure 12 (cut 22, 30 in.), of bone, has the lower end cut away to a shoulder just above the tang. There were three barbs, near the tip, so small as to be hardly more than notches. Plate 34, figure 13, from cut 9, depth 24 inches, is an ivory point set in the end of a wooden shaft, with the baleen wrapping still preserved. In shape the point is more like a foreshaft than anything else, although it has no line hole. On the other hand the manner of hafting is that of an arrow, which it probably was, in spite of the blunt end; it may have been a child’s arrow, never intended for serious use. With this was found a “cache” of finished and unfinished projectile points, foreshafts, etc., including several arrow points which were triangular in cross- section. This is a recent type on St. Lawrence Island, and its occur- rence, together with a quantity of other ivory artifacts, mostly un- finished, at this one place in the midden, but not elsewhere, suggests that the entire ‘‘ cache” may have been later than the other material in this part of the midden. None of the tangs have knobs or projections; they are either per- fectly smooth or merely roughened by hacking. It should also be noted that all of the blade slits, or sockets, are around 4 mm wide, evidently in order to accommodate a rather thick, chipped stone blade like that of plate 34, figure I. We may now consider the relative ages of the types. It will be noted that five of the type 5 and type 6 specimens come from the later, north- western section of the midden, and eight from the upper levels of the older, southeastern section. On the other hand all 10 of the examples of types 2, 2a, 3, and 4, in addition to plate 34, figure 11, came from the older part of the midden, most of them from around the 3-foot level with the two examples of type 2a coming from depths of 5 and 6 feet. It seems clear, therefore, that types 2, 2a, 3, and 4 are older than types 5 and 6. As pointed out above, the principal distinc- tion between the two groups is the unusual appearance of the barbs: in the older types they lie usually close to the shank, with a deep line, extending upward, and opposite, a similar line which may or may not terminate in a barb. It will be recalled that both of the barbed arrow points from the Hillside site (pl. 29, fig. 6) had barbs made in this way. Arrow shafts—The lower ends of two arrow shafts are shown in plate 57, figs. 4, 5 (cut 19, 32 in., and cut 18, 66 in.). Neither of them shows evidence of having been feathered, although another specimen, from the Hillside site, has two splits extending through the shaft, evidently for this purpose. 138 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 The ivory object shown in plate 48, figure 24 (cut 18, 27 in.) may have been used for setting the feathers in the arrow shaft; the sharp, straight lower end would have been well adapted to such a purpose. Wounp PLucs Wooden plugs like those shown in plate 35, figures 11-14, are de- scribed by Nelson as follows: To insure the seals floating while being towed, it is a common practice to make slits in the skin at various points and, with a long pointed instrument of deer- horn, to loosen the blubber from the muscle for a space of a foot or more in diameter. Then, by use of a hollow tube, made from the wing-bone of a bird or from other material, air is blown in and the place inflated; wooden plugs are then inserted in the slits and driven in tightly to prevent the air from escaping. By the aid of several such inflated spots the seal is floated and the danger of losing it is avoided. [Nelson, 1899, p. 131, pl. 52, fig. 19.] It is likely that they were also used as in the Central region and Greenland to close the wound made by the harpoon. Twenty-three were found at Miyowagh, as follows: Cut 1, 60 inches. Cut 5, 30 inches; 42 inches. Cut 9, depth unknown (3). Cut 10, 31 inches (2); 54 inches. Cut 13, 40 inches. Cut 16, 12 inches; 26 inches; 46 inches; 52 inches. Cut 18, 49 inches; 53 inches. Cut 19, 45 inches (pl. 35, fig. 11) ; 59 inches (pl. 35, fig. 14) ; 86 inches (pl. 35, fig. 12). Cut 23, 45 inches; 64 inches. Cut 24, 64 inches. Cut 25, 72 inches (pl. 35, fig. 13). None was found at the Hillside site. The general shape is wedgelike, with the upper end pointed and tapering and the lower thicker end either rounded or squared off. They were perforated at the upper end, so that a number of them could be strung together on a line. Plate 35, figure 12, is unique in having two small wooden plugs set in rectangular slots at the larger end. The distribution shows the wound plug to have been an Old Bering Sea type which seems to have been less extensively used during the Punuk period. Meat Hooks Although no longer in use on St. Lawrence Island, these imple- ments no doubt correspond in function with the blubber hooks from Point Barrow described by Murdoch: *‘ For catching hold of pieces of blubber or flesh when ‘ cutting in’ a whale or walrus, or dragging them round on shore or on the ice, or in the blubber rooms, they use hooks made by fastening a backward-pointing prong of ivory on the end of a wooden handle, which is bent into a crook at the other end.” (Murdoch, 1892, p. 310). From Kotzebue Sound southward to the NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 139 Kuskokwim similar hooks are used as boat hooks ( Nelson, 1899, pp. 222-223), and it may well be that some of the larger examples in the present series were used for this purpose. Forty-six hooks of ivory were excavated at Miyowagh, in addition to six unfinished forms and nine fragments; one specimen was found at the Hillside site. They fall into the following types : Type r (pl. 35, fig. 1).—Single, long lashing slot through the top; small slot through base of prong; prong about same length as body. Twenty examples, distributed as follows: Cut 1, depth 48 inches. Cut 2, 24 inches. Cut 3, 20 inches; 24 inches; 50 inches; 54 inches. Maes, 36 inches (2) ; 43 inches (2). Cut 7, 28 inches; 33 inches ; 37 inches (pl. 35, fig. 1). Cut 8, 40 inches. Cut 9, depth unknown. Cut 18, 27 inches (2). Cut 19, 24 inches. Cut 22, 41 inches. Cut 24, 54 inches. Type ra (pl. 35, fig. 2) —Same as type I, except that it is larger, with prong longer than body. Seven examples: Cut 2, 48 inches (pl. 35, fig. 2, the end reworked into a drill). Cut 3, 20 inches. Cut 9, 48 inches; another, depth un- known. Cut 19, 42 inches; 54 inches. Cut 23, 8 inches. Type 2 (pl. 35, fig. 3). —Two short lashing slots through top; other- wise same as type I. Five examples: Cut 3, 12 inches. Cut 4, 70 inches. Cut 19, 51 inches ; 56 inches (pl. 35, fig. 3). Cut 27, 29 inches. Type 3 (pl. 35, fig. 4) —Two or three circular lashing holes through the top; a similar hole through base of prong ; prong about same length as body. Five examples: Cut 1, 55 inches. Cut 3, 24 inches; 54 inches. Cut 22, 24 inches; 36 inches (pl. 35, fig. 4). The single example from the Hillside site belongs to this type. Type 3a—Same as type 3, except that it is larger, with prong longer than body. Two examples, one from cut 15, 40 inches deep, the other picked up on the surface. Type 4 (pl. 35, fig. 6).—Five or six circular lashing holes through center of body; body longer than spur. Two examples, from cut 4, 31 inches, and cut 20, 26 inches (pl. 35, fig. 6). Type 5 (pl. 35, fig. 5).—A wide lashing groove across top and one or two circular holes through base of prong; prong about same length as body. Three examples: Cut 7, 85 inches (pl. 35, fig. 5). Cut 9, depth unknown. Cut 18, 40 inches. I40 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 Type 6 (pl. 35, fig. 7).—Long and slender; lashing groove across top; no holes; prong longer than body. One example, from cut 3, 24 inches (pl. 35, fig. 7). Type 7 (pl. 35, fig. 8) —Large and heavy ; deep, wide lashing groove across top; prominent upstanding back end; prong about same length as body. One example, from cut 17, 20 inches (pl. 35, fig. 8). Little can be said regarding the chronology of the forms. Types 4 and 7 were restricted to the later section of the midden and are therefore Punuk types, and presumably later than types I to 3a, 5, and 6. The latter, however, were generally distributed throughout the midden, suggesting that there was no segregation according to period, the several types having been more or less contemporaneous. The two small ivory hooks, plate 35, figures 9, 10 (cut 5, 32 in., and cut 23, 78 in.), may be barbs for salmon spears. They have flat bases, like the meat hooks, so that if they were salmon barbs they differ from those of the Thule Culture (Mathiassen, 1927, vol. 1, pl. 12, figs. 10, 11;-pl. 43, figs. 1, 3-5; pl. 71, figs. 2, 3; pl: V77maeeee and of the modern Alaskan Eskimo (Nelson, 1899, fig. 42, fig. 1). FISHLINE SINKERS The plummet-shaped fishline sinker is an important element of the Old Bering Sea culture. Seventy complete and eleven broken speci- mens were found at Miyowagh, all of ivory except two which were of bone; four, of ivory, were found at the Hillside site. Although relatively simple, there is sufficient variation in size, shape, and posi- tion of the line holes to warrant a division of the fishline sinkers into several types, as follows: Type r (pl. 36, figs. 1-4.).—Thick; line holes at right angles ; lower end tapers abruptly, upper end gradually, so that the maximum thick- ness is below the center; sometimes a prominent side bulge below center, projecting in plane of the lower line hole. Nineteen examples were found, distributed as follows: Cut 1, depth 38 inches. Cut 7, 46 inches (pl. 36, fig. 3). Cut ga, 12 inches, Cut 10, 50 inches. Cut 18, 51 inches (pl. 36, fig. 2); 57 inches; 59 inches (pl. 36, fig. 1) ; 60 inches (2) ; 66 inches. Cut 19, 24 inches; 56 inches (2). Cut 20, 12 inches. Cut 21, 29 inches. Cut 23, 68 inches. Cut 24, 36 inches (pl. 36, fig. 4). Cut 27, 38 inches. Cut 29, 29 inches. This is perhaps the most striking and clear-cut of the several types and also one of the oldest. Only five examples (those from cuts 1, 19, and 20) were found in the northwestern part of the midden, the others coming from the older, southeastern section. It should be noted NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS I4I in this connection that one of the three sinkers from the Hillside site was of this type. Type ra (pl. 36, figs. 5, 6).—An extreme form of type I, with the characteristic side bulge accentuated and the upper end very long and slender. Two examples were found: Cut 4, 48 inches (pl. 36, fig. 6) and cut 13, 20 inches (pl. 36, fig. 5). The upper end of figure 6 is notched, as in the case of four examples of type 1; in figure 5 the notch is enlarged so as to suggest a fish’s tail, and the line holes are parallel. Type 2 (pl. 36, figs. 7-9).—Thick to slender; more symmetrical than type 1; the side bulge, when present, is more centrally placed ; from this it grades into a more slender, symmetrical form, oval to almost round in cross-section, with evenly tapering ends; line holes at right angles. Twenty-two examples, as follows: Cut 6, 20 inches. Cut 7, 51 inches; 92 inches. Cut 9, depth unknown (pl. 36, fig. 7). Cut ga, memenes (2). Cut 10, 31 inches. Cut 18, 27 inches (pl. 36, fig. 8) ; 60 inches. Cut 19, 17 inches; 38 inches (pl. 36, fig. 9); 48 inches ; 51 inches; 72 inches. Cut 21, 24 inches; 42 inches; 46 inches. Cut 23, 68 inches. Cut 25, 96 inches. Cut 27, 9 inches; 22 inches. Cut 29, 24 inches. This, the most common type, is the least clear-cut, including forms as diverse as plate 36, figures 7 and g. However, there is no sharp line dividing them; from an asymmetrical form with a side bulge which differs from that of type 1 only in that it is nearer the center, there is a gradual transition into the more slender, symmetrical forms with evenly tapering ends. The distribution according to cut and depth (almost all of them coming from the southeastern section of the midden) shows it to be an old type, probably contemporaneous with type 1, to which it is also typologically related. Type 3 (pl. 36, figs. 10-12).—Thick to slender, with an inclination to flatness ; no side bulge ; symmetrical or nearly so; line holes parallel. Nine examples, as follows: Cut 3, 16 inches (pl. 36, fig. 12). Cut 8, 72 inches. Cut 9, depth unknown. Cut 18, 12 inches (pl. 36, fig. IO) ; 25 inches; 40 inches. Cut 19, 5 inches (pl. 36, fig. 11). Cut 23, 14 inches. Cut 25, 52 inches. Like the preceding, this is a variable type, the range of which is shown by the three specimens figured. Plate 36, figure 12, is unusual in that the two line holes are in the direction of the greatest breadth ; in the eight other examples of this type, the two line holes are at right angles to the greatest breadth. Judging from the depth and distribu- tion, this appears to have been another of the older types, only two examples coming from the northwestern section. [42 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 Type 4 (pl. 36, fig. 13).—Slender; symmetrical; round in cross- section; maximum thickness at center; both ends long and tapering with long, triangular, parallel line holes. This is a special form, represented by a single specimen from cut 7, 75 inches deep (pl. 36, fig. 13). Type 5 (pl. 36, figs. 14, 15).—Flat and relatively wide; small; line holes at right angles. Three examples: Cut 1, 38 inches (pl. 36, fig. 14). Cut 4, 31 inches. Cut 19, 34 inches (pl. 36, fig. 15). With all three examples having been found in the later, north- western section, the indications point to this type being later than those described above. Type 6 (pl. 36, figs. 16, 17).—Same as type 5 except that line holes are parallel. Six examples: Cut 4, 43 inches (pl. 36, fig. 16). Cut 7, 28 inches. Cut 18, 12 inches; 36 inches (pl. 36, fig. 17). Cut 23, 64 inches. Cut 26, 38 inches. This type is somewhat variable as to shape: three others approxi- mate plate 36, figure 16 (which is of bone), in outline; plate 36, figure 17, is unique in being very flat and narrow; the remaining example of this type is wider than the others and had also been made to serve as a drill rest. Age rather uncertain, but present indications suggest contemporaneity with type 5. Type 7 (pl. 36, fig. 18) —Small and thick ; line holes at right angles. One example, from cut 6, 12 mehes (pl. 36, fig. 18): Type § (pl. 36, figs. 19, 20).—Same as type 7 except that line holes are parallel. Three examples: Cut 1, 30 inches (pl. 36, fig. 20)2) Guteizaeem inches. Cut 23, 79 inches (pl. 36, fig. 19). Plate 36, figure 20, and the third example of the type (cut 17, 26 in.) are walrus teeth, the former with both ends slightly worked for the line holes, the latter with only one end worked. Age uncertain: the two specimens last mentioned are from the later section of the midden, although plate 36, figure 19, must be old. Four of the sinkers, two of which are shown in plate 36, figures 21, 22, do not conform to any of the above types. Plate 36, figure 21, from cut 15, depth 4o inches, was made to represent a fish; curving lines around the gills and the peculiar arrangement of the lower line hole, suggesting a mouth, add to the general impression. The lower line hole is a perforation just above the end and to the side, at right angles to the upper line hole. NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 143 Plate 36, figure 22, also from cut 15, 4o inches deep, is an unfinished harpoon head, which has been converted into a sinker by cutting line holes at the tip and through the end of the spur; both ends are deeply grooved and notched. The distribution of the various types in the two sections of the Miyowagh midden may be summarized as follows: Type | Type | Type | Type | Type | Type | Type | Type | Type | Total 1 2 3 4 6 8 | la 5 Southeastern SECHIOM. 2... 5. 12 I 17 af I 5 I 44 Northwestern SECHON. 4.2... ; 5 I 5 2 B I I By 20 It appears from the above that of the 44 sinkers from the older, southeastern section of Miyowagh, 38, or 86.3 percent, belonged to types I to 4, and 6, or 13.6 percent, to types 5 to 8; of the 20 sinkers from the later, northwestern section, 13, or 65 percent, belonged to types I to 3, and 7, or 35 percent, to types 5 to 8. Another feature which seems to possess chronological value is the notched end (pl. 36, figs. 3-7, 13, 17, 22). Fifteen sinkers had one or both ends notched, and of these, twelve belonged to types I to 4. The shape of the line hole seems to have no particular significance, although triangular or slightly irregular holes are more frequently associated with types 1 to 4 than with the later types. MENn’s KNIVES IVooden handles —Four wooden knife handles were found at the Hillside site and twenty at Miyowagh. Two of the Hillside specimens are shown in plate 29, figures 11, 12. The first of these, which has the stone blade still in place, was found beneath the floor stones of house no. 1. The wooden handle is 19.8 cm long and is notched at the lower end for suspension. The blade is of chipped chert, with one edge curved and one practically straight. At the socket end the handle was cut down somewhat, providing a slight depression for the lashing, which was prevented from slipping off by the elevated tri- angular tip. A shorter knife handle is shown in plate 29, figure 12. It is flat, pointed at the lower end, distinctly curved along one edge and straight or very slightly curved on the other. The socket is like that of the preceding specimen, with a slightly sunken area for the lashing below the thickened tip; another handle similar to this in shape has the socket end flush with the rest of the surface. The fourth 144 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 handle from the Hillside site is straight-sided and oval in cross-section, with the tip fashioned like plate 38, figure 2. The 20 wooden handles from Miyowagh fall into three distinct groups. Type 1 (pl. 38, fig. 1) —For end blade. Short (length 10 to 13 cm), and flat; blade slit 2.5 to 4.5 cm long; a slightly elevated, wedge-shaped tip below which the lashing was wrapped. Ten examples, distributed as follows: Cut 7, 37 inches. Cut 18, 42 inches. Cut 19, 54 inches ; 63 inches ; 79 inches. Cut 21, 35 inches. Cut 24, 64 inches (2). Cut 26, 55 inches (pl. 38, fig. 1). Cut 27,38 inches. The specimen illustrated is the only one of the type found with the blade intact, but another has the tang remaining in place. All handles of this type seem designed for such end blades, either of rubbed slate or of chipped stone like that from the Hillside site shown on plate 29, figure II. Type 2 (pl. 38, fig. 2).—For side blade. Longer (14.5 cm or more), and either round or rounded oval in cross-section; blade slit at the end 5.5 to 8.3 cm long; plain, pointed tip with ivory rivet for holding the blade in place. Three examples: Cut 13, 29 inches (pl. 38, fig. 2). Cut 18, 36 inches. Cut 19, 79 inches. Type 3 (pl. 38, figs. 3, 4).—For side blade. 13.8 to 19 cm long; differs from type 2 in having the side blade set in a long deep socket instead of an end slit; the socket does not extend to the tip; blade held in place by lashings above and below. Four examples: Cut 18, 42 inches. Cut 19, 63 inches. Cut 23, 64 inches (pl. 38, fig. 3). Cut 24, 36 inches (pl. 38, fig. 4). In the length of the blade this type approaches the ulu, or woman’s knife. The distribution of these three types of knife handles is striking, all having been found in the older, southeastern section of the midden or in the deeper levels of cut 19, the southernmost of the cuts in the northwestern section. These are, therefore, types of the Old Bering Sea culture. Two comparable examples were found at the Hillside site—plate 29, figure 11, and one other; these are to be included in type 1; the other type from the Hillside site (pl. 29, fig. 12) is not represented at Miyowagh. In addition to the knife handles enumerated there are two which do not fall into any of the above types and one which is doubtfully classed as a handle. The latter, from cut 25, 89 inches deep, is 26.8 cm long, oval in cross-section and has a blade slit only 2.2 cm long; NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 145 it lacks the rivet hole characteristic of type 2. Another (cut 29, 36 in.), is half of a compound knife handle, made to enclose a long, taper- ing, and slightly curved tang. The third handle, from cut 7, 51 inches, approaches the second main class of knife handles, of bone or ivory, described below. It has a socket 3 cm long and 1.2 cm deep opening both on the side and at the end; there is a small shoulder to hold the lashing, which passed around the side on which the socket opens, showing that the cutting part of the blade projected above the tip— in other words that the handle was provided with an end blade in spite of the appearance of a side socket. The width of the blade slits or sockets—from 2 to 5 mm—indicates that all of the wooden knife handles were fitted with stone blades. Ivory and bone handles—rThere are 62 knife handles of ivory and bone from Miyowagh, among which the following types can be recognized. Type 1 (pl. 38, figs. 5-7). —Consists of two pieces which were lashed together to enclose a small blade at one end, sometimes at both ends, blade being set in a short narrow socket with the opposite side rough- ened or cut away, leaving a more or less prominent tip for holding the lashing in place; occasionally a short transverse or longitudinal slot for pegging the two halves together. Length range, 5 to 15 cm; average length 8 to 9 cm. This is the well-known Eskimo tool used for cutting bone or ivory— the “ antler chisel ’’ of Murdoch, the “ whittling knife ” of Mathiassen. In all, 56 were found, 48 of ivory and 8 of bone. Since the size and shape of the blade socket may have a bearing on the kind of blade used, it may be well to consider this feature. On this basis the type 1 handles fall into three rather distinct groups: (1) with rather large, straight blade sockets—12 to 30 mm long, about 2 mm wide, and 3 to 10mm deep; (2) with very small, straight sockets—4 to 8 mm long, about I mm wide, and 1.5 to 3 mm deep; (3) with large, curved socket, 11 to 28 mm long, 2 to 5 mm wide, and 2.5 to 8 mm deep. There are 19 of the first group, distributed as follows: Cut 7, 28 inches. Cut 8, 32 inches (2, one of bone). Cut 17, 36 inches. Cut 18, 30 inches (pl. 38, fig. 6) ; 40 inches ; 42 inches. Cut 19, 12 inches; 27 inches ; 32 inches. Cut 23, 14 inches (bone) ; 18 inches ; 52 inches ; 68 inches. Cut 27, 16 inches (socket at both ends) ; 44 inches (bone) ; 46 inches ; 52 inches (socket at both ends). Cut 28, 31 inches. Of the second group, with very small blade sockets, there are 31 examples: Cut 1, 55 inches. Cut 2, 28 inches (2). Cut 7, 28 inches; 46 inches (2, one with sockets at both ends—pl. 38, fig. 7). Cut 8, 32 inches. Cut 9, 25 inches; depth unknown (2). Cut 17, 27 inches; 146 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. g6 28 inches. Cut 18, 12 inches; 27 inches (2); 81 inches. Cut 19, 12 inches ; 17 inches ; 32 inches (pl. 38, fig. 5). Cut 20, 12 inches (3, all of bone) ; 22 inches; 26 inches (2, both bone). Cut 23, 14 inches (2) ; 18 inches. Cut 24, 45 inches. Cut 27, 29 inches; 38 inches. One handle of this kind was found at the Hillside site. The sockets of the first group are large enough to have held stone blades, but those of the second group are so small that they must have been designed for metal blades. The distribution is not very enlighten- ing; the older, southeastern section yielded proportionately more of the larger socketed type, and the northwestern section more of the small socketed type; furthermore, the five small socketed bone handles from cut 20 may safely be regarded as the most recent of the entire lot. However, the distribution as a whole does not point to a clear- cut chronological distinction between the two groups. Of the third group, those with rather large socket with distinct curve at lower end, there are seven examples, distributed as follows: Cut 4, 43 inches. Cut 9, 36inches. Cut 19, 12 inches (2) ; 32 inches; 56 inches. Cut 20, 22 inches. It will be observed that with the exception of the specimen from cut 9, all of these come from the later, northwestern section of the midden. One of the handles from cut 19, depth 12 inches, has the lower end of the socket in the shape of a T; in the others, the lower end is curved. The purpose of such sockets is to hold a curved tang in place, and this would seem to indicate beyond a doubt that the blade was of metal. Type 2 (pl. 38, fig. 8) —Heavy, around 15 cm long; a deep, rect- angular, open socket at the end, in which the blade was held by means of a plug; opposite side cut down or roughened for lashing. Socket around 30 mm long, 8 to 12 mm wide, and 6 mm deep. Two examples: cut 19, 46 inches and 73 inches (the latter shown on pl. 38, fig. 8). Both have the lower end perforated for a suspension thong. Type 3 (pl. 38, figs. 9, 10). —Ten to 12 cm long ; a narrow, rectangu- lar, enclosed blade socket at one or both ends. Three examples, all of bone: Cut 18, 48 inches (pl. 38, fig. 10). Cut 19, 79 inches. Cut 25, 72 inches (pl. 38, fig. 9). The last mentioned has a blade socket at the end 14 mm long, 3 mm wide, and 13 mm deep; at the lower end is a narrow slot for a sus- pension thong. The handle from cut 18 has a narrow, rectangular blade socket in each end, one 7 and one 10 mm deep. These three handles were found at considerable depths, indicating that the type is an old one. NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 147 KNIFE SHARPENERS On plate 38, figures 11-14, are shown four knife sharpeners, made from very young walrus tusks. Knife sharpeners of this type are still used on St. Lawrence Island. Twelve were found at Miyowagh, as follows: Cut 3, 40 inches. Cut 4, 20 inches. Cut 18, 27 inches (pl. 38, fig. 12) ; 36 inches (2, pl. 38, figs. 13 and 14). Cut 19, 5 inches ; 12 inches ; iaamenes; 24 inches; 42 inches; 76 inches (pl. 38, fig. 11). Cut 24, 54 inches. They are usually unworked except for the upper end which is beveled and perforated for the suspension cord. Exceptions are plate 38, figure 14, the central part of which has been cut down and plate 38, figures 12, 13, with the upper ends carved to represent whales’ tails. In the first of these there is no perforation, the line evidently having been tied on. In the case of plate 38, figure 13, the perforation is a slot, placed about midway, with part of the baleen line still attached STONE IMPLEMENTS One of the most striking features of the Old Bering Sea culture is the extensive use of chipped stone implements. At the Hillside site 242 artifacts of chipped stone were found, far more than at all the other Gambell sites combined. Of these, 101 were of chert or jasper ; 141 of slate. There were also 140 specimens of rubbed slate. Chipped stone implements were also fairly common at Miyowagh, but here there were many more of rubbed slate. At the three later sites chipped implements were rarely found. In plates 39 to 43 are shown the vari- ous types of stone implements found at the Hillside site ; the same types were found to occur also at Miyowagh. The classification of stone implements is always difficult. Some types, like drills, gravers, and ulu blades, can be recognized easily enough, but there are others which are more questionable. Thus, it is not possible to assign a definite function to many of the cruder implements, such as flakes with a retouched edge which could have served either as knives or scrapers, and it is still more difficult to distinguish in many cases between projectile points and knife blades. Harpoon blades—No harpoon heads with an end blade in place were found at the Hillside site and only one at Miyowagh. I have designated as harpoon blades, therefore, only those tangless forms of rubbed slate which are very thin, symmetrical, and triangular in shape. Examples are shown in plate 39, figures 1-5, all broken; in all, 23 were found, none of them complete. It is probable that these 148 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 blades were used only on the smaller harpoon heads. They are bev- eled only at the tip or just along the edges ; none has the median ridge or broad faceted edges such as are found on the Thule harpoon blades ; they are also longer and somewhat narrower at the base than the Thule types. At about the center they are usually no more than 1 mm in thickness, at the tip about 2 mm. Harpoon or knife blades ——Under this heading comes a much larger number of blades of rubbed slate, chipped slate, and chipped chert or jasper. a. Rubbed slate—Those of rubbed slate are tanged, symmetrical in outline, and usually have a median ridge extending from tip to about center; below the ridge and to each side of it the blade is bev- eled, producing a three-faceted surface (pl. 39, figs. 6-12). The average length is 5 to 6 cm and breadth 3 to 3.5 cm. The greatest thickness is at the lower or proximal end of the median ridge, where it averages 4 to 6 mm. Undoubtedly, some of these were used as har- poon blades. Many of the Old Bering Sea harpoon heads have blade slits 3 mm wide and up to 4 cm long, which would require a blade of just this length and thickness but which would be too large for the very thin blades of the type shown in plate 39, figures 1-5. On the other hand, it is also certain that some of these larger tanged speci- mens were used as knife blades (e. g. pl. 38, fig. 1), and since there seems to be no way of distinguishing between the two, they are here grouped together. Plate 39, figure 13, is the lower end of a large slate blade for either a knife or a lance. Blades of this size and shape were rare at the Hillside site. b. Chipped slate—The 76 blades of chipped slate from the Hillside site are difficult to classify because of uncertainty as to which are finished and which unfinished specimens. In plate 40, figures I-10, are shown some of the tanged forms, most of which were no doubt used as knife blades, although plate 40, figure 8, might possibly be an unfinished blade of the type shown in plate 30, figures 6-12. The two slate blades, plate 40, figures 20-21, are unfinished forms or rejects, and there are numerous others similar in shape but smaller, which may have been arrow points or unfinished harpoon blades of the thin, small, tangless type. c. Chipped chert or jasper.—There are 20 of these, the types of which are shown in plate 40, figures 11-19. The smaller blades, such as figures 12-14, were probably used as arrow points, and figures 15 and 16 may have been side blades for harpoon heads. The larger ones, figures II, 17-19, were no doubt knife blades. NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 149 Knife blades. a. End blades for men’s knives—Among the rubbed slate blades I have definitely classed as end blades for men’s knives only those which are (1) slender with one edge straight or almost so and the other curved inward or outward (pl. 39, figs. 14-17), and (2) slender with symmetrical edges but too long to have been suitable for harpoon blades (pl. 39, fig. 18). All seem to have been tanged and most of them have a median ridge, which, however, is not very prominent. b. Blades for ulus or women’s knives—The ulu blades, 30 1n number, are all of rubbed slate (pl. 39, figs. 23-25). The prevailing shape is oval to rectangular, and the cutting edge is either curved or straight. Figure 23 is unusual in shape, being very high and with the upper edges thickened and smoothed where ordinarily they would have been left unfinished ; the smooth, wide, rounded upper end sug- gests that it was used without a handle. A few other blades, of which figure 24 is an example, have two cutting edges, and these also may have been used without handles. Implements with rubbed edges ——Under this heading are four slate artifacts of unknown use (pl. 39, figs. 19-22) representing a type of implement which does not appear to have been described previously, although the smaller end is suggestive of certain “ drill points’ from Greenland (Solberg, 1907, pls. 5 and 6). They are flat pieces of ground slate, with one straight smoothed edge (the left as shown on the plate) about 3 mm thick and 3 to 3.5 cm long. The tip is rubbed down, leaving a short upper edge 7 mm to 1 cm long which joins the longer edge either at a right angle or a 45° angle. Below this the right edge is smoothed in a similar manner, leaving one or two straight sections (pl. 39, figs. 19, 20) or a somewhat rounded edge (fig. 21, 22). The function of such implements is uncertain. The smoothed edges could well have resulted from their use as rubbing stones but the peculiar shape is surely intentional and suggests that they had a specific use, possibly in connection with the dressing of skins, or as boot sole creasers. Drill points —Seven stone drill points, two of them problematical, were found at the Hillside site; all are shown in plate 41, figures 1-7. Two are of ground slate, three of chipped slate, one of chert, and one of quartz. Plate 41, figure 1, is a four-sided point of rubbed slate, the sides of which taper very gradually down to the tip; the tip is rubbed, producing three facets and one short sharp cutting edge. Plate 41, figure 2, is a natural quartz crystal, columnar, and hexagonal in cross- section. Its use as a drill is uncertain ; its shape is perfectly adapted for the purpose, although it shows no certain signs of wear. Plate 41, I 150 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 figure 3, is included among the drills because of its shape, although it is quite possible that it was used as a scraper. It is of dark gray chert roughly four-sided in cross-section and is rubbed down to a chisel- like edge at both ends. Plate 41, figures 4 and 7, are of chipped slate, somewhat similar in shape to figure 3, although somewhat flatter ; figure 4 is the only one of the seven drill points showing unmistakable signs of wear. Slender drill points of this type were undoubtedly hafted, but those with widely flaring upper ends such as plate 41, figures 5 and 6, were probably used as hand drills or perforators, instead of being set in a wooden shank; figure 5, has been rather care- fully chipped all over and brought to a long tapering point. Figure 6 is a flake, only the tip having been chipped out ; the upper curved edge has been retouched, suggesting that it also served as a knife or scraper. Gravers.—E leven artifacts are identified as gravers. The character- istic feature is one or more sharp points, even though there may be chipping along other edges or over the entire surface. Plate 41, figure 8, is made from a thin flake of dark-colored chert. The sharp tip has been smoothed somewhat from use. On the side visible on the plate a curving edge has been chipped’a short distance up from the point while on the reverse the straight edge has been chipped from tip to upper corner. The remaining gravers were fashioned from flakes of jasper, chalcedony and quartz. Plate 41, figure 13, of white, translucent quartz, has a thin cutting edge 5 mm wide which is worn smooth through use; it has not been worked otherwise. Plate 41, figures 9-12 and 14, are thick flakes. Figure 12 is chipped on both surfaces, the others only on one surface. All have finely rechipped edges, indicating that they were used for cutting or scraping as well as engraving. Figure 14 has been chipped with more than ordinary skill; flakes have been thrown off along an abrupt curving ridge which runs from tip to tip; the upper shorter tip is slightly worn, the lower broken. The tips of figure 12 and two others not figured are also considerably worn. Plate 56, figures 1 and 2, are two engraving tools of jasper, from Miyowagh, complete with wooden handles and baleen lashings. These show the probable manner in which stone gravers such as plate 41, figures 10-13, were hafted. Scrapers. a. Side scrapers—In dealing with stone artifacts with chipped edges it is not always possible to differentiate between those used as scrapers and those used as knives. I have, therefore, arbitrarily classed as side scrapers all irregularly shaped stone implements—almost all of them flakes—which have been chipped along one or more edges, even though some were in all probability designed for cutting rather NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 15] than scraping. Those that are relatively thick, with an abruptly chipped concave edge, may safely be classed as scrapers; in this category come those shown in plate 41, figures 15-26. Plate 41, figures 18, 19, and 20, are included, for even though they are straight-edged or almost so, the abruptly chipped edges would be poorly adapted for cutting. The thinner flakes, such as plate 41, figures 27-31, are more doubtful, for these, with their finely chipped, relatively straight edges, could have been used effectively as knives. On these five implements a thin and already sharp edge has been chipped just enough to prevent dulling. The chipping on the others, plate 41, figures 15-26, seems to have been designed to produce a strong rather than a sharp edge. Figures 15 and 16 have been chipped only along one edge, the left; figure 17, on two edges; figures 18 and 19 are chipped along three edges ; and figures 20 to 23, all the way around. All of these flakes are more or less concave-convex, having been struck off from a core, and the chipping is invariably restricted to the edges of the convex surface. Plate 41, figures 24, 25, 26 are not flakes but cores showing all-over chipping. On plate 42, figures 1-3, are shown three-sided scrapers which differ from those described above in that edges are chipped on both surfaces. Plate 42, figure 1, is a piece of dark-colored silicious slate, with chip- ping along the curved edge on the side that is visible and on the straight edge on the reverse side. Plate 42, figure 2, of greenish prase, is triangular in shape with one abruptly chipped edge shown at the left and a similar one opposite it on the other side. Plate 42, figure 3, is a flake of jasper with chipping along all edges; on the convex side, which is visible in the photograph, chipping extends along the left edge and around the bottom; about half way up the right edge the chipping on the convex surface stops but continues on the opposite or concave surface from that point around the upper right corner and along the straight upper edge. b. End scrapers——This is the well known type of scraper that is referred to under various names, such as “ turtle-back”, “ hump- backed”, “ snub-nosed ”, “ thumb-nail”’, etc. Twelve were found, seven of which are illustrated in plate 42, figures 4-10. They are made from thick heavy flakes of jasper, the concave or under side being left unworked while the convex side is chipped either along the lower rounded edge or over the entire surface. The more finished examples of this type of scraper are thickest at the lower end, just above the blunt, steeply sloping edge. All-over chipping of the con- vex surface is shown in plate 42, figures 4, 6, and 7. The lower ends of figures 4 and 5 are thick and wide, the upper ends tapering. Figures 152 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 6 and 7 are shorter, and are rounded or squarish in outline. Figure 8 has the rounded sloping lower edge characteristic of end scrapers, whereas the chipping along the two other edges suggests that the im- plement may also have served as a side scraper. Plate 42, figure 9, is a flat type of end scraper, unfinished, with the shape blocked out but with only crude primary chipping along the edges. Figure 10 also appears unfinished, although the polished and worn lower edge shows that it has been put to use. The flat slate implement, figure IT, is included among the end scrapers because of its shape, although the chipped edges indicate that it has been used also as a side scraper. Adzlike scrapers.—We now come to a type of implement which is characteristic of the Old Bering Sea culture but which, as far as I know, has not been reported from outside the Eskimo territory. In general shape and size these resemble an adz blade with a rough chipped body and a straight, sharply beveled, smoothed lower edge (pl. 42, figs. 12-14). The edge, however, instead of being merely beveled or sharpened as on the usual adz blade, is very abrupt, in some cases being ground down practically flat so that it forms a right angle with the sloping lower portion of the blade. Four scrapers of this type, from Miyowagh, are illustrated in text figure 16. In plate 42, figure 13, the under side meets the straight smoothed lower edge at a right angle, and the upper side, visible in the photograph, descends almost as abruptly. The exact function of implements of this type is not clear, although a reasonable assumption would be that they were used in scraping skins or wood since the edges are too abrupt to have been effective for cutting. Furthermore, the edges are so sharp and perfect that they could hardly have been used for scraping anything as hard as bone or ivory. Another feature of these adzlike scrapers may be seen in plate 42, figure 12 (of dark-colored chert), namely, the smoothed and sharp- ened upper end. Unlike the abruptly beveled lower end, it 1s smoothed down from both sides, producing an ordinary cutting or scraping edge. The fact that both ends are fashioned for use would seem to show that it was not hafted but held directly in the hand. Figure 14, also of chert, is somewhat different in shape, the sides tapering down to the scraping edge, which is only 1.8 cm long instead of the customary 3 to 4 cm; the upper end appears to have been broken off. Figure 15, of prase, differs from the first three in having a thickened lower end which descends abruptly to the scraping edge, in this respect resembling the end scrapers shown in plate 42, figures 4 to 7. Figure 16, of prase, is very similar, although it is not clear whether the chipped lower edge is unfinished or was used as it is; the upper edge is sharpened as on figure 12. en NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 153 In plate 42, figures 17-19, are shown three smaller implements of greenish prase, with four, six, and four faceted edges, respectively. Although at first glance these might appear to be the broken ends of scrapers of the type described above, the arrangement of the faceted surfaces shows that they were used just as they are. In figure 18 there is a smooth, perfectly flat lower edge 12 mm wide and 36 mm long at right angles to the smoothed part of the outer face. The smaller end has also been rubbed dewn, leaving a small smoothed surface Fic. 16.—Adzlike scrapers of stone from Miyowagh. which meets the other two at right angles as in the case of the thin slate forms shown in plate 39, figures 19-22. Just above the squared- off end is a fourth smoothed surface, rubbed down to a 45° angle. Figure 17 is generally similar in shape, but the faceted surfaces, except those at either end, are not perfectly flat and level; the upper surface is slightly concave while the two along the lower end are somewhat rounded or convex. Both corners are squared off, and at the smaller end a small area on the back side has been rubbed down, making in all six faceted surfaces. Figure 19 has four smoothed 154 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 surfaces, tapering to a point below. Figure 20 is a piece of carbon- aceous shale with four faceted surfaces, which indicate its use as a rubbing tool; a deep groove along one edge is evidence of sawing or rubbing with an edged implement. Adz blades—Six adz blades were found at the Hillside site, the type being illustrated in plate 42, figure 21. This was made from a small beach boulder of basalt, the natural surface of which forms the outer face. It has been shaped by crude chipping on the back and along two edges. The cutting edge is somewhat rounded and — beveled, meeting the flat edge of the back side at an angle of about 45° Rubbing stones and whetstones——Rubbing stones of different ma- terials and a variety of forms were found at the Hillside site. The most striking of these were columnar or pyramidal blocks with five smoothed surfaces. Five of these were found at the Hillside site (pl. 43, figs. 1-5) and several others at Miyowagh. Figure 1, of basalt, is shown resting on its broadest surface, which is 4 cm wide at the upper end, tapering to a width of 3.5 cm at the lower end. The right side, which is seen foreshortened in the photograph, rises vertically, while the left side rises at a 60° angle. The right side is slightly wider than the left, the average widths (at center) being 3 cm and 2.7 cm respectively. The two upper surfaces, averaging 2.6 cm in width, rise from the sides at an angle of 45°. All five of the slightly tapering sides are smoothed down to a perfect plane; the two ends are unworked. Plate 43, figure 2, of vesicular basalt, is identical with plate 43, figure 1, in the arrangement and proportions of the five smoothed surfaces. It differs in that it is less tapering. The upper end is the rounded water-worn surface of the original beach boulder from which it was fashioned. The lower end is broken. Plate 43, figure 3, is of fine- grained sandstone, pyramidal or tapering in shape, with the five sur- faces rubbed down at the same angles as plate 43, figures 1 and 2. Figure 4, of basalt, is still more distinctly pyramidal in shape, but here again the angles formed by the faceted areas are essentially the same as those already described. Plate 43, figure 5, is made from a small, rounded beach boulder of vesicular basalt. It also has five smoothed sides, but the inclinations of these differ somewhat from those just described. There is also a rounded unworked space between two of the facets. Plate 43, figure 6, is a piece of scoria with three sides and the smaller end smoothed down. Plate 43, figure 7, is a block of basalt with the two larger surfaces smoothed. Unlike the rubbing stones described above, on which the sides were smoothed down to a perfect plane, there is here a slight concavity of the smoothed surfaces, indicative of a whetstone. The NO:, I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 155 same function should probably be assigned to plate 43, figure 8, for the surface is slightly concave. One edge is also smoothed. The ma- terial is a fine grained sandstone. Plate 43, figure 9, is a small rectangular piece of sandstone with the sides, edges, and one end smoothed ; the opposite end is broken. Plate 43, figure 10, is a slab of diorite which has evidently been used as a whetstone to judge from the slightly concave rubbing surface. The opposite side has also been rubbed to a slight extent. The two pieces of scoria shown in plate 43, figures 11 and 12, were no doubt used for rubbing skins. The first is wedge-shaped, with the thicker end (below) rounded off. Figure 12 has been worked on all surfaces except the upper end. Even the edges, which have a breadth of 2.4 to 2.7 cm, have been rubbed flat. Plate 43, figure 13, is a small beach boulder of basalt, unworked except for a small flattening of the lower end, produced by rubbing. A shallow depression just below the center and two similar ones on the opposite side allowed a firm grip. The flat beach boulder of diabase shown in plate 43, figure 14, was probably used as a hammer stone; the lower end is broken. Finger grips have been pecked in on both sides near the upper rounded edge, and smaller pitted areas occur lower down. The upper edge and parts of the lateral edges show slight signs of wear; there is also a small notch on one edge opposite the finger grip. SLEDGES, Boats, Evrc. It is a point of considerable importance that at the four oldest Gambell sites no objects were found which could be recognized as having been used in connection with the modern type of dog sledge— no flat bone shoes for the built-up sledge, no trace buckles, swivels, or ferules for whip handles. As we shall see later, such objects are not found at any of the older Alaskan sites, although they are common at the late prehistoric and modern sites. If we may judge from the Gambell finds, the sledge of the Old Bering Sea Eskimos. was very similar to the short low type with heavy ivory runners such as is still used on St. Lawrence Island and at Point Barrow for hauling umiaks and loads of blubber and meat over the ice ( Nelson, 1899, pl. 76, fig. 1 ; Murdoch, 1892, p. 358). Sledge runners.—Six ivory sledge runners were found at the Hill- side site, five of which are shown on plate 44. They are of two dis- tinctive types. Type 1, illustrated by plate 44, figures 3, 4, 5, is a heavy runner made from a walrus tusk which was but little modified. 156 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL, 96 At the upper or proximal end of the tusk a single large hole was cut through, and above it a wide shallow notch was made for holding the end of the cross bar. There is no certain evidence that more than one cross bar was used; only at the lower end of figure 5 is there a small roughened area which may have been made for the attachment of another cross bar. The lower ends of the two tusks may have been fastened together with a lashing. A characteristic feature is the flatten- ing of the under side which seems to have been caused entirely by wear, as there is no evidence of cutting or shaping; figures 4 and 5 show a pronounced flattening of the under side, figure 3 only a slight flatten- ing. There is some question as to whether the latter was really used as a sledge runner. The large hole and notch above it show that it was intended for such a use, but the beveled and sharpened lower end shows that it was also used as a pick. In addition, one side (visible in the photograph) has been adzed down to a flat surface, apparently in order that it might serve still another purpose, possibly as a sleeper or prop for timbers. The other type of runner, type 2, is made from a section of tusk, its height being about the same as the original diameter of the tusk, its width or thickness around 1 cm. The three other sledge runners from the Hillside site are of this type, two of them being illustrated in plate 44, figures 1, 2. Figure 1, which is rather badly warped and weathered, is a thin, narrow strip of ivory 44.5 cm long and 1.4 cm thick ; one end, broken, is 5 cm high, tapering to 2.5 cm at the opposite end, at the first hole. Near the upper border are five round holes (probably six originally) about 6.4 cm apart, and above each of them a shallow flat-bottomed notch for holding the end of the cross bar. Figure 2 is a broken runner of the same type, 25 cm long, 4.8 to 5.8 cm high and .g cm thick. It differs from the other in having the lower edge somewhat beveled so that in use it leaned slightly outward in- stead of upright. The distribution of all of the sledge runners and shoes from Gambell is given in tabular form on p. 230, following the description of the Punuk types. In plate 45 are shown nine sledge runners and shoes from Miyowagh. In all, 48 complete and fragmentary specimens of ivory were found at this site, including several types which were not represented at the Hillside site. In addition to the ivory runners there were two whale ribs, which by their notched ends and smoothed and worn outer surfaces are shown to have been used as sledge runners. The type 1 sledge runner of ivory, described above, was represented at Miyowagh by only two examples. However, there is a modification NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 157 of this type, here designated as type 1 a, and illustrated by the two run- ners shown in plate 45, figures 1, 2, and four others not shown. They have the large hole and the wide notch of type 1, but they differ in that they have been cut down to the same shape as type 2. Type 2 is represented at Miyowagh by five examples, including the well pre- served specimen shown on plate 45, figure 3. It is 26cm long, from 5.7 to 7.9 cm high, and 1.3 cm thick; it differs from the two runners of the same type shown on plate 44, figures I, 2, by having a greater number of holes—nine, each with a shallow notch above, and two additional holes at one end and one at the other. Sledge shoes—Oft sledge shoes there were 31 specimens, all of ivory and only one of them complete. Type 1, the most common form (14 specimens) is illustrated by the two shown in plate 45, figures 4, 5. The upper surface is flat, but the lower edges are somewhat curved, giving a slight convexity to the under side; they range from 5 to 7.4 cm in width and from 1.2 to 2.3 cm in thickness. They lack the peg holes found on modern sledge shoes, the only method of at- tachment being a pair of round holes connected by a sunken groove on the bottom to protect the lashing. The complete specimen, figure 4, shown with the lower side up, has in addition two holes at the back end with grooves leading from each, showing that it has been attached to a similar shoe. The type 2 shoe (pl. 45, figs. 6, 7, 8—four specimens in all) re- sembles type 1 as to thickness and width, but has an unusual arrange- ment by means of which it is attached to the runner—a knoblike elevation at the end 1 cm or more high, beneath which passes a trans- verse perforation. The type 3 shoe (13 specimens) is illustrated in plate 45, figure 9. It is of about the same thickness as the preceding types but somewhat narrower. The method of attachment consists of a single hole at the end; this hole either extends straight through the end, which slants upward from the bottom so as to protect the lashing; or it is drilled through at an oblique angle, as in figure 9, connecting with a sunken groove on the lower side for holding the lashing. From the later, northwestern section of Miyowagh there are four sledge shoe-runners of Punuk type 1 a, which will be described later. None of the sledge shoes above described conform to any known types, ancient or modern. Rather large sledge shoes, usually of bone, are known from the Thule culture, and some of them have pairs of holes connected by sunken grooves for a lashing; but on the whole they seem to have been longer, narrower, and thinner than the St. Lawrence forms, in addition to which they have holes for pegs as the 158 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 usual means of attachment to the runner (Mathiassen, 1927, vol. 2, pp. 60-61). We know nothing as to the nature of the runners to which these heavy ivory shoes were attached ; most likely, however, they were short and low, forming with the shoe itself a composite runner com- parable to those described above. Cross bar.—Plate 50, figure 3, is a sledge cross bar, made of walrus penis bone, from Miyowagh; both ends are notched and roughened and the under side is flattened where it rests on the runner. Toy sledges—The parts of a toy sledge of wood, all found to- gether in cut 25 at a depth of 43 inches, are shown in plate 59, figures 8-10; the two cross bars (fig. 10) were lashed to the runners (figs. 8, g) just above the holes. Figure 11 (cut 18, 12 in. deep) is a toy sledge runner of ivory, of the type illustrated in plate 44, figures 1 and 2, and plate 45, figure 3. Baleen toboggan.—The Old Bering Sea Eskimos also used the baleen toboggan, such as is known from the Thule culture and the Sadlermiut and which the modern St. Lawrence and Point Barrow Eskimos use for hauling loads of meat and blubber over the ice (Mathiassen, 1927, vol. 2, p. 45; Boas, 1901, p. 71; Murdoch, 1892, pp. 356-357; Bogoras, 1904-09, p. 107). None of the baleen slabs which formed the toboggans were found at the Hillside site, but two cross pieces show that they had been used. These are strips of baleen slightly more than 30 cm long and 6.5 cm wide with three pairs of round holes by means of which they were lashed to the slabs of baleen beneath. At Miyowagh a number of these cross pieces were found, in addition to one practically entire toboggan and numerous fragments. The most complete specimen was the one from cut 2, shown 1 situ in plate 6, figure 3. It is 18 inches wide at the front end and slightly more than 6 feet long. Umiak paddle——tIn plate 50, figure 1, is shown a section of an umiak or kayak paddle, from cut 23, depth 81 inches. It was 8 cm wide (one edge is broken off), and the entire surface is covered with a coat of red paint, made from pulverized hematite or ocher. Plate 37, figure g, is one of the cross braces for the framework of a kayak, from cut 19, depth 59 inches. Umiak keel—The wooden object shown in plate 46, figure 14 (cut 18, depth 44 in.), is probably the end of an umiak keel; there is a transverse groove across the thickened front end; below this is a round hole and at the opposite end a smaller perforation. Plate 59, figure 3 (cut 18, depth 32 in.), shows a replica of such a keel piece, from a toy boat; a baleen lashing remains in place in the hole at one end and the opposite end is spliced by means of a carefully made dove- NO; I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 159 tailed joint. Plate 59, figure 2 (cut 13, 54 in.) is one of the cross pieces from the bottom of a toy umiak; figure 7, from cut 18, 32 inches—of bark—is one of the end blocks, or seats, for the prow or stern. A general idea of the external appearance of the boats may be ob- tained from the two toys shown in plate 59, figures 1 and 6, from cut 18, depth 44 inches, and cut 19, 54 inches respectively ; the first repre- sents a kayak; the second an umiak, of bark. Plate 59, figures 4 and 5 (cut 18, depth 44 in., and cut 6, 42 in.), seem to be stylized repre- sentations of the man seated in the kayak, the projecting upper part possibly representing a hunting helmet. Ten other toy wooden boats were found at Miyowagh, mostly from the lower levels of the midden, as follows: Cut 10, depth 50 inches ; cut 18, 32 inches (2) ; 44 inches; 53 inches; cut 19, 48 inches; cut 24, 64 inches; 72 inches; cut 25, 84 inches (2). Two fragmentary toy boats were found at the Hillside site (pl. 30, fig. 25, from house no. 2; and fig. 24, from beneath the floor stones of the same house). ApzeEs, Picks, WepGEs, Etc. Adze handles and heads.—In plate 46, figures 1-7, are shown several types of Old Bering Sea adz handles and heads from Miyowagh. Figure 1 is an ivory handle from cut 19, 76 inches deep. It bears an Old Bering Sea decoration consisting of narrow bands of obliquely incised parallel lines and four nucleated, concentric circles which are given the appearance of scrolls by the attachment of pairs of short parallel lines. The handle is broken off at the notch which formed the finger grip. The upper end was made to rest in an oval pit on the adz head and was held in place by lashings which passed through two cir- cular holes. The fragmentary ivory handle, figure 2 (cut 9, depth unknown), has two deep, wide finger grooves; it seems to have been rectangular at the upper end; it was lashed to the socket, the lashing passing through the large triangular opening at the end. The only adz handle found at the Hillside site is the well-preserved wooden specimen shown in plate 29, figure 16. It came from house no. 2. It is slightly curved, 24 cm long, square in cross-section with rounded edges. The lower end is slightly enlarged, the upper end is cut square to fit the hole in the socket, and a small wooden wedge, still in place, was driven in to insure a tight fit. Plate 46, figure 3 (purchased from an Eskimo) is a rather elaborate adz head of ivory. The stone blade was attached to the lower end, which was sunken for the purpose, and the lower edge on the opposite 160 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 (back) side was provided with a lip or ridge to hold the lashing in place. The end of the handle rested in a shallow, oval-rectangular con- cavity on the back side and was held in place by thongs passing around the two circular elevated bosses which extend from the back side around to the front; there are also four pairs of lashing holes on the back side. At the center of the two bosses are circular pits 3 mm deep, in one of which is a wooden plug; a row of smaller pits extends along the periphery, with five more in between the bosses. The upper end, now broken, seems to have been bilobed. Figure 5 (cut 25, depth 84 in.) is a wooden adz head with a deep oval depression on the. back side for the end of the handle, and an opposite groove for the lashing ; on the opposite—or front—side there is a deep sloping concavity for the blade. Figure 4 (cut 18, 25 in.) shows the front side of an ivory adz head with a slightly sunken bed 6.8 cm long for the blade; on the opposite side is a raised lower edge to hold the lashing in place. On the back side the upper end is cut down slightly to provide a resting place for the end of the handle ; opposite this is a deep wide groove for the lashing. Figures 6 and 7 are two adz heads made of whale rib (from cuts 15, 33 in., and cut 24, 52 in. deep, respectively), which ditter from those just described in that the blade was set in a deep socket in the end instead of being lashed on. In the first of these there is a transverse oval hole for the handle, in the second a rectangular pit with an opposite lashing groove. In addition to those illustrated there are four other adz sockets with a transverse hole for the handle; these are from cut 7, 51 inches deep; cut 15, 40 inches; cut 16, 30 inches; cut 18, 49 inches. There are also six others with a sunken bed for the end of the handle, from the following locations: Cut 4, depth 55 inches; cut 7, 51 inches; cut 9, 25 inches; cut 17, 27 inches; cut 18, 42 inches; cut 23, 81 inches. Pick handle.—Plate 46, figure 8, is a wooden pick handle from cut 27, 46 inches deep. The lower, enlarged end is perforated for a sus- pension thong, and there is a deep notch or finger rest at the center ; the upper end is somewhat concave and slanting so that the bone or ivory pick, which must have been a light one, might project at an angle; there is a wide groove on the back side of the upper end for the lashing which held the pick in place. The implement was probably used for digging roots. Ivory picks—Heavy ivory picks are among the most common implements of the Old Bering Sea culture. They were made from a whole walrus tusk or a large section, sharpened at the lower end and designed for digging in frozen ground or ice. Seven of the picks from the Hillside site are shown in plate 49. The lower end was NO: I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 161 given a rounded cutting edge by beveling from both sides. In a few instances the ends are pointed; one of these, plate 49, figure 3, also differs in the manner of hafting: at the upper end of the convex side there are two deep grooves separated by a prominent ridge, and the opposite side has been hacked to a flat surface for a distance of 10.5 cm in order to provide a firm resting place for the end of the handle. Usually the upper end of the pick was prepared for the handle by scarfing along one side with the opposite side roughened to hold the lashing (pl. 49, figs. 2, 5, 6). Some, like plate 49, figure 1, have two large widely spaced grooves on one side with a single groove opposite. The most finished type is illustrated by plate 42, figure 4, which has two deeply cut, rounded, transverse grooves. Plate 49, figure 7, was lashed to the handle by means of two small notches. Ivory wedges.—Ivory wedges, mainly for splitting wood, were found in large numbers both at the Hillside site and Miyowagh (pl. 47, figs. 8,9). They are usually made from the distal end of the tusk, which was beveled to a rather sharp edge; the upper end usually shows signs of hammering. Such wedges were often used secondarily as drill rests, as in the case of one of those illustrated. The average length is around 12 cm. In plate 29, figure 17, and plate 47, figures 12 and 13, are shown three wedge-shaped ivory objects from the Hillside site and Miyowagh respectively, which are characteristic of the Old Bering Sea culture, but the use of which is uncertain. Plate 47, figure 12, is 6 cm long; the upper end is 1.4 cm wide by 1 cm thick, and the lower end 2.3 cm wide by .3 cm thick. On both sides the right half has been cut down lower than the left, thus producing two opposite offsets which would help to hold the object firmly in place if it were inserted wedge- like in a socket shaped in a similar manner. The small hole in the lower, thin end was probably for a peg or rivet for fastening to the socket. Figure 13 is made in the same way, but is somewhat thicker and less tapering. Plate 29, figure 17, from the Hillside site, shows the same characteristic features. It is possible that these objects were wedges which were driven into the end of the wooden adz handle to hold it tightly in place in the socket. Mattock blade.—Pilate 50, figure 2, is a broken mattock blade, made from a section of a whale rib. The upper end is roughened for the lashing. Snow shovels.—In plate 50, figure 6 (cut 18, depth 49 in.) is an example of the Old Bering Sea type of snow shovel, made from a walrus scapula. The concave inner surface, shown in the photograph, served as the front or upper side of the shovel; it was smoothed down 162 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 by cutting away the processes at the proximal end, the longitudinal crest or spine, and the thickened opposite edge. The handle was lashed on by means of two round holes. Bone knife.—Plate 50, figure 4, is a bone knife from cut 19, depth 67 inches, the principal use of which, if we may judge from modern analogy, was to scrape the snow from fur clothing before entering the house. It is made from a section of walrus penis bone, split length- wise, flat on one side, convex on the other. The four notches and intervening knobs at the handle are suggestive of those found on modern ivory snow knives from western Alaska, and also of those on a bone knife for squeezing out water, which Mathiassen describes from the Thule culture (1927, vol. 1, p. 48, pl. 13, fig. 14). The present specimen may represent the prototype of these knives; at any rate, the small knobbed grip and the curving shape would have ren- dered it ineffective for cutting or stabbing. It was not a common type, however, as this and a few doubtful fragments were the only examples found at definite Old Bering Sea levels; as we shall see later, another type of bone knife, evidently a dagger, comes in with the Punuk period. Dritts, REAMERS, AwLs, Etc. Drill rests—Ivory drill rests were found in large numbers both at the Hillside site and Miyowagh. Plate 46, figure 13, shows one of these, a section of tusk with five deep holes on one side and three on the other. Drill rests of this kind were held in the hand and pressed down on the upper end of the drill shank as it was being rapidly revolved by the bow which was held in the other hand. Many of them had served more than one purpose; usually they were wedges, like the one shown in plate 47, figure 9, or picks or pieces of heavy sledge runners which later had been utilized as drill rests. They were used no doubt both for fire drilling and for drilling holes in implements. Mouthpieces—Only a few mouthpieces for bow drills were found. Two of these, from the Hillside site, are shown in plate 30, figures 2, 3. The first is a small segment-shaped piece of ivory 5.7 cm long and 2.5 cm wide with just enough of a groove on each side to insure a firm grip with the teeth. Figure 3 is a walrus tooth unworked except for two longitudinal grooves to facilitate grasping between the teeth. Drill bows.—It is not always possible to distinguish between drill bows and handles for pails, boxes, etc. Assuming the longer ones to have been drill bows, however, we see that these were made either of wood or bone. Those of bone were usually made of a section of walrus rib, which already possessed the necessary curve; those of wood are NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 163 usually somewhat less curved (pl. 47, fig. 1) ; all have the ends notched for attachment of the cord. Drill shafts—Numerous wooden drill shafts were found, some of which (from Miyowagh) were painted red. Plate 30, figure 1, is a drill shaft from the Hillside site ; it is round in cross-section, slightly enlarged at the lower end and has the tip charred. Two specimens from Miyowagh are shown in plate 57, figures 2, 3. Drill points—Drill points of bone were found in great numbers at Miyowagh; eight were found at the Hillside site. Most of them were made from seal fibulae (pl. 48, figs. 1-4). The enlarged distal end of the bone, after being trimmed down, was set into the end of the wooden shaft; the proximal end, below a somewhat abrupt shoulder, was made cylindrical and sharpened to a chisellike tip. In some cases, as in figures 3 and 4, the shoulder was less pronounced. Hand drills —Three hand drills are shown in plate 48, figures 5-7. These have very small, sharp points, about 2 mm in diameter; the upper end, which was held in the hand, is unworked. Such drills were usually made from dog or seal ulnae or slivers of ivory. Reamers.—We may consider as reamers those implements having a blunt cylindrical point and an unworked upper end for holding in the hand. One of these, from the Hillside site, is shown in plate 30, figure 15. Four examples from Miyowagh are shown on plate 48, figures 8-9, 13-14. The first is made from the distal end of a walrus rib, the second is of ivory; figures 13 and 14 are also made from sections of walrus ribs. The principal function of such reamers was no doubt that of enlarging or smoothing out holes in skin, wood, baleen, etc. Rubbing or scraping tools——In plate 48, figures 10-12, are shown three examples of an implement which was found in considerable numbers and which seems to have been used primarily as a rubbing or scraping tool, probably for working hides. They are almost invaria- bly made of walrus ribs, with one end smoothed and spatulate, the other unworked as in figure 12, or smoothed and blunt as in figures 10 and 11. The upper, blunt end of such implements would seem well adapted for stone flaking, and they may have been so used. Stone flakers——The three objects shown in plate 48, figures 18-20, may safely be regarded as stone flakers. In all, 19 of these flakers were found at Miyowagh. They are shorter than the type just de- scribed ; both ends are usually provided with a blunt point, less fre- quently one end is somewhat flattened, as in figure 18. They may have been held directly in the hand ; at any rate no recognizable handles were found. Three stone flakers from the Hillside site are shown in plate 30, figures 9-11; 10 others were found at this site. 164 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 Awls.—Bone awls (pl. 48, figs. 15-17) do not show any features of particular interest. Most of them are made of bird bones, usually the humerus of larger birds, such as cormorants, geese, gulls, etc. Figure 15 is made from the ulna of a fox. In addition to those of bone there are also awls made of ivory, usually sharpened at both ends. From the sharp, smooth, well-made examples such as figures 21 and 22, there is a gradation into cruder types—ivory slivers which may have had a wide variety of uses. Two specimens from the Hillside site are shown in plate 30, figures 16, 17. Needles Needles were made of small bird bones, usually ribs. The only complete specimen was one from the Hillside site. It is 3 cm long; the eye is a round drilled hole about half a millimeter in diameter, just large enough to receive the point of a common pin. Women’s KNIVES The ulus or women’s knives from Miyowagh and the Hillside site had all been provided with stone blades; most of the handles were of wood, a few of ivory. In plate 51, figures 1-7, are shown two complete ulus and five wooden handles from Miyowagh which il- lustrate the range of types. In addition to these there were 16 other handles rather uniformly distributed throughout the midden. Most of the handles are straight or only slightly curved, the crescent shape of figure 6 (cut 19, depth 45 in.) being unusual. When there was a tendency for the handle to split, it was grooved at one or both ends and tied (pl. 51, figs. 3, 4—from cut 9, 20 in., and cut 19, 86 in.). Figure 1 (cut 22, 86 in.) is unusual in having the blade held secure by a thong placed around a central groove and through a hole in the blade. An interesting feature, shown in figure 5, from cut 25, depth 33 inches, is a broad groove or notch at one end for the index finger. Figure 7 (which is shown accidentally reversed) has a narrow rect- angular inset of ivory on the top, and there seems to have been an attachment at either end which fitted over a small cylindrical core projecting from the handle; it comes from cut 27, depth 36 inches. Fat SCRAPERS Another typical woman’s implement was the ivory scraper for removing fat from intestines and skins. A few of these approach an oval or rounded shape, somewhat like the modern forms, but the pre- vailing types are flat-bottomed, with upright or slightly flaring walls, and open at both ends; or trough-shaped, with flat or rounded bottom and upturned ends. The edges along the side are sharpened; on the NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 165 type with open ends the ends may or may not be sharpened. Four examples, of the rectangular type, were found at the Hillside site. One is the decorated specimen already described (pl. 13, fig. 8). The other three are shown in plate 30, figures 5-7. Plate 30, figure 5, from house no. I, is 14.1 cm long and about 3 cm wide. The bottom is flat; the sides, which are for the most part broken away, were almost vertical, with a height of 2.4 cm. The one small remaining section of rim is beveled but not sharp. The upper end is slightly upturned, and through it a small rectangular slot has been cut, evi- dently for suspension; the opposite end has no rise, but was cut off straight. Plate 30, figure 6, in fwo pieces, also from house no. 1, is of the same general shape but somewhat shorter and broader. It is 12.6 cm long, but 4.3 cm wide and 1.1 to 1.8 cm high. The sides have a slight outward flare and slope gradually downward from the upper end. The upper end, like plate 30, figure 5, turns upward and is perforated for suspension while the lower end continues level with the bottom. Plate 30, figure 7, from the third level of the midden below house no. 1, is half of a smaller fat scraper which differs from the first two in that both ends turn upward, making it boat-shaped. The drilled hole and the rectangular slot show that it had been mended after breaking. From Miyowagh there are 26 more or less complete scrapers and fragments. Two of these, bearing Old Bering Sea ornamentation have already been described (pl. 17, figs. 1, 2, from cut ga, depth 50 in., and cut 10, 54 in.). Three others are shown in plate 51, figures 8, 9, 11. The first of these (cut 12, depth 25 in.) is rectangular in shape with open ends like the two decorated specimens just mentioned. In addition to plate 51, figure 8, there are nine other fairly complete specimens and five fragments of the same type, distributed as follows: Cut 1, depth 55 inches; cut 3, 45 inches; cut 5, 32 inches; cut 9, 25 inches (2), 36 inches; cut 16, 18 inches; cut 18, 44 inches, 60 inches ; cut 19, 17 inches, 67 inches; cut 22, 18 inches; cut 24, 36 inches; cut 27, 52 inches. Plate 51, figure 9 (exact provenience unknown) and figure 11 (from cut 10, depth 22 in.) are examples of the more rounded type with curved ends. Five other examples of this type were found, as follows: Cut 7, 51 inches; cut 9, depth unknown (2) ; cut 18, 81 inches; cut 19, 67 inches. Another specimen from cut 16, depth 18 inches, differs from the last mentioned type in being rectangular in outline, with a flat bottom and straight deep sides and ends. Lastly, there is to be mentioned a small ivory scraper, or perhaps a cup or ladle from cut 29, depth 17 inches. It is short and deep, I2 166 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 squarish in shape, with a flat bottom, rounded corners, and smoothed edges. Plate 51, figure 10, is a long, narrow section of a beluga jaw, with one edge sharpened for use as a scraper, from cut 10, depth 22 inches. Another, less common type of scraper, found only at the Hillside site, was made from a dog femur, with a section of the shaft cut away, leaving two parallel and fairly sharp scraping edges. Three of these are shown in plate 30, figures 12-14. The first two are from the midden, first and fourth levels respectively, the third from house no. 1; the fourth specimen is also from house no. 1. From Miyowagh, cut 19, depth 51 inches, there is a larger scraper of this same general type made from the metatarsal bone of a reindeer. This is the only example from Gambell of the two-handed scraper, a type which occurs at Point Barrow, among several groups of Central and Eastern Eski- mos and among many Indian tribes (Birket-Smith, 1929, vol. 2, p. 37). LapLEs, VESSELS, ETc. Ivory vessel—The ivory vessel (pl. 51, fig. 12, cut 16, depth 23 in.) in general shape resembles some of the scrapers, but the incurving edges would make it unsuitable for such a purpose. It was a receptacle, probably designed for some special use, as in a ceremony. A line of very small round pits encircles the rim; at the smaller end is the broken remnant of a handle. Ladies —Plate 51, figure 14 (cut 19, depth 30 in.) is a much smaller utensil of ivory, probably a ladle. Figure 15, from cut 19, depth 67 inches, is a small spoon or ladle of antler, with the larger end split and then lashed together with baleen to make it concave. Figures 13 and 16 are two wooden ladles from cut 23, depth 72 inches, and cut 18, 48 inches. In plate 47, figure 16 (cut 23, depth 81 in.) is shown a ladle or spoon of a different type made from a beluga jaw. It has a long, narrow handle, perforated at the end, and a flaring and slightly con- cave lower end, made from the proximal part of the jaw. Three pairs of narrow rectangular slots were cut in the lower end to hold it together after it had split. POTTERY In a treeless region like St. Lawrence Island the oil-burning lamp is of vital importance, performing as it does the indispensable func- tions of ‘heating and lighting the interior of the house, of drying wet NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 167 clothing, and of cooking. From the Old Bering Sea period down to modern times lamps and cooking pots on St. Lawrence Island have been made of earthenware; steatite and sandstone lamps, such as are used by the modern Eskimo from Bering Strait northward and east- ward did not reach St. Lawrence. Considering the numbers of vessels that must have been used and their fragile nature, it is not surprising that pottery fragments should be found in such abundance at the old sites. Potsherds were by far the most numerous of all artifacts found at the Hillside site and Miyowagh. Compared with that from other parts of the world, Eskimo pottery is crude and poorly made. The paste is coarse and rather crumbly in texture, the walls of the vessels are usually quite thick, the surface finish is poor, and there is usually no decoration. The poor quality of the ware was probably due in large part to the absence of suitable clay, although a contributing factor must also have been an imperfect knowledge of the technique of mixing, shaping, and firing. In plate 52 are shown the various types of ware found at the Hillside site. Paste——The prevailing color of the paste is a rich sooty black ; this color, with but few exceptions, is uniform throughout the thick- ness of the sherd, showing that the material from which the pottery was made was originally black and that it had been subjected to very little firing; otherwise the surfaces would have been burned to a different color. Ina few instances this happened, one or both surfaces having been burned to a light buff or gray, with the rest of the paste remaining black. A very few sherds, like plate 52, figure 7, are light brown in color, but this apparently should be attributed to the use of a brownish clay rather than to intensive firing. Temper—The tempering material most commonly used was coarse sand and gravel. The latter seems to have been carelessly selected, for it is often very coarse, containing particles up to the size of a kidney bean. Such coarse tempering material could not have been very effective as a binding agent. Fine grass was also used for tempering. The small grooves left by the grass stems and even the carbonized stems themselves are plainly visible on sherds which have weathered or flaked off (pl. 52, figs. 9, 10). The modern Alaskan [’skimo are known to have made use also of mammal hair in tempering, and the present material has a superficial resemblance to hair. However, microscopic examination shows plainly that it was grass. The sherds with grass tempering are appreciably lighter in weight than those in which sand and gravel had been used, and they show a more pro- nounced tendency to flake off in layers. I have found no instance in which both methods of tempering had been used in the same vessel. 168 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 Thickness.—In addition to its coarse texture, Eskimo pottery is remarkable for its thickness, and in this respect the sherds from the Hillside site are no exception. The average thickness of the ware is 1 to 1.5cm. The thinnest sherd found measured .5 cm, but those with a thickness of more than 2 cm were much more common; one sherd was 2.7 cm thick. Grass-tempered sherds were either thin or medium in thickness; sand- and gravel-tempered sherds were thin, medium, or thick. Method of manufacture—The vessels seem to have been shaped by molding directly from the clay rather than by the coiling process. A single rim sherd (pl. 52, fig. 2) bears evidence of coiling, three distinct strips being visible at the rim. However, this would appear to be an exceptional case, for the many other sherds show no evidence of coil- ing, and their strong tendency to flake off in layers shows that in most cases at least, this method could not have been employed. Shapes—No entire vessels were found, but many of the sherds were large enough to show that they were rounded in shape. The bases and rims were uniformly rounded; not a trace was found of a square, flat bottom or a rectangular rim, features which are invaria- bly present on modern St. Lawrence cooking pots and lamp rests. There was also no evidence of the lugs and suspension holes such as are found on modern St. Lawrence cooking pots, nor of the ridges which are so typical of the modern St. Lawrence lamps. It appears that the lamps were shallow, thick-walled, citcular vessels, somewhat similar to those used by the modern Eskimo of the Yukon-Kuskokwim region, but flatter, with a straight slope from the rim to the somewhat conical base. Plate 84, figure 3, shows half of a lamp of this type from the Punuk site, levoghiyoq. Plate 53, figure 2, is a lamp of similar type excavated by the writer at an old site on Imaruk Basin, Seward Peninsula. The cooking pots were thinner, much deeper, and some- what cylindrical in shape. On many of the sherds there is a thick black incrustation, caused by the burning of oil which adhered to the surface. Surface finish and decoration—The surfaces were finished by smoothing, no attempt having been made at polishing; a slip or wash was never applied. Although the pottery was for the most part un- decorated, a large number of the sherds were corrugated on the ex- terior, a decoration which had resulted from the application of a paddle or other implement bearing parallel grooves. One of these wooden pottery paddles, from Miyowagh, is shown in plate 47, figure 17. It is 27 cm long and 8.4 cm wide at the lower end; the back side is NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 169 convex; the front, with six deeply cut, parallel oblique grooves, has just enough of a concavity to fit the sides of a large rounded vessel. A large sherd, picked up from the surface at the northwestern sec- tion of Miyowagh, was decorated with a typical “ check-stamp ”’ pat- tern (text fig. 17). The indented squares were rather deeply im- printed and of uniform size, measuring about 6 mm across. pteeak oo nae aw Lr Fic. 17—Potsherd with check-stamp decoration, Miyowagh. In plate 58, figure 14, is shown the only example of a molded decora- tion; this is a small human figure applied to the inner rim of a thick- walled vessel, apparently a lamp. BALEEN AND WOODEN VESSELS Parts of baleen vessels—the round or oval wooden bottoms and the flattened sides of baleen—were found in considerable numbers at Mi- yowagh. Four were found at the Hillside site, one of them the well- preserved specimen shown in plate 53 found together with three others in what appeared to have been a small storage space beneath the rear floor stones of house no. 2 at the Hillside site. It is straight-sided, 8.5 cm high, with an oval wooden bottom 9.4 and 10.7 cm in diameter. The upper part of the vessel is formed of a piece of baleen bent around and held together by an in-and-out stitching of baleen. The wooden 170 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. g6 bottom is held securely in place by four long stitches of the same strip of baleen, which passes in and out through four rather widely spaced splits in the wood. The bottom is set about 1 cm above the lower edge of the baleen, resting in a shallow groove which extends all the way around. Most of the baleen vessels from Miyowagh and the few others from the Hillside site differ from this one in that the baleen stitching which holds the ends together is narrower and is set in a narrow shallow groove; occasionally there is a double row of stitching. Some vessels were made higher by having another band of baleen of the same size attached to the lower one. !Vooden bottoms.—Three wooden bottoms for baleen vessels, from houses nos. I and 2, Hillside site, are shown in plate 54, figures 1-3. The first of these was attached to the upper part of the vessel by means of loops of baleen, one at either end; the three circular holes along the left edge and the one opposite were made to hold the pieces together after they had split. Plate 54, figures 2 and 3, were attached in the same manner as plate 53 as shown by the remaining slits and pieces of baleen lashings. Plate 54, figure 3, which is thicker than the others, has a small wooden peg set in the edge which no doubt helped to hold the bottom in place. Handles of wood or bone-——Some, at least, of the baleen vessels had handles of wood or bone, of which the three shown in plate 47, figures 2-4, are examples. The first of these, from cut 23, depth 72 inches, is of whale bone; it is widened at the middle and a decorative effect is provided by four rather deeply cut lines; figures 3 and 4 (cut 25, 89 in., and cut 2, 4o in.) made of a walrus rib and wood, respec- tively, are undecorated. Wooden vessels —Two fragments of wooden vessels from the Hill- side site are shown in plate 54, figures 4, 5. The first is crudely made, with thick uneven sides and bottom; it seems to have been narrow and rather shallow. Figure 5 shows better workmanship, the cutting being smooth and even on both sides. The bottom has a thickness of about 5 mm, the rim of about 7 mm. There are three small pits along the rim which appear to be intentional. From Miyowagh there are several fragments of long, trough-shaped wooden vessels. OsjECTS OF BALEEN Baleen was relatively scarce at the Hillside site. Besides the baleen vessels, the collection contains several toboggan cross pieces, and a larger number of knots, strands and fragments of various sizes, but nothing approaching the quantities of this material found at Miyowagh and the later sites. The same was true of whale bones, only a very NO; I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS Ws few being found at the Hillside site. On the whole it would seem that whaling, if practiced at all, was only incidental with the earlier Old Bering Sea Eskimos; certainly it did not occupy the prominent place it came to have later. The baleen objects described below are all from Miyowagh. Ice scoop.—Plate 55, figure 7, from cut 18, depth 44 inches, illus- trates the type of the Old Bering Sea ice scoop, made entirely of baleen. The rim and part of the handle consist of a single strip about 2 cm wide bent around and forming a triangle with rounded corners. The netting is of smaller strands of baleen, one looped around the other, the loops occurring at regular intervals and radiating in nine bands from the center. The netting is fastened to the rim at 11 places and is further strengthened by being looped around a straight strip which extends from the apex to the opposite side. Four more baleen ice scoops were found at Miyowagh, two in cut 9, depth unknown, one in cut 18, depth 48 inches, and one in cut 19, depth 45 inches. Ventilator rim (?)—The heavy grass ring with baleen wrapping (pl. 55, fig. 4—cut 23, depth 32 in.) was thought by the I:skimos to be a border or rim for the circular opening which serves to ventilate the inner sleeping room. In the modern houses the opening is cut in the upper part of the reindeer skin which hangs as a heavy curtain and forms the front of the sleeping compartment. A similar grass ring was found in cut 10, 40 inches deep, and also one made from the epi- physis of a whale vertebra, in cut 23, depth 81 inches. These grass rings also resemble the hoops used by the Plains Indians in various kinds of hoop and pole games (Culin, 1907, pp. 458, 483, 495-6). Plate 55, figure 2 (cut 23, depth 39 in.), is a slender strip of baleen, rounded in cross-section and pointed and notched at one end. Both ends appear to have been bent over and may possibly have been held by the baleen loop placed somewhat below the center. It may have been part of a snare. Baleen knots.—Plate 55, figure 3, is one of a great number of baleen knots, the exact use of which is unknown; this specimen con- sists of two connected knots, both formed of two layers of square plaiting, the same technique as shown by the two smaller baleen objects in plate 56, figures 8, 9. Figure 6 is a broken wooden shaft, composed of two sections spliced together and held with a wrapping of baleen. Figure 8 is a baleen ring, from cut 9, 48 inches deep, the ends held together with baleen wrapping; small strands of baleen, attached to the rim, may have extended across. 172 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 Fishing line —Figure 9 is a fishing line from cut 26, depth 39 inches, made of thin strips of baleen tied together with small tight knots; it has been carefully looped and tied together at the middle so as to unroll without tangling. Fishhook.—Pilate 56, figure 3, is probably a fishhook;; it consists of a wooden shaft, with a small pebble for a weight, and a sharpened bird bone for a barb, both attached by means of baleen lashings. Figures 8 and g (cut 18, 27 in., and cut 23, 22 in.) are two small objects of unknown use, square in cross-section, made by plaiting together four interlocking strips of baleen. The technique has a wide, if sporadic, distribution. An object identical with those illustrated, but made of bark and used as a child’s toy, is described from the Tver government in the Volga region, U. S. S. R. (Zolotarev, 1926, p- 154, fig. 10, 9). The British Guiana Indians make toys of leaf strands which are exactly similar ; and specimens in the United States National Museum show that similar plaited ornaments were made by the Mohave and by certain Mexican tribes. Figure 10 (cut 18, 40 in.) is a fragmentary object consisting of three small pieces of wood lashed together with baleen; use unknown. Figure 11 (cut 18, 32 in.) is a strip, or strips, of baleen, neatly wound and tied up with a strand of the same material. Figure 12, from the same location, is a flat sliver of wood with a small baleen loop—a slip knot—tied to one end; possibly a snare. Figure 13 (cut 27, 16 in.) is a piece of wood to which a small sharp sliver of bird bone has been lashed by means of a strand of baleen. Figure 16 (cut 21, 24 in.) is a small piece of wood with a wisp of human hair attached. Figure 14 (cut 27, 26 in.) is a small bundle of grass, tapering to a bifurcated end, and wrapped with baleen. Its use is unknown, although it resembles the arrow targets of grass wrapped with sinew used by the northern Plains Indians (Culin, 1907, p. 384, 391). The toy bows of baleen (pl. 55, fig. 1, and pl. 56, fig. 15) have already been described. Oxsyects or Woop Many wooden objects and fragments were found in the permanently frozen soil at the lower levels of the Hillside site and Miyowagh, but their identification presents unusual difficulties. Most of them are broken and incomplete, and others—as is also the case with many artifacts of ivory, bone, and stone—seem to have no counterparts in modern Eskimo culture. Furthermore, a great many of them are of a rather generalized nature and could easily have served more than one purpose. Some of these wooden objects, the exact functions of which are uncertain, are described below. a a a NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 173 Shovels (?).—In plate 54, figures 6-8, are shown three flat wooden objects from house no. 1, Hillside site, which may have been used as spades or shovels for some particular purpose such as cleaning out the bottom of the umiak. Such a function is suggested by the fact that on each of them the tapering lower edge, and the edge only, is battered and worn as if through constant scraping. The photograph shows the flat, presumably upper, surface; the under side is somewhat convex. The edges of plate 54, figure 7, are rather sharp, those of the other two thicker and more rounded. The two pairs of narrow slots or small holes near the upper ends were evidently for the lashing by which the handle was held on. A few strands of the lashing—of baleen—remain in place in one of the slots in figure 6. Plate 47, figure 14, cut 22, depth 41 inches, seems to have been a hook, with an oblique perforation through the enlarged lower end for holding the prong; there is an incrustation of some charred greasy substance around the lower hole. Plate 47, figure 15 (cut 10, 54 in.), represents a highly stylized human face on the enlarged end of a wooden object of unknown use. The eyes and mouth are indicated by depressions made when the nose was carved out in relief. The gougelike lower end of plate 47, figure 10 (cut 27, 46 in.), and plate 37, figure 7 (cut 8, 32 in.), suggests that these objects may have been used as scrapers, possibly for removing the fat from intestines. Plate 47, figure 11 (cut 27, 42 in.), represents a rather common type of implement, probably a scraper handle, with a deep rounded socket for the blade; on the opposite side there is a prominent lip at the rim for holding the lashing in place. Seven other handles of this type were found at Miyowagh, as follows: Cut 7, depth 37 inches, 62 inches ; cut 18, 42 inches, 53 inches; cut 19, 45 inches, 79 inches; cut 24, 54 inches. Engraving tool——The wooden object shown on plate 57, figure 6 (cut 19, 79 in.), is probably the handle of an engraving tool. In the upper end is a small pit or socket about 1.5 cm deep and 2 to 3 mm in diameter. Such a socket would seem to call for a metal point, al- though it might have been of bone or ivory. There is a lashing of baleen around the upper end; the lower end is widened and flanged, producing a shape somewhat suggestive of certain Alaskan needle cases. Plate 57, figure 7, is a long, narrow piece of wood with one end pointed and notched—possibly part of a trap or snare. 174 SMITHSONIAN MISCELLANEOUS COILECTIONS VOL. 96 Plate 57, figures 9 and 10 (cut 23, 81 in.) are two fragmentary wooden rods carefully carved and painted with red pigment, over which designs in black were applied; the latter consisted of rows of very small circles between heavy V-shaped figures. The painting of wooden objects with red pigment was a common practice with the Old Bering Sea Eskimos. Drying rack.—Plate 57, figure 22 (cut 19, 45 in.), is part of the wooden frame of a drying rack. The wooden slats or cross pieces, of which figure 20 is a fragment, were set in deep slots along the edge averaging 2.5 cm long, .£8 cm wide and 1 cm deep. One complete wooden slat for a drying rack was 60 cm long. A number of pieces of such slats were found in other cuts at Miyowagh; there were also several questionable fragments from the Hillside site. Plate 57, figures 18 and 19 (cut 23, 32 in. and 45 in. deep), are two small wooden cylinders 2.7 and 2.2 cm long respectively ; both have a groove around the middle containing a small strand of baleen. These objects were found in considerable numbers at Miyowagh, many of them, like figure 19, being painted red; they may have been floats for small bird snares of looped baleen. Plate 57, figures 11 and 13 (cut 24, 68 in., and cut 18, 36 in.), are pointed and slightly curved wooden objects; use unknown. Figures ‘14-16 (cut 18, 49 in., 59 in.; cut 28, 31 in.) are three more wooden objects of uncertain use; figure 15 is painted red. Plate 57, figure 12 (cut 18, 36 in.) is flat on the bottom and con- stricted at center, has a wooden peg in one end and a small piece of wood set in a transverse groove at the opposite end. Figure 17 (cut 23, 81 in.), apparently a fragment, has a small piece of wood inserted through a central perforation. Figure 21 (cut 18, 40 in.) is carved in a single piece and is possibly one of the end pieces for the frame- work of a toy umiak. Drum rim.—A small section of a drum rim is shown in plate 57, figure 8 (cut 16, 46 in.). It is of wood, 1.9 cm wide and has a con- tinuous groove around the outer side. Pieces of drum rims were also found in cut g; cut 18, depth 44 inches; cut 23, 72 inches; cut 24, 25 inches; cut 25, 8g inches. Drum handle—Only one drum handle was found at an Old Bering Sea level (pl. 55, fig. 5—cut 16, 23 in.). Unlike the later handles, this one is carved from the same piece of wood which forms the rim. Enough of the latter was found to show that the diameter was around 35 cm. In plate 37, figure 8, is shown a wooden object of unknown use, purchased from an Eskimo who dug it up at Miyowagh. It is 28.5 NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 175 cm long, oval in cross-section, and somewhat pointed at one end. A narrow longitudinal tapering section was cut out and the channel enlarged and cut down to a depth of about 1.8 cm. The longitudinal strip was then replaced, leaving a squarish to rectangular canal through the middle, apparently in order that the object might be used as a tube of some kind. The loose strip was held in place by means of two tightly wound loops of baleen. Several small rolls of birch bark were also found, evidently ob- tained from drift wood. Opjects oF Ivory AND BONE Comb.—Plate 58, figure 10, is a small ivory comb, with eight teeth, from cut 7, 67 inches; the under side is flat, the handle part of the upper side is convex, with an irregular freehand circle at the highest point; the upper end was perforated. This is the only example of an Old Bering Sea comb in the entire collection. Shuttle —Plate 58, figure I1, is a small ivory shuttle (cut 7, 67 in.), used probably for making baleen netting. A similar specimen comes from cut 7, depth 75 inches and two fragments from cut 25, depths 61 and 72 inches. The only examples of netting were a few small sections which seemed to be parts of a baleen dip net, such as the modern St. Lawrence Eskimos use for catching tom cod (Nelson, 1899, pl. 70, fig. 12). Four bone and ivory objects of unknown use are shown in plate 58, figures 13, 15, 17, 18. Figure 13 (cut 23, 58 in.), of ivory, is rounded at one end, bifurcated at the other; it has sharp tips and a circular hole through the center. Figure 15 (cut 5, 32 in.) is a small plaque, made of a flat square piece of bone. Figures 17 and 18 (cut 7, 67 in., and cut 8, 20 in.) are two broken ivory objects which are flat on the back side, somewhat rounded on the front, and widened or flanged at the center; there is a perforation through the upper, thinner, end, from which a deep V-shaped groove extends down the front side; paralleling the lower end of the groove and continuing to the end are two rows of small round pits; both objects are broken at the point where a large transverse hole has been drilled through. What seems to have been a similar object, from the fourth level of the Hillside midden, is shown on plate 30, figure 27. Plate 58, figure 16 (cut 16, 36 in.), shows two walrus teeth, fused along the approximal surfaces, a rare condition; a small pit on the Opposite site shows that they had been used as a drill rest. In plate 30, figures 18-20, are shown three heart-shaped ivory objects of unknown use. Figure 18 was found in house no. 2, the two others 176 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 (pl. 30, figs. 19 and 20) between the stones of house no. 1. All three show the same characters, a deep, wide notch above with a slight beveling of the two upper edges, and a small notch below. Plate 30, figure 18, differs from the other two in having an oval-rectangular perforation through the center. The small button-shaped ivory object, plate 30, figure 21, was found among the rocks at the north end of the site. It is 2.3 cm square with rounded corners, a flat base and a rounded top. Plate 30, figures 22 and 23, are small plaques of ivory, both broken; their average thickness is 3 mm. Plate 30, figure 28, is a smaller, thinner piece, with a longitudinal groove (on the opposite side), and two small round holes at the larger end; it is 3 mm thick. Plate 30, figure 29, is a section of a small walrus tusk, notched at the end. In plate 46, figures 9-12, are shown four ivory objects which seem to be parts of what Nelson has described as “ sinew spinners ” from St. Lawrence Island (Nelson, 1899, pp. 111-112, fig. 31) but which in all probability were instead parts of toy “ spindle buzzes’ such as Bogoras has described from the Chukchee and Kerek (1904-09, vol. I, pp. 273-274, fig. 197, d, e) and which are also known from the Ammasalik and Egedesminde districts in Greenland (Thalbitzer, 1914, pp. 654-655, figs. 379, 380; Birket-Smith, 1924, fig. 300). Plate 46, figures 9 and 10 (cut 9g, 48 in., and cut 27, 22 in.), are probably the bases of such toys. Figure g is rectangular in form, with a flat base and a convex upper surface in the center of which is a round pit 12 mm deep. Figure 10 is of the same shape but has the corners rounded and the central hole extending completely through. Two others were found, one oval, from cut 21, 42 inches, and one circular, from cut 7, 80 inches. We found none of the flattened rods or handles ; the small ivory object, figure 11 (cut 24, 64 in.), perforated at one end, is the only specimen in the collection which might have had such a function. Figure 12 is one of several spike-shaped ivory objects corresponding to the slender knobbed rod which was inserted in the ivory base. The latter, it should be noted, might in the round form be regarded as tops. In plate 37, figure 6, is shown an ivory object of unknown use from cut 24, depth 80 inches. It resembles in certain respects an ivory object previously described by the writer (Collins, 1929, pl. 5). It is 11.8 cm long and 5 cm wide and has the under side hollowed. In addition to the two perforations shown there are two smaller transverse openings | near the upper end. This end of the object was carved to represent the blunt head or snout of an animal with a large mouth and with eyes indicated by round deep holes in which were set plugs of wood. The | surface has been ornamented with incised designs—Old Bering Sea— but these are so weathered as to be hardly visible. NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 77 CLOTHING ORNAMENTS, Etc. We know very little as to the kind of clothing worn by the Old Bering Sea Eskimos, though there is no reason to believe fhat it dif- fered essentially from that of the modern Eskimos. The gut parka was used, for several large fragments were found at Miyowagh. The heavier garments were made of seal and bird skins, to judge from fragmentary pieces found in association with the burial back of house no. 3 at Miyowagh. Ornaments were surprisingly rare. In addition to the few previ- ously described, they are represented only by an occasional button, pendant, or brow band. Buttons —Three buttons are shown in plate 58, figures 3-5. Figure 3 (cut 24, depth 64 in.), is an oval button of bone, undecorated; on the under side, which is shown, the center is raised and slotted; the front side is convex, rising to a pointed center. Figure 4 (cut 19, 30 in.), is half of a larger button, of ivory; it is also slotted on the base. Figure 5 (cut 24, 25 in.), is a piece of a smaller ivory button, oval in shape, flat on the under side, with the upper side convex and rising to something of a peak at center; around the edge are three incised lines, the inner one spurred; encircling the raised center is another spurred line, with four equidistant hatchured triangles pointing in- ward; the base is slotted in the same manner as figure 4. Brow bands——Plate 58, figures 6-9 and 12, are probably to be in- terpreted as brow bands; all are of ivory. The smallest, figure 9 (cut e752 i.) is 5.3 cm long; the largest, figure 7 (cut 7, 75 in.) 9.5 cm. Figure 12 is the only one that is decorated ; it comes from cut 3, depth 60 inches. The design consists of a continuous wavy line made by opposed spurs thickened into triangles; beneath the base line on which the lower row of triangles rests, is another line which curves upward at the ends and forms two crude, irregular circles. The ob- ject appears to have been broken, probably at about the center, and the remaining half used again ; there are two circular holes, remarkable for their smallness, at either end, the smallest of them being about half a millimeter in diameter. Plate 58, figures 6 and 8, come from cut 18, depth 81 inches, and cut 23, 72 inches. Eleven additional brow bands or fragments were found, as follows: Cut 1, 12 inches, 18 inches ; cut 2, 12 inches; cut 4, 8 inches; cut 17, 20 inches (2); cut 20, 12 inches; cut 23, 72 inches; cut 26, 55 inches; cut 27, 38 inches, 42 inches (pl. 15, fig. 11). A single example was found at the Hillside site, the broken specimen shown on plate 30, figure 26, found in the third level of the midden. 178 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 Snow goggles —Two wooden snow goggles are shown on plate 58, figures I, 2 (cut 19, 54 in., and cut 24, 39 in.). Both are well made and have the nose clearly indicated. In figure 1, the eyeslits are dis- tinctly oblique; in figure 2, the eyebrows are carved in low relief. Unfinished snow goggles of wood were found in cut 18, depth 32 inches, and cut 19, 59 inches. Toys In addition to the toy boats, sledge, and bows already described, several other toys were found at Miyowagh (pl. 59). Plate 59, figure 12 (cut 18, 12 in.), is a small ivory disk, perforated at about the center and presumably used as atop. There are two others, one made of baleen, from cut 9, 36 inches, and one from the epiphysis of a walrus vertebra, from cut 7, 33 inches. Plate 59, figures 13 (exact location unknown) and 14 (cut 7, 75 in.), are small bark figures that seem to represent whales. Four others were found, in cut 5, depth 12 inches ; cut 18, 44 inches; cut 23, 68 inches; cut 24, 64 inches. Plate 59, figures 16 and 17, are dolls; the first, from cut 19, 34 inches, of wood, has legs but no arms; the second, of bark, from cut 9, 48 inches, lacks both arms and legs. Figures 18-21 are toy harpoon heads, all with closed sockets, found in cut 24, depth 54 inches, cut 18, 59 inches, cut 19, 56 inches, and cut 7, 75 inches. Figures 22-24 (cut 18, 60 in., cut 27, 46 in., cut 5, 23 in.) are three small ivory replicas of the elaborate winged objects that are so characteristic of the Old Bering Sea culture; figures 22 and 24 are well made and have all of the distinguishing features—a rectangular basal socket, graceful, rounded wings, and a small indentation at the end of the projecting central part ; the ornamentation consists of only a few simple lines; figure 22 has two small cylindrical plugs of baleen through the wings. Figure 23, showing the opposite side, is unfinished. Figure 25 (cut 24, 54 in.) represents a small ivory polar bear, the eyes indicated by shallow pits and the teeth by short straight lines. Figure 15 (cut 10, 54 in.) may not be a toy, although it probably had no serious function; this exam- ple of a continuous spiral groove is of interest as showing a knowledge of the simple principle on which the screw is based, even though it was never utilized. Plate 50, figure 5 (cut 7, 37 in.), is probably an ajagagq of the sim- plest form—a seal humerus; the roughened center would indicate the attachment of the line to which the stick or peg’ was tied and the foramen at the upper end the opening in which the end of the stick was caught. NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 179 Some of the baleen objects illustrated in plate 56 were in all proba- bility also toys. Figures 4, 5, and 6 (from cut 19, 45 in., 38 in., and cut 27, 26 in.), would seem to come in this category, representing probably a snow knife and paddles; figure 7 (cut 19, 42 in), is more doubtful; its rather heavy sharpened end might have been intended for use. IMPLEMENT TYPES, EARLY PUNUEK In plate 60 are shown a number of objects representing types which, at Miyowagh, were found either in the later, northwestern section of the midden or in the upper layers of the older, southeastern section, but neither at the Hillside site nor from the lower levels of Miyowagh. Such types may therefore be regarded as belonging to the early Punuk stage. There are undoubtedly other types characteristic of this transi- tional period, but an attempt to segregate them, when they occur in the same midden with Old Bering Sea and later Punuk material, would result in considerable uncertainty. Full knowledge of the type forms of the early Punuk stage must await the discovery of a pure site of this period; for the present our information regarding it must be limited to the evidence afforded by harpoon heads and art styles, and the few objects such as shown in plate 60. Bone dish.—Plate 60, figure 1, is half of a bone dish from cut 17, depth 34 inches, made from the epiphysis of a whale vertebra; the entire inner surface has been carefully smoothed and slightly hollowed out; the outer surface is worked only along the periphery. Adz head.—F igure 2 is an ivory adz head from cut 19, 42 inches deep. The two lateral “ ears’ for the lashing thong are similar to those on the Old Bering Sea specimen shown in plate 46, figure 3. On the under side there is a shallow oval pit for the end of the handle, and just above it two holes for the lashing. The blade rested in the concave lower end of the front side. The decoration is simple, consisting of a crescent-shaped and circular figure, both divided into sections by means of short cross lines; circles at the center of the “ ears’; and a few slightly curving lines on both sides at the upper end. Plate 60, figure 3 (cut 9, 36 in.), is an ivory object of unknown use; one end—the upper as here shown—is rather sharp edged and spatulate in shape, the opposite end is wider with a longitudinal groove 3 mm deep and 3 mm wide in the base. Figures 4 and 5 (cut 19, 17 in., and cut 23, 24 in.) are two ivory objects which may have been cord handles. Figure 4 appears to be unfinished. 180 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 Snow goggles.—Figure 6 is one half of a pair of ivory snow goggles picked up on the surface of the midden. The eye slit is small, only 2.5 cm long and .5 cm wide. Decoration consists of a few curving lines with spurs attached. Plug.—Figure 7 (cut 2, 28 in.) is an ivory plug, probably an orna- ment of some kind. It has a rough conical tang and a flat smooth upper face on which there are five small round pits, drilled to a depth of between 2 and 3 mm. Toy bow.—Figure 8 (cut 19, 42 in.) is a toy bow of wood, heavily reinforced with baleen. A thick strand of baleen, loosely twisted into a cable, extends along the back, being held on by cross lashings at regular intervals. The under side is reinforced by a strip of wood bound on with baleen. In this toy we have the earliest evidence of the use of the sinew-backed bow on St. Lawrence Island—for in all likelihood the baleen on the toy is a substitute for the sinew backing of the real bow. Drill shaft-—F igure 9 is a drill shaft of unusual type, the upper part of wood, the lower part of ivory. It was purchased from an Eskimo who had excavated it at Miyowagh, from a depth of between 1 and 2 feet. The upper end of the wooden section is smoothed and worn where it revolved in the drill rest; the lower end is set into the hollowed out upper end of the ivory section. The latter is decor- ated in Punuk style, with four encircling spurred lines and inter- vening circular pits, some drilled to a depth of 3 mm. At the lower end is a shallow concavity for the point which was probably of metal; opposite this is a lipped edge to prevent the lashing from slipping. There is also a small round pit in the end 1.5 mm deep, which would seem too shallow to have held a point, even of metal. Another drill shaft of bone, with particles of iron remaining in the small socket at the lower end, was found in cut 20, at a depth of 22 inches. Engraving tools —The two small ivory objects, plate 60, figures 10 and 11, are engraving tools from cut 27, 22 inches, and cut 3, 20 inches. Another comes from cut 23, depth 52 inches. They are both delicately made; with the maximum thickness—around 6 mm—at the center, and both ends constricted. Figure 10 is 6.2 cm long. The upper end of figure II is broken; the center is pierced by four slots. These imple- ments are of particular interest because they were provided with iron points, a few corroded particles of which still remain on figure IT. These, presumably, were the implements used by the early Punuk Eskimos in engraving designs on ivory—designs which, as we have already seen, show clear evidence of having been produced by metal tools. NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 181 Shovel_—F igure 12 is a small shovel, made from a walrus scapula, from cut 4, depth 15 inches. It is 13.2 cm long and 10.4 cm wide, much smaller than the more common type made from an entire scapula, illustrated in plate 50, figure 6; it is likewise more carefully made than the larger forms. At the upper end are two small drilled holes for the lashing which held the handle on. Fragments of two similar shovels were found in cut 5, 23 inches deep, and cut 19, 32 inches. IEVOGHIYO0Q Some 200 yards to the north of Miyowagh is the smaller site, levoghiyoq, a grass-covered midden rising like a small island from the flat gravel plain (pl. 61, fig. 1 and text figs. 2 and 18). It is separated from Miyowagh by four old beach lines, and to the northward six more beach lines lie between it and the sea. Its position would thus indicate that it had been built later than Miyowagh and that during the time of its occupancy and after its abandonment the sea had re- ceded still farther, piling up still more gravel ridges. The Ievoghiyog midden is roughly oval in outline, with an average diameter of around 150 feet. The shallow depressions of about a dozen house pits are visible on the surface; two of these were excavated—houses no. 6 and 7. In recent years, probably after the famine and epidemic of 1878-79, burials were made on the surface of the midden, and there is an old cemetery on the gravel ridge extending eastward toward the hillside. The burials on the ridge had been enclosed by oval rows of stones, similar to those observed by Nelson on the south side of the Island (Nelson, 1899, p. 321). Although the Ievoghiyoq midden stands out more prominently than Miyowagh, its total depth was only 5 feet as compared with the 8-foot depth of the older midden. Its apparently greater height is due to the fact that not so much gravel had piled up along the edges. Being situ- ated on the gravel plain, the Ievoghiyoq midden is rather dry, and the material found in it was in a much better state of preservation than that from the southeastern section of Miyowagh. The ivory was likewise more uniform in color, ranging from a light cream to light brown, in contrast to much of that from Miyowagh, which had been turned to a dark brown or black from the action of water. In 1930 seven cuts (one incompleted) were excavated at Ievoghiyoq, and two others in 1931. These will not be described individually as the material obtained was remarkably uniform from top to bottom in contrast to that from Miyowagh, where there was evidence of cultural modifications along many lines. Ievoghiyoq proved to be a pure Punuk site, one which had been established and abandoned during the period 13 182 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 i / ) y / 7 / N J i a Ba MZ yy @) 25 50 15 100 FEET Fic. 18—Outline of Ievoghiyoq, showing excavations of 1930 and 1931. NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 183 when Punuk culture was flourishing on St. Lawrence Island. But as we shall see later, the material obtained from this site was not identical with the Punuk material from Miyowagh. It represents a more de- veloped, more typical aspect of Punuk culture, showing that levoghiyoq was established just before the time Miyowagh was abandoned, and that it continued to be occupied for many years afterward. Around 2,000 artifacts were excavated at Ievoghiyoq, in addition to the usual quantities of potsherds, animal bones, pieces of slate blades and other fragmentary objects. The Ievoghiyoq midden was structurally much like Miyowagh ex- cept that in some places, particularly cut 4, there was more gravel. There was also a larger content of mussel shells; these were par- ticularly abundant in the upper levels of cuts 5 and 6, and they formed a distinct stratum about 6 inches thick in cut 1, sloping from a depth of 1 to 3 feet. Cuts 1, 2, 5, and 6 extended through thick, solid refuse which thawed very slowly. These cuts were prolific in artifacts and yielded large quantities of mammal, bird, and fish bones as well as baleen. The above mentioned cuts in the midden revealed no remains of house structures and parts of only two caches. Both of the caches were found in cut 4, which was mostly gravel. One of them, at the southeast corner, had been built of whale skulls, two of which were visible, resting on logs and whale vertebrae. The other cache was only 2 feet away; it had been constructed of stones, one above the other, the upper tiers evidently having been removed, as no fallen stones were found nearby. HOUSES AT IEVOGHIYOO Two houses were excavated at Ievoghiyoq. One of them, house no. 6, closely resembled house no. 5 at Miyowagh, but was much better preserved. The other, house no. 7, was excavated by Mr. Ford dur- ing my absence from the island. It was larger, apparently older, and differed in important structural features from house no. 6. ifouse no. 6 (text fig. 19, and pl. 62, figs. 3, 4)—This house was Ig feet 4 inches long by 19 feet wide; it had stone walls and floor ; the entrance passage, facing east, was 8 feet 8 inches long and 25 inches wide. The floor of the passage was at the same level as that of the house, a little less than 2 feet below the surface ; the roof, walls, and floor of the passage were of stone slabs. The end of the passage was enlarged into a rounded antechamber (pl. 62, fig. 4) 74 feet long by 64 feet wide. It had been roofed with whale ribs, had a stone floor, and walls built of stones, walrus skulls, whale vertebrae, and five small VOL. 96 SMITHSONIAN MISCELLANEOUS COLLECTIONS 184 je Si eS ee J SOSA? Gs ' \) : “NB Q \Q) ae a US: 0 WO ONS CY &o ONS, | MG L | yer y fomon= 20 no. 6, levoghtyoq. Fic. 19.—Plan of house ~ NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 185 whale skulls. The antechamber evidently had been used for cooking, for on the north side was found a mass of burnt gravel and bones. No artifacts were found in the house or passage. House no. 7 (text fig. 20, and pl. 61, figs. 2-4; pl. 62, figs. 1, 2). — This house faced toward the south, was 20 feet 10 inches long and 194 feet wide, and at the back had a narrow recess or annex; the floor was 4 feet g inches below the surface. The house was entered by a passage 23 feet long by 21 inches wide, with an enlarged end 4 feet 10 inches wide. The narrow part of the passage seemed to have been roofed with stones, but the enlarged outer end had been roofed with timbers. The passage had a stone floor and sides, but unlike those e. z a : UP 20 a Fee ee eee COIS NSS Ss r=Ip\- ENESCO Re bSOVS icons LONGITUDINAL CROSS SECTION A TOE » 8 le Fic. 20.—Plan and section of house no. 7, levoghiyoq. previously described, it had three different levels, the third or outer one being higher than the floor of the house. The house floor was also of stone slabs, but the walls and roof had been constructed entirely of timbers. These were fairly well preserved, and here for the first time a clear view could be had of the unique type of wall structure that had been so imperfectly preserved at Miyowagh. The timbers were laid one over another horizontally, passing around a large up- right timber or whale jaw (roof support) at each corner, and several smaller uprights in between. The wall appeared to have had a height of about 4 feet. The roof timbers had collapsed and covered the entire floor, but unfortunately the roof structure could not be made out, for the excavation of this house was not begun until the middle of August, 186 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 and by the time the roof timbers were completely exposed the ground had begun to freeze again, so that the excavation had to be left unfinished. Artifacts from house no. 7 include a number of potsherds; slate blades ; adz blade ; adz handle; ivory fishline sinker ; drill rest ; wooden drum handle (pl. 81, fig. 6) ; kayak paddle (pl. 84, fig. 2) ; board used in making fire (by drilling) (pl. 84, fig. 1); and 35 small wooden pegs, for bird snares. Two poorly preserved human skeletons, with crushed skulls, were found beneath the fallen roof timbers. SEKLOWAGHYAGET Seklowaghyaget, the largest of the old Gambell sites, is a low, spreading midden situated at the northwest end of the lake and im- mediately behind—to the east of—the present village. It has an aver- age breadth, north and south, of 200 feet and a length of some 200 yards but its exact boundary on the west cannot be determined, since it merges imperceptibly into the older section of the present village. Numerous house pits are visible and the surface has been much dis- turbed by the digging of the Eskimos. On the south side the midden stands out rather prominently as it slopes down to the low land bor- dering the lake, but one of our cuts, no. 4, sunk just back of the highest point, showed that the refuse extended to a depth of only 4 feet 4 inches. The appearance of greater depth was due to the fact that on the south side the refuse had been distributed along the top of the gravel ridge bordering the lake, some of it falling down to the low ground below. Seven 12-foot cuts were made at Seklowaghyaget, four in 1930 and three in 1931. The cuts ranged in depth from 4 feet 1 inch to 5 feet 7 inches. There was more gravel than at the three older sites and not nearly as much cultural material—less than 700 specimens were found. However, these were sufficient to indicate that the site had been established during the Punuk stage and occupied until a comparatively late period, when the Punuk culture had begun to be supplanted by the modern. The midden differed from those previ- ously described in that much of the gravel and refuse was saturated with blubber, a condition similar to that observed at Point Hope on the Arctic coast. This is an indication not only of the younger age of the Seklowaghyaget midden as compared with the three already described, but also of the fact that whaling had now become an 1m- portant occupation (as is also shown by the relatively large number of whaling harpoon heads) and that larger stores of whale blubber NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 187 had to be preserved. It was no doubt this whale blubber, stored in underground caches all over the midden, that permeated the gravel. Remains of stone and whale bone caches were found in three of the cuts. Plate 63, figure 1, shows a mass of fallen whale bones and stones which seemed to be part of such a cache, lying above a piece of blubber soaked walrus hide. HOUSE AT SEKLOWAGHYAGET House no. S—This house, the best preserved of its type that we excavated, was not situated on the Seklowaghyaget midden but on g79 eo feg PEO OUI yeu as oOU VERORCIDN UPax 5 CSNICS aps () IDO OA ESCM kf Rae 5, Age, = SIGS vevyeuvuy yur” O Ra cece Pare UY QA MEAY “(eS WHPSS!sq VIT Ay o G8 BAER AKRINICE OW Ke A ay, “ey De nh eoemanen-en Amgen SUG NI MAING CN SUA De | Se EWEN VE 3S (oo 35200 Afr 1 ae Si a Qe SAC CSE HO SIC IIR OFS % SPOR NO MARES en Pana mae SRP AAC > a ‘ LONGITUDINAL CROSS SECTION A TOE ‘E o a ; c y ] yyy UY wy, HY MMM Y YY) Uy Up Yf WW Wf Y fj Fic. 21.—Plan and section of house no. 8, Seklowaghyaget. fj YW, WM the gravel flat a short distance to the eastward. It consisted of a rect- angular inner room 18 feet long by 15 feet wide (text fig. 21 and pl. 63, figs. 2, 3) and an entrance passage 30 feet long facing to the south. The walls of the inner room, mostly fallen, had been built of walrus skulls, whale vertebrae, and stones. The floor, which is 24 to 445 feet below the surface, was covered with flat stone slabs and was somewhat sunken at the center. At the entrance, on the south side, a small whale skull was driven into the gravel, the flat occipital portion serving as a threshold over which the occupants crawled as they entered or left the house. To either side of the whale skull was a long upright stone slab. On the east side of the entrance a number of stones and whale vertebrae were piled up around a projecting whale jaw, the only one of the roof supports still standing, and just beyond, 188 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 in the corner, was a circular space about 2 feet in diameter, probably for storage. The entrance passage was 29 inches wide and the floor level 16 inches below that of the inner room. The floor, like that of the inner room, was of flat stone slabs; the walls were of heavier stones piled one above the other, and across the tops of these were laid the whale ribs and jaws forming the roof. Plate 63, figure 5, shows the entrance passage before it had been entirely uncovered. The man at the center is standing in the narrow part of the passage, between the whale ribs which formed the roof. In the foreground are seen the fallen whale ribs of the roof of an anteroom—an enlargement of the passage—13 feet beyond the entrance to the inner room. This anteroom was an oval chamber 8 feet 8 inches long and 7 feet 8 inches wide. Plate 63, figure 4, shows the west side of it after the fallen roof bones and gravel had been cleared out. Its floor was of stone slabs sloping from the sides toward the center, and the walls of stones with an occasional walrus humerus or whale vertebra included. The height of the walls here was slightly over 3 feet. The roof had been constructed of whale ribs and jaws. In this enlargement of the passage, which had no doubt served originally as a store room, were found six human burials. These had been made after the house had been abandoned, for they were resting on about 2 feet of gravel which had accumulated above the floor. The roof, however, was at least partly intact when the bodies were placed there for the whale bones had fallen in on top of them. At the southern end the anteroom contracts and then continues again as a narrow passage 25 inches wide. The walls and floor were built in the same manner as those just described, but the floor level was 4 inches above that of the anteroom. Across the outer end of the passage was laid a long narrow slab of stone and resting partly on this and partly on a whale rib laid across the wall was a whale skull. Three feet beyond was another whale skull which had fallen and partly blocked the passage. The entrance to the passage was apparently be- tween these two skulls. Two human skulls and a few long bones were found in the outer entrance at the same level as those in the anteroom. The tops of the walls of the passage and anteroom are at present about 1 foot below the surface. The cultural material found in house no. 8 indicates that it was con- temporaneous with the latest period at Seklowaghyaget, probably early eighteenth century. The fragments of lamps and cooking pots were all of the recent types, and one half of a blue glass bead was found. The bulk of the material, however, is to be classed as late Punuk as is NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 189 shown by such objects as the harpoon heads (pl. 71, figs. 15-20), adz socket (pl. 78, fig. 21), plates of bone armor, bone knife, arrow point, harpoon socket piece, etc. Oxp SECTION OF GAMBELL The latest of the old sites in the vicinity of Gambell is that at the north end of the long narrow gravel bar which lies between the lake and the sea. Here, immediately to the south of the present village, are found the ruins of the underground houses of the nineteenth cen- tury, the last of which were abandoned between 40 and 50 years ago. The site is nothing more than a continuation of Seklowaghyaget, al- though the Eskimos use that name only in referring to the larger midden extending along the north end of the lake. The house pits in this latest section are deeper and more sharply outlined than those at the older sites, and the midden deposits around them are much shal- lower, perhaps not exceeding 3 feet in depth. Two cuts were made, one of them through refuse 20 inches thick, the other 32 inches. The midden here was no less prolific in artifacts than the older ones (about 250 artifacts being found), and the material obtained was of value as illustrating the latest stage in the transition from the Punuk to the modern culture. The evidence furnished by harpoon heads was of particular interest in this connection, since it showed that the modern thick, rounded form with closed socket had evolved directly, and rather suddenly, from the immediately preceding form which was thinner in cross-section, with a wedge-shaped socket, which, though still open, had functioned as a closed socket. HOUSES AT OLD SECTION OF GAMBELL House no. 9—On the low ground between the midden and the lake and about 100 feet SW. of Seklowaghyaget is a house ruin very simi- lar to house no. 8. The inner room is roughly square in outline with a stone floor 3 feet below the surface (text fig. 22). The walls, which are now mostly fallen, had been built of whale vertebrae, walrus skulls, and stones, with a single upright whale jaw in the rear (west) wall. At the NE. corner there was an upright wooden roof support, and another at the NW. corner together with a small upright whale jaw and a whale skull. The stone flooring does not extend to the wall on the north, west, and south sides, and it may be that sleeping platforms had occupied these spaces. Just as in house no. 8, an inverted whale skull had been placed at the entrance. The narrow passage way was built of stones and roofed with whale ribs; the enlarged anteroom was I9QO SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. g6 constructed in a similar fashion, with a stone floor, walls of long up- right stones, above which were placed smaller stones, and above these a roof of whale ribs. A unique feature of this house was an oval annex, evidently a cache, opening from the anteroom on the south side (pl. 64, fig. 1). The floor of this annex was of dirt, and it was 3 feet above the floor of the passage. It had stone walls and a roof of whale ribs. Two whale skulls were placed at the point where the anteroom narrows at the outer end, and another at the extreme outer end of the passage. = N oe Joc g SOG 5 aos Y Nels Ns C c IC Ae 2 g¥0 Nec ¢ sat IBOVOREGI sibueg Wate OVLYN eke E AIST SOR EO SES oD cy ve C VW NSO 2 wa, we <} Q ON 5 eas : Yeon mAnAAAR TY LONG/TUOINAL CROSS SECTION A TOE Ya) E Al 8B 7]; yl UY YU si TTT? 7 YE Y fy Yf YY YY Uff Yf Fic. 22.—Plan and section of house no. 9, old section of Gambell. The cultural material from house no. g was scanty but sufficient to date it as contemporaneous with house no. 8. Fragmentary lamps, lamp rests, and cooking pots were of the modern types and identical with those found in house no. 8; a bone adz socket of the type il- lustrated in plate 78, figure 19, from Seklowaghyaget, was also found ; there were only two harpoon heads, one of them, weathered and pos- sibly a relic, was of open socket type III (a) x; the other was of the later type III (b) x with wedge-shaped socket. House no. 10.—This was one of the latest of the underground houses at Gambell, having been abandoned, according to native testi- mony, around 50 years ago. It was situated on the midden and appeared NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS IOI as a rather deep depression ; excavation showed about 2 feet of gravel over the center and about 4 feet at the corners. The house was not completely excavated, since it belonged to the same type as several much better preserved houses previously excavated at the eastern end of St. Lawrence Island. This type is illustrated in plate 2, figure 3, and plate 3, figure 4. House no. 10 differed from all of the older Gambell houses in having a wooden floor and sleeping platforms, fea- tures also present in the Punuk Island houses. Our excavation un- covered only the floor (pl. 64, fig. 2), parts of the sleeping platforms on three sides, and the entrance, which faced to the east. The floor had been built of split logs neatly fitted together, with flat sides up; it measured 7 feet 6 inches north and south by 7 feet 3 inches east and west. The platforms, which extended along three sides, had been built of similar timbers laid horizontally across timbers extend- ing out from the wall along the floor. The rear wall, which was only partially uncovered, had been constructed of upright timbers, whale jaws, and a few stones. The passage had floor and walls of stones like those of houses no. 8 and g, but no anteroom. The material found in house no. 10 was modern in every respect, including a large quantity of metal—hoop iron, files, large knife blades, and sheets of copper. There were also numerous fragments of modern type lamps, lamp rests, and cooking pots, a stone sinker for a tom cod net, an ivory bird figure, blunt arrow tip of ivory, whetstones, and long, flat sledge shoes made of whale bone. At Mirrukta, a deserted village about 3 miles east of Gambell, ruins of stone-walled houses somewhat different from those described above were found. These may be somewhat more recent than most of the stone-walled structures at Gambell, as they are in a better state of preservation. They are about 18 feet square, and the walls, built en- tirely of stones, are mostly above ground (pl. 64, fig. 5). We did not excavate any of them, so the depth of the floor beneath the surface is uncertain, but it was probably no more than a foot. A stone wall extending the length of the room formed an inner partition about 6 feet wide, the purpose of which is uncertain unless it served as a storage space. Some of the houses had attached at one corner, on the outside, a roughly circular stone-walled cache. The roof con- struction could not always be determined, but in some cases two large parallel whale jaws had formed the main framework of the roof, which had no doubt been formed of walrus skins. These houses lacked the long sunken entrance passages such as were attached to the other early forms of St. Lawrence dwellings; the only entrance now visible was a low opening in the east wall of one of the houses across which IQ@2 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. g6 was laid a large flat stone (pl. 64, fig. 5). Houses of this type ap- parently represent summer dwellings, used along with the semi- subterranean house with walls and roof of wood and whale bones. DECORATED OBJECTS, PUNUK, FROM JEVOGHIY0Q, SEKLOWAGHYAGET, AND OLD SECTION OF GAMBELL Decorated objects from the three later sites are illustrated on plates 65 to 6g. At none of these sites was Old Bering Sea art found, and there is only one doubtful example of Punuk style 1. The objects shown in plates 65 and 66 are from Ievoghiyogq, all of them examples of Punuk style 2. Plate 65, figure 1, is a broken needle case from cut 5, depth 43 inches. The flaring end distinguishes it from the Old Bering Sea forms shown on plate 17. Ornamentation is in phase 2 of Punuk style 2, consisting of deeply incised lines, spurs, dots, and compass-made circles. The other needle case, plate 65, figure 2 (cut 9, depth 12 in.), is similarly decorated; it also has a flaring end, and somewhat below the center two protuberances or “wings”? upon each of which is inscribed a nucleated circle. These lateral “ wings ” are set off by curving lines which extend from the enlarged and flattened upper end to the constricted, tubular lower end. This “ winged ” needle case is similar both in form and ornamenta- tion to two others from Point Barrow described by Mathiassen (1929, vol. 2, p. 95). Plate 65, figure 3, is the handle of an ivory knife from cut 1, depth 28 inches, a common implement type of the Punuk period (other ex- amples illustrated on pl. 79, figs. 4-6). It is decorated with pairs of longitudinal lines, nucleated circles and small spurs. The latter, which are attached obliquely to both the circles and lines, are deeply gouged pits, very different from the straight, slender spurs of the preceding stage, Punuk 1. These short, deeply cut spurs, usually attached in pairs to lines and circles, are distinguishing characteristics of phase 3 of Punuk style 2. The two fragmentary objects (dart socket pieces ?) shown in plate 65, figures 4 and 5, are also examples of phase 3; these come from cut 1, 31 inches, and cut 5, 24 inches deep, respectively. The same arrangement of lines and spurs may be seen in plate 65, figure 6 (cut 8, 8 in.). This is a link ornament, which was probably attached to a drum handle; the end has been carved to represent a seal’s head. This is one of the examples where the pairs of oblique, deeply gouged spurs of phase 3 resemble very closely the modern Eskimo Y-figure. Plate 65, figure 7 (cut I, 42 in.), is an oval ivory block of unknown use, with a piece of wood in the central perforation. It bears a decora- NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 193 tion (phase 2) of nucleated circles at the ends of short lines extending from the opposite borders, with a bold zigzag line formed of large connected Y figures, running down the center. Plate 65, figure 8, is an ivory object of unknown use from cut 5, depth 31 inches. The pairs of deeply cut lines at the center are ar- ranged chevronlike; the lower one takes the form of a bold Y. This is a favorite device of Punuk art, variations of which are illustrated in plate 65, figures 2, 9; plate 67, figure 8; plate 68, figure 7; plate 70, figures 22, 23; plate 74, figure 21. It was also employed during the Old Bering Sea period (e. g. pl. 17, fig. 3; pl. 19, fig. 2). In plate 65, figures 9-11, are shown three wrist guards, implements which are characteristic of the Punuk, but which seem not to have been known to the Old Bering Sea culture. Figures 9 and 10 are made of ivory, figure 11 of bone. They all came from cut 1, at depths of 36 inches, 44 inches, and 36 inches respectively. The incised design on figure 9 consists of pairs of curved V-shaped figures; arcs; short, straight lines and spurs. The designs on figures 10 and II are very much alike and are recorded in table 1 under the heading “ Modern.” The upper rounded end, like figure 9, is divided off by a pair of trans- verse lines, and a V-shaped figure is placed in the space so formed. Below, a narrow strip down the center is left plain, being outlined by two pairs of parallel, longitudinal lines. Pairs of short transverse lines are placed between these and the bordering lines; on figure 10 the transverse lines are placed obliquely, on figure 11 they are at right angles. Short spurs are attached to some of the lines in various arrangements, some as single rows on the inside, some on the outside, and some as opposite or alternate rows; between the oblique lines of figure 10 they take the form of short connecting lines. The four ivory objects from Ievoghiyoq shown in plate 66, figures 1-3, 6, are examples of the more linear style of Punuk ornamentation which in some cases cannot be distinguished from the work of the modern Eskimos of southwest Alaska and which is here designated as phase 4 of style 2. Since most of the pieces decorated in this linear style were purchased from the Eskimos and are therefore lacking in depth records, they are not included in table 1. Plate 66, figure 2, is a wrist guard of rather unusual form. The decoration is in three sections, each consisting of a transverse band of closely spaced, deeply incised parallel lines, at the center of which is a series of short vertical lines arranged in pairs. Plate 66, figure I, is a fragment of another wrist guard. It has the same arrangement of transverse bands of parallel lines, with the ad- dition of occasional long, sharp spurs. 1Q4 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 Plate 66, figure 3, is a broken socket piece for a harpoon. The dec- oration is very simple, consisting principally of pairs of encircling lines connected by shorter pairs of lines forming H-figures; at inter- vals there are deep, circular pits containing remnants of wooden plugs. Plate 66, figure 6, is an object of unknown use, which is ornamented with parallel longitudinal lines and pairs of short transverse lines, with four bold spurs in the center. It comes from cut 8, depth 44 inches. Plate 66, figure 7, is a broken needle case similar to several previ- ously described from St. Lawrence Island (Collins, 1929, pl. 17, b-d). It has a “modern” decoration of encircling lines to which are at- tached alternating spurs. The function of the object shown on plate 66, figure 8, is uncertain. The central notch and the two holes above show that it was lashed to something, possibly as a cleat on an umiak for holding a line. The surface is profusely ornamented with straight longitudinal lines, single and double; and numerous short, oblique, deeply cut spurs attached in pairs to the longer lines. It is a typical exe of phase 3 of Punuk style 2 Plate 66, Apnities 4 and 5, are two other objects of unknown use; they come respectively from cut 1, depth 42 inches, and cut 5, 14 inches. On figure 4 the decoration consists of nothing more than three rows of short detached vertical lines arranged in pairs. Figure 5 shows the combination of lines and short, deeply incised, oblique spurs (phase 3 of Punuk style 2) that was observed in plate 65, figures 3-5, and plate 59, figure 8. Among the 144 harpoon heads and fragments from Ievoghiyoq, only 29 are decorated. This is in striking contrast to the conditions observed at Miyowagh and the Hillside site, where out of a total of 255 heads and fragments, 162 were decorated. Eight of the decorated harpoon heads from TIevoghiyoq are illus- trated in plate 70. There was no clear example of Punuk style 1, which had been so prominent at Miyowagh (pl. 24, figs. 19-21, 23; pl. 26, figs. 6-9; pl. 28, figs. 8-17, 19-22, 25-27). Plate 70, figure 21 (cut 4, 42 in.), is the only possible exception; stylistically it is quite conformable, but the lines and pits are much deeper than on the Miyowagh heads which were decorated in Punuk style 1. Plate 70, figure 8 (from cut 4, depth 59 in.), is one of six harpoon heads from Ievoghiyoq decorated in phase 1 of Punuk style 2; the lines are rather deeply incised, mostly straight, and have very long spurs attached. This particular arrangement of lines and spurs (which was also found on several harpoon heads from Miyowagh—pl. 28, NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 195 figs. 13, 14) appears to be the oldest style of ornamentation at levoghi- yoq, for the six heads so decorated conform structurally in every par- ticular to the Miyowagh specimens mentioned, and lack the specialized characters possessed by the other Ievoghiyog heads; furthermore they all come from the lower levels of the midden, from depths of 31 to 59 inches. Phase 3 of Punuk style 2 was the more common form of decoration applied to harpoon heads at Ievoghiyoq (pl. 70, figs. 15, 18, 20, 22, 23). The lines are very deeply incised, the nucleated circles (when present) are mechanically perfect, and the spurs have taken the form of small, deeply gouged pits which are attached obliquely to the lines. Very few decorated objects were found at the two latest sites, Seklowaghyaget and the old section of Gambell, partly, no doubt, because of the relatively small amount of digging done there. How- ever, the scarcity of decorated objects is also an indication of the fact that by this time the artistic impulse was dying out, and the imple- ment forms were assuming the heavy, utilitarian aspect so character- istic of the culture of St. Lawrence Island in the nineteenth century. Plate 67, figure 7, is an ivory wrist guard from cut 3, depth 20 inches, Seklowaghyaget. It is identical in form with another wrist guard from the same site illustrated in plate 74, figure 22. The decora- tion, which is rather crude, consists of short, vertical lines attached to pairs of horizontal lines, all very deeply incised; it is to be classed as phase 4 of Punuk style 2. The ivory wrist guard shown in plate 67, figure 5, comes from cut I, depth 8 inches, old section of Gambell. It is straight at both ends and has two round holes on one side and two slots on the other for the lashing. The decoration is modern; four curved lines, or arcs, the inner three with light spurs attached, rise from the rim and enclose both of the slots; at the lower end is a band of seven lines with an alternate arrangement of spurs, and at the upper end a similar band of nine lines. | The other six objects in plate 67 were purchased from Eskimos and are reported to have been excavated at Seklowaghyaget. Plate 67, figure 3, is a broken dart socket piece bearing a charac- teristic phase 2, Punuk style 2 decoration of lines, circles, dots, and spurs. Plate 67, figure 4, is a bodkin with a simple decoration of lines and spurs, and at the center two V-shaped figures. Plate 67, figure 1, is an ivory ulu handle, bearing a phase 3, style 2 decoration consisting of deeply cut straight and curving lines, spurs, and compass-made circles. The bordering bands are of particular 196 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 interest in that they show so clearly the relationship of the spurred line to the zigzag, a negative design resulting from the thickening of the bases of the spurs. The lower band contains a perfect zigzag or wavy line and so does the left half of the upper band. On the right half of this band, however, the alternate spurs are longer and more slender and present the usual appearance of the “ toothed line.” For some reason the artist chose to leave the spurs at this stage instead of widening the bases so as to produce the negative zigzag design. This is the only Punuk example I know of in which the very common alternate spur design has been transformed into a zigzag, although the process is constantly recurring in modern Eskimo art, as well as in Melanesian, European Iron Age, and other styles that employ the spurred line motive. We might also note here that although the single spurred line and the ladder design are characteristic of Old Bering Sea art, the simple alternate spur design has not appeared. This, how- ever, may be due to accident, for on the fragment of an Old Bering Sea brow band shown in plate 58, figure 12, there appears a fully developed zigzag pattern. Plate 67, figure 2, is an ivory object similar to several found at Miyowagh (pl. 46, figs. 9, 19), and which apparently is to be regarded as part of a spindle buzz. The present specimen is flat on the under side, convex on the upper, and perforated through the center. The upper surface is rather elaborately ornamented in phase 2 of style 2. Around the periphery is a band formed of two lines with alternate spurs pointing inward, and about half way up there is a similar band. The spaces between the bands and at the center are divided off by means of narrow inverted V-shaped figures, and between these are pairs of slightly curving lines connected at the center by a small nucleated circle; rising from the curving lines are pairs of short lines terminating in circles to which are attached three equidistant spurs. Plate 67, figures 6 and 8, are two objects which I am unable to identify. They are straight across the lower end, rounded above, and have in the base a deep cylindrical socket. Figure 8 seems to have had a link pendant at the end; the decoration consists of very deeply incised lines, some straight, some V-shaped and others curving around to the opposite side. The lines on figure 6 are less deeply incised ; at the base is a band containing long alternating spurs and from this rises a pair of vertical lines which are joined at the center by two oblique bands of three lines each in the open spaces; above these are two short detached lines each with a pair of downward-pointing spurs attached. Red pigment still remains in most of the incisions, as 1s so often the case with objects bearing a Punuk decoration. NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 197 Just as at Ievoghiyoq, the harpoon heads at the two latest sites were for the most part undecorated. Of the 80 harpoon heads and frag- ments found at Seklowaghyaget, 14 were ornamented ; of the 68 from the old section of Gambell, only 2. One of the decorated heads from Seklowaghyaget (cut 2, 49 in.) is shown in plate 71, figure 2. It has the phase I, style 2 decoration of lines and long spurs which seems to have been the earliest style present at Ievoghiyoq. Three other Seklowaghyaget heads have this same decoration and all of them came from the deeper levels—34 to 52 inches. On the basis of these few specimens, therefore, it would appear that the earliest period at Seklowaghyaget was approximately contemporaneous with the earli- est period at levoghiyoq. However, from the evidence afforded by other implement types, particularly the forms of the harpoon heads themselves, it is clear that Seklowaghyaget was occupied for a con- siderable period after Ievoghiyoq had been abandoned. Of the 10 other decorated harpoon heads from Seklowaghyaget, 3 are to be classed as phase 2 of style 2,6 as phase 3 (including plate 71, figure 14), and I as indeterminable. Of the two decorated heads from the old section of Gambell, one (from cut 1, depth 16 in.) has the deep lines, spurs, and dots of phase 2, style 2; the other specimen (from cut I, 24 in.) is so badly weathered that the design is unrecognizable. The early Punuk stage witnessed an interesting development in those artifacts which for want of a better name have been referred to on previous pages as “ winged objects’, the Old Bering Sea forms of which were illustrated in plates 12, 20, and 21. In plates 68 and 69 we have views of both sides of eight specimens showing the forms that these “ winged ” objects assumed during the Punuk stage. The initial step in the transition from the Old Bering Sea to the Punuk forms is not illustrated by the material excavated at Gambell, but fortunately there are two specimens which do show it. The first of these is the winged object from Point Hope figured by Mathiassen (1929, p. 45). The second, which is very similar in form and orna- mentation, is shown in text figure 23; this specimen was purchased by Dr. Hrdlicka, and although the exact provenience is unknown, it probably came from St. Lawrence Island or Little Diomede Island. The central projection is broken off on both this object and the one figured by Mathiassen. Structurally, these two objects are closer to the Old Bering Sea than the Punuk type: both have the Old Bering Sea form of socket—square or rectangular—and in the specimen shown in text figure 23 there are round lashing slots leading into the socket from the sides; the wings are still rather wide, although they 14 198 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 are pointed instead of rounded at the ends and their sharp upward inclination approaches the Punuk trident; the notches at the bases of the wings are the same as on the Old Bering Sea forms. On the back side (text fig. 23, b, and Mathiassen, fig. 19, b) the raised central section retains the general Old Bering Sea outline, although the short inner “ wings”? are narrower and are directed upward rather than outward. This is the crucial step in the metamorphosis, for it was this Fic. 23—Ivory “ winged” object with early Punuk decoration, northern Alaska. raised central portion, divested of the flaring outer wings, that later developed into the Punuk trident. The ornamentation on both speci- mens is typically Punuk—straight or slightly curving lines and dots, the latter either free or enclosed in the spaces of a “ ladder ”’ design. Another stage in the transition is represented by two specimens from Punuk Island and Cape Kialegak previously described by the writer (Collins, 1929, pl. 10, a-b; pl. 13, b). Here we have the square basal socket characteristic of the older forms, and wings that, although pointing upward in a Punuk fashion, are still relatively broad and NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 199 flaring. The central projection and the ornamentation on these two specimens are typically Punuk. From this we come to the dominant, or later, Punuk form, that of a trident, as illustrated in plate 68, figures 1-5. Figures 1, 2, and 5 are from levoghiyoq; figures 1 and 2 are both from cut 4, depth 31 inches, figure 5 from cut 5, depth 24 inches. There are also two frag- ments—pieces of wings—from cut 6, depth 36 inches, and cut 8, depth 13 inches. Figures 3 and 4 were purchased from the Eskimos and are said to have come from Seklowaghyaget. The sockets on all of these are perfectly round, about 1.2 cm in diameter and 2.1 to 3.2 cm deep, and there are no lashing holes. A piece of the wooden handle or shaft remains in place on the broken specimen, plate 68, figure 5. The central projection is rather prominent, being of approximately the same length as the wings, and at the end is the shallow pit that is always present, both on the Old Bering Sea and Punuk forms. The wings are much narrower than formerly and are homologous to the inner wings of the Old Bering Sea type, which were visible only on the back side. The flaring outer wings, which were so prominent on the Old Bering Sea pieces and which gave them their beautiful sym- metry, are greatly reduced and appear only in the form of small tri- angular projections at the base. On plate 68, figure 4, even this vestige of the outer wings has been lost. The incised designs on these five specimens are all generally similar in arrangement. All are to be classed as Punuk style 2, mainly because of the deep, sharply cut lines; plate 68, figures 1-3 and 5, have the lines, dots and compass-made circles characteristic of phase 2 of style 2, whereas figure 4 with its straight, short cross lines, falls into phase 4. Plate 68, figures 7 and 8, are two objects which, though wingless, are genetically related to the above. Both were purchased from the Eskimos who had excavated them at Seklowaghyaget. They appear to be late forms—probably the latest—of the winged-trident objects, which have completely lost the wings but retained the thickened body and central projecting element. The basal socket and the opposite notch are identical with those described above. Figure 7 is decorated in phase 4 of Punuk style 2. Plate 68, figure 6 (from Seklowaghyaget, purchased) represents another Punuk variant of the “ winged ” object, several of which have been described previously (Collins, 1929, pl. 10, c-d, e; Hrdlicka, 1930, pl. 23, b). Here again we have the cylindrical basal socket and upper notch, but the wings, instead of rising free, are attached at the top to the upright central element. It has vestigial outer wings—small triangular projections along the sides—just as in the case of plate 68, figures 1-3 and 5. The surface decoration of deeply incised lines— 200 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 single, double, or in horizontal bands of four—is to be classed as phase 4 of Punuk style 2; most of the incisions are filled with red pigment. There is no stratigraphic evidence to indicate the age of this form in relation to the trident. Typology, however, points clearly to the trident as the older form, one which through a series of intermediate stages is linked with the winged form of the Old Bering Sea period. The decoration on the “ turreted”’ pieces is likewise of a later style (phase 4 of style 2), whereas that most commonly seen on the tridents is phase 2 of the same style. On the other hand, there is strong proba- bility that the latest of the tridents, e. g., plate 68, figure 4, were con- temporaneous with the turreted forms. In no other class of artifacts is the Eskimo’s feeling for form and symmetry better exemplified than in these “ winged ”, “ trident ’’, and “turreted”’ objects. The Old Bering Sea forms with their broad spreading wings, some of them resembling a whale’s tail, some a bird or a butterfly, possess all the grace and symmetry of these same forms in nature. Through gradual modification, in which the wings become more and more attenuated, there evolves a different form, the trident; this form, although having in its perfected stage no exact counterpart in nature, is a concept possessing an intrinsic esthetic appeal which no doubt accounts for its wide distribution in the Old World among both primitive and civilized peoples. As a final stage there is the more angular “ turreted”” form, in which the wings and center are united. Here we have strict artificiality, somewhat remi- niscent of a heraldic device; but this last stage, though far removed from the flowing beauty of the earlier forms, has a straightened and pleasing symmetry of its own. What may we say as to the function of these curious objects? Ex- planations advanced by the modern Eskimos are of little assistance, since they have no recollection of even the most recent forms. Some of the St. Lawrence Island Eskimos thought that the tridents might have been ornaments for war helmets and the fact that the Japanese warriors of feudal times wore metal helmet ornaments of a somewhat similar shape would seem to lend plausibility to such an explanation. However, this would fail to explain the use of the earlier winged forms of the Old Bering Sea culture since there is no evidence that defensive armor of any kind was used on St. Lawrence Island until well into the Punuk period. Gordon (1916, p. 62) was of the opinion that the winged objects were used in whaling ceremonies, and on the whole this seems as good an explanation as any, even though Gordon was undoubtedly mistaken in assuming that they had been so used in modern times. Gordon also considered that they were genetically re- lated to the prehistoric banner stones of the eastern part of the United NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 201 States, but as | have suggested before (1929, p. 9) this seems unlikely in view of the great distance separating the two forms. In a way these winged objects recall some of the canoe ornaments from Micro- nesia (Miller-Wismar, 1912, fig. 11), particularly the Caroline Island form in which the pointed, flaring ends are connected to the center, just as in the case of the “turreted”” Eskimo form; but it would seem hardly likely that these were related in origin. There is, however, still another possibility to be mentioned, namely, that these objects may have been “ wings” for the butt end of a harpoon or dart used with the throwing board. It will be recalled that all of them have a basal socket seemingly designed to hold a handle or shaft tightly in place. This may take the form either of a deep cylindrical pit in which a shaft would fit tightly, or a square or rectangular opening with lateral slots for a lashing. More significant, however, is the invariable shallow notch or groove at the end of the central projection. This is pre- cisely the kind of a pit that would be necessary to engage the bone or ivory spur at the end of the throwing board; the fact that such a pit is present on every specimen that has been found and that very often these pits show signs of wear, strongly suggest that they had a definite function. It would not be difficult to think of the “ trident ” and “ turreted ” forms as adapted to such a use, for they would proba- bly have been as effective as the bone harpoon “ wings” used by the Greenland Eskimos. It must be admitted, however, that some of the larger and heavier “ winged” forms of the Old Bering Sea period would seem very poorly adapted to such a purpose, and this alone would be sufficient reason for looking for a better explanation. For the present, therefore, the interpretation of these curious objects must remain in doubt, but whatever it may be, it must satisfactorily explain the presence of the small notch at the upper end. This seem- ingly insignificant feature evidently had a function just as definite as the basal socket, and it will, no doubt, provide the final explanation of the manner in which these objects were used. Were it not for the fact that the tridents all have this shallow notch, and the further fact that the form represents only one stage in a demonstrable sequence, we might be inclined to look for a genetic relationship between them and certain more or less comparable Asiatic trident forms such as the Yenisei shaman’s staff (Karutz, 1925, p. 95, figs. 2-4), old Japanese helmet ornaments, Indian, Chinese, and Tibetan standards, etc. In view of other Asiatic affinities of prehis- toric Eskimo culture such a connection cannot be ruled out in advance ; the present evidence, however, points to convergence as the most likely explanation of the resemblances between these widely separated trident forms. SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 202 Lgz v C1 z zz of gz 09 61 I¢ Sz LY 8 a aaa (2110) gz a I Cy rome rome arene aus apistiH g i z I Pe eg ate eee ut 96 03 zZ £z ¢ ( 9 TES. |e Se ales ut 12 0} gt gz z Z II 9 ee Se StL On he gz I 8 © 9 v 17 "ut €Z OF BRJANG “a pue ‘Ss ‘ysemoalpy z z Soe ET QOWOIeL ZI 9 ¢ z v Suey Z OnUg ir, ov I + ve + z I 1 Seniesa LVOl ee is I Zz oI ol ze I S I I “ur €Z OF VdejJANG “AY pue “NX ‘ysemodrpy 14 z z TS Gate Lovey If z v I 9 ZI 9 She POLARS, cr pout iS ag OL z AM “ul © OF VDBJANG : boArysoaay] I I Ae CL ON IGY 9 I I I e aS ey LO Nes g ater mais 9 z alo “- ur €Z 0} BeVPING : :JoseAyseMOpyaS € I I I “* + Taquuresy ‘uOT}DaS PID b aseyg | ¢ aseyg | Z aseud | JT aseyd | Z eed | I 2eUd o[qeurur 9[(qeuru ¢ Z T qe -Ioyopuy -I9}opuy 3TAWS ayAIs JAS 1H0L | usrapow Z PAIS T 2149S qe yung He vag suLeg PIO SINS NaQuUevy aary ay, 1p sq2alqQ payv4soraq fo uoyngrajsy~q— | AAV L NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 203 Harpoon HEADS FROM IEVOGHIYOQ From Ievoghiyog there are 133 harpoon heads sufficiently complete for type identification, in addition to a number of fragments and un- finished specimens. Oren Socket Harpoon Heaps Open socket type II (b) *«—Two lashing slots; bifurcated, sym- metrical (median) spur; round line hole ; two side blades parallel with the line hole (pl. 70, fig. 1.) A comparatively unimportant type, represented by the single speci- men shown in plate 70, figure 1, from cut 1, depth 12 inches. The spur, although broken, was symmetrical like that of plate 28, figure 5. The side blades had been inserted in a slot which passes through from one side to the other ; no end blade. Open socket type II (b) y—Same as above except that the side blades are at right angles to the line hole and that the latter may be triangular in shape (pl. 70, fig. 2). This type is also represented by the single specimen illustrated, from cut 5, depth 18 inches. The material is bone, and the remaining side blade is made of shell; the line hole is triangular. From cut 5, depth 50 inches, there is the upper end of a head corresponding to this, but made of ivory and with a round line hole. Open socket type II (c) x-——Two lashing slots; a single, lateral spur; triangular line hole; two side blades, parallel with the line hole (pl. 70, fig. 3). The specimen illustrated, from cut 1, depth 22 inches, is the only example of the type. It is of bone and the side blades were of shell ; the latter were set in a slot which passes completely through from one side to the other. Open socket type II (c) y—Same as above but with the side blades at right angles to the line hole; the latter is round. Two examples, both fragmentary and of bone, from cut 1, depth 28 inches, and cut 9, 34 inches. Open socket type III (a) x«—Defined on page 118. (PI. 70, figs. 8-15). This, the dominant form of harpoon head at Ievoghiyogq, is represented by 98 specimens, distributed as follows: Cut 1, depth 12 inches (4); 22 inches; 28 inches; 31 inches; 34 inches; 40 inches; 44 inches (2). Cut 2, 16 inches ; 30 inches ; 37 inches. Cut 3, 12 inches. Cut 4, 42 inches (2) ; 59 inches (2) (pl. 70, fig. 8). Cut 5, 12 inches (12); 18 inches (3) ; 24 inches (2) ; 28 inches (2) ; 31 inches (4); 33 inches (2) (pl. 63, fig. 10) ; 39 inches (4). Cut 6, 5 inches (6) 204 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. g6 (pl. 70, figs. 13, 14); 9 inches (4) ; 22 inches (5) (pl. 70, fig. 12) ; 30 inches (pl. 70, fig. 11) ; 36 inches (2) (pl. 70, fig. 9). Cut 7, 8 inches ; 18 inches (2) ; 24 inches (2). Cut 8, 8 inches (2) ; 25 inches; 28 inches; 30 inches; 34 inches (2); 39 inches (3); 44 inches (3) (pl. 70, fig. 15). Cut 9, 12 inches (4) ; 19 inches (4) ; 31 inches (3) ; 34 inches (3). This large group includes a few harpoon heads which are identical in every respect with some of the type III (a) x heads from the northwestern section of Miyowagh. The greater number, however, although still falling within the limits of the type as defined, have become modified in several respects, principally in the shape of the spur. We will consider first the specimens which are directly com- parable with those from Miyowagh. Plate 70, figure 8, from cut 4, depth 59 inches, is one of these; in both form and ornamentation it is identical with the two heads from Miyowagh shown in plate 28, 13 and 14. There are five other open socket heads from Ievoghiyoq with the same ornamentation, and significantly enough, all of these come from the lower levels, from depths ranging from 31 to 59 inches. Plate 70, figures g and 15 are two more specimens that are very simi- lar in form to the Miyowagh examples shown in plate 28, figures 8- 14. However, they lack the small elevations around the line holes which were present on the Miyowagh heads; and the decoration in figure 15, consisting of deeply cut lines, spurs, and mechanically made circles, can be compared only with two of the latest of the Miyowagh heads, plate 28, figures 24 and 28. It is, therefore, only the latest style of ornamentation at Miyowagh (pl. 28, figs. 13, 14, 24, 28) that is carried over on these open socket heads at levoghiyoq ; among all of the decorated harpoon heads from the latter site, with the pos- sible exception of plate 70, figure 21, there is not one with the lightly incised line, spur, and dot ornamentation that was so characteristic of the earlier phase of the Punuk at Miyowagh as illustrated by such specimens as plate 28, figures 8-12, 15-17, 19-22, 25-27. It is to be noted further that only a small number of the type III (a) x heads at Ievoghiyoq are decorated at all—17 out of a total of 98, in con- trast to Miyowagh where all but one of the 4o heads of this type were decorated. The majority of the open socket heads from lIevoghiyoq, although possessing the general features that bring them within the defined limits of type III (a) x, have nevertheless become modified to such an extent that they stand clearly apart from the older examples of the same type from both Ievoghiyog and Miyowagh. The principal differ- ence is in the altered form of the spur. In the Miyowagh heads the NOS). I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 205 spur extends straight down and is fairly wide across the tip; in the later Ilevoghiyog heads the spur flares outward and becomes narrow and arched, the sides having been cut down at an acute angle from the median ridge, which is thereby made to stand out more prominently (pl. 63, figs. 10-14). This arched form of spur, in the most typical examples, is still further accentuated by having the tip cut off obliquely, giving it the appearance of curving inward, as in some cases (e. g., fig. 12) it actually does. About 90 percent of the Ievoghiyoq heads have this arched form of spur in some degree, and, just as in the case of ornamentation, those which do not have it come for the most part from the lower levels of the midden. As a rule the line holes on the levoghiyog heads are somewhat larger than those from Miyowagh and the shallow groove beneath the line hole is usually lacking. In addition to such specific differences in individual features, many of the later heads from levoghiyog are heavier and thicker, particularly around the lower end (pl. 70, figs. 10, 11, and 14). Plate 70, figure 9, has part of the baleen lashing remaining in one of the slots, and the same is true of one other specimen from cut 5. Plate 70, figure 16, from cut 9, depth 12 inches, possesses several features that set it apart from type III (a) x. It has a bifurcated spur, a groove opposite the single lashing slot, and a deeply cut Y-figure above the line hole on both sides; furthermore, it is of bone, whereas all but one of the other type III (a) x heads are of ivory. Open socket type III (a) y.—This type, with the blade slit at right angles to the line hole, is not represented at levoghiyog. As noted previously, the distribution of this type at Miyowagh indicated that it was somewhat earlier than type III (a) x, in which the blade slit and line hole were parallel. The present evidence is confirmatory in this respect, the type apparently having been discontinued before levoghiyoq was established. Open socket type IV.—Two lashing slots; single, lateral spur; round or triangular line hole; two prominent lateral barbs, at right angles to line hole; no end blade; undecorated. Differs from type II (e) in the lashing arrangement and size of the barbs. (Pl. 70, figs. 4, 5.) There are nine examples of this type, distributed as follows: Cut I, depth 22 inches (pl. 70, fig. 4) ; 31 inches. Cut 5, 24 inches (2, one bone, one ivory—pl. 70, fig. 5). Cut 6, 9 inches; 36 inches. Cut 8, 25 inches. Cut 9, 31 inches; 34 inches (bone). In form these har- poon heads are identical with Mathiassen’s Thule type 2. However, they do not have the Y-shaped figure above the line hole as do the 206 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 Thule heads, and most of them are made of ivory, whereas the Thule heads are of bone. Six of the Ievoghiyoq specimens have triangular line holes, like plate 70, figure 4; three, including figure 5, have round line holes. One of them—from cut 8, 25 inches—departs somewhat from the type in having the end of the spur notched. Open socket type IV (a) «—Same as preceding type but with two pairs of barbs, and an end blade, parallel with the line hole. (PI. 70, fig. 6.) This variant of type IV is represented by the single specimen illus- trated, from cut 8, depth 28 inches. The material is bone. CLosEpD Socket Harpoon Heaps Closed socket harpoon heads occupied a subordinate position at Ievoghiyoq; only 19 were found as compared with 114 of the open socket forms. Furthermore, only 5 of the 19 are directly comparable with the Miyowagh heads, the others having undergone modifications that give them a distinctive appearance, just as was the case with the open socket type III (a) x heads from Ievoghiyoq. Closed socket type V «—Definition on page 110. (PI. 70, figs. 17, 18.) There are six specimens that fall into this group: Plate 70, figure 17, cut 9, depth 31 inches; plate 70, figure 18, cut 5, 12 inches; and four others from depths of 5, 18 (2), and 44 inches. The small specimen, figure 17, in a general way resembles the two heads from Miyowagh shown on plate 28, figures 21, 22, although its contour is different. The ornamentation is basically the same as that on other closed socket heads from Miyowagh, e. g., plate 28, figures 19, 20, 25. Plate 70, figure 18, has a sharply beveled spur like those on the open socket heads, figures 10-14. However, it is pointed at the end instead of being cut off obliquely. This specimen bears a typical late Punuk ornamentation of deeply cut lines and spurs, which were originally filled with red pigment, traces of which still remain. Another head of this type is ornamented in a similar fashion with lines, spurs, and small. deep, round pits, and in this case the red pigment is still in place. The three remaining heads of type V x are undecorated. Closed socket type V (b) «—Long, tapering, pointed spur; round line hole, beveled below; end blade parallel with line hole; oval to almost round in cross-section, with a constriction between line hole and tip; decoration Punuk, or plain. (PI. 70, figs. 19, 20.) Five examples, as follows: Cut 1, 22 inches; 40 inches (pl. 70. fig. 20) ; 50 inches (pl. 70, fig. 19). Cut 2, 30 inches. Cut 9, 19 inches. This is a specialized form of the general type with line hole and blade slit parallel, but one which is so different in contour, shape, and NO. 1 ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 207 style of spur that it seems advisable to describe it as a subtype. To judge from the depth distribution, it came into use at a rather early period at Ievoghiyog. The spur, instead of being sharply arched as in the case of the prevailing open socket forms, is relatively wide above, tapering down to a long, sharp point that flares out from the body. Four of these heads are undecorated. The fifth (pl. 70, fig. 20) bears an ornamentation generally similar to that of plate 70, figure 15— deeply cut lines, short spurs, and compass-made circles with central pits. Plugs of baleen were set in the latter, and all of the incisions were filled with red pigment. Closed socket type V y—Defined on page 111. (Pl. 70, fig. 21.) This type, which is carried over from Miyowagh without change of form, is represented by only two examples, plate 70, figure 21, cut 4, depth 42 inches, and an unfinished specimen from cut 8, depth 39 inches. Figure 21 has the same graceful outline, the median ridge between line hole and tip, and the small rounded elevations around the line hole that were observed on some of the Miyowagh examples, such as plate 28, figures 25, 26. The decoration, furthermore, falls within the defined limits of Punuk style 1, consisting of incised lines and long spurs terminating in small round pits although the lines are much more deeply incised than on the comparable examples from Miyowagh. As in the case of so many of the decorated objects from this site, the deep incisions and pits were originally filled with red pigment. Closed socket type V (a) y.—Differs from type V y in general contour and in the shape of the spur. The upper part of the body is straight and the median ridge between tip and line hole is less pro- nounced than in type V y. The spur is like that of open socket type III (a) x (pl. 70, figs. 10-14), having an outward flare and being sharply beveled or arched, with the end cut off obliquely. (PI. 70, figs. 3, 23.) There are five heads of this type, as follows: Cut 2, depth 22 inches (pl. 70, fig. 23). Cut 5, depth unknown (2). Cut 6, 9 inches (2) (pl. 70, fig. 22). Four of these are decorated, all with deeply incised lines and short, oblique spurs; all of them have had red pigment rubbed into the lines. Closed socket type VI.—Single, lateral spur; triangular line hole; two prominent lateral barbs, at right angles to line hole; no end blade ; undecorated. (PI. 70, fig. 7.) The specimen illustrated, from cut 1, depth 44 inches, was the only one of the type found at Ievoghiyoq. In spite of the closed socket its closest relationship is to the group of open socket heads shown on 208 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 plate 70, figures 4-6, i. e., open socket types IV and IV (a) x. It was apparently an unimportant type on St. Lawrence Island, although it occupies a prominent place on the Arctic Coast, where it appears as a comparatively late derivative of the Thule type 2. SUMMARY The harpoon heads from Ievoghiyoq, although including some ex- amples identical with those from the later section of Miyowagh, have for the most part undergone considerable modification. The pre- dominant form, open socket type III (a) x (pl. 70, figs. 8-15), in- cludes a few heads from the lower levels, e. g., figure 8, which are indistinguishable from some of the Miyowagh specimens. The great majority of heads of this type, however, have taken on a distinctive appearance ; the spur has become thin, rather sharp edged and flaring, with the end cut off obliquely; the line hole has become somewhat larger, and there is a general tendency toward heaviness, particularly of the lower end. Correlated with this change of form is an absence of ornamentation. None of these later heads are decorated, whereas on the other hand most of those that have retained something of the old form bear a Punuk ornamentation comparable to that which appeared during a late stage at Miyowagh. The closed socket heads show similar resemblances and differences. A definite linkage between levoghiyog and Miyowagh is provided by a few of the closed socket forms such as plate 70, figures 17 and 21 (types V x and V y), but a majority of them have become specialized in much the same manner as the open socket heads. The spur has either become arched as on the open socket forms or long and pointed, and the graceful, tapering outline has been lost. Among the new forms at Ievoghiyog is open socket type IV, with two prominent lateral barbs but no blades of any kind. In general outline these heads correspond exactly with Mathiassen’s Thule type 2; open socket type IV (a) x is a closely related form with two pairs of barbs and also an end blade. The subtypes of type II, plate 70, figures I-3, intergrade with similar forms from the later section of Miyowagh (pl. 28, figs. 1-3) and are probably contemporaneous ; and just as at Miyowagh, most of these type II heads are made of bone. Another point to be noted is the preponderance of open socket heads at levoghiyoq—114 as compared with only 19 with closed sockets. At Miyowagh, on the other hand, 4o percent of all the heads had closed sockets. NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 209 Harpoon HEADS FROM SEKLOWAGHYAGET The harpoon heads from Seklowaghyaget, numbering 54, are of particular interest inasmuch as they provide a definite linkage between the prehistoric forms already described and those in use at the present time. Open SockeT Harpoon Heaps Open socket type III (a) +—Defined on page 118. (PI. 71, figs. 2-4.) This type of harpoon head, which was found in such large numbers at both Ievoghiyog and Miyowagh, is also well represented at Seklowaghyaget, even though it has begun to give way to another type, to be described presently. The 20 type III (a) x heads from Seklowaghyaget were distributed as follows: Cut 1, depth 38 inches. e)2, 18 inches; 49 inches (pl. 71, fig. 2). Cut 3, 12’ inches (2); ie inches (2) (pl. 71, fig. 3); 30 inches; 34 inches (2); 42 inches. Cut 4, 12 inches; 18 inches (pl. 71, fig. 4) ; 44 inches (2). Cut 7, 12 inches; 20 inches. Cut 8, 12 inches; 19 inches ; 67 inches. Eight of the above heads, including plate 71, figure 2, are decorated, all of them in phase 1 of Punuk style 2 which was found at Ievoghiyoq and also at the later section of Miyowagh. Three of them have traces of red pigment remaining in the incisions. The older type of spur— relatively wide and only moderately beveled—was present in only three instances: on the head shown in plate 71, figure 2, and on two other heads which are practically identical in form and ornamentation, from depths of 34 and 38 inches. Eleven, including plate 71, figures 3 and 4, have the later form of spur—sharply beveled and with a slight inward flare—which was so common at Ievoghiyoq. On the remain- ing six the spur is not clearly distinguishable. Open socket type III (b) x.—Two lashing slots, one slot with an opposite groove, or no slots; a single lateral spur which is usually made asymmetrical by an acute beveling on the socket side; slanting walls produce a wedge-shaped socket in which the lower (outer) end and the floor are wider than the upper end and the top; line hole usually round or triangular, sometimes squarish; end blade, parallel with line hole; undecorated. (PI. 71, figs. 5-9.) Twenty-one examples, distributed as follows: Cut 1, 12 inches. Cut 2, 12 inches (3) (pl. 71, figs. 6 and 8); 18 inches (2) (pl. 71, fig. 5) ; 22 inches (5) (pl. 71, fig. 9) ; 28 inches (pl. 71, fig. 7). Cut 3, I2 inches. Cut 7, 20 inches (3) ; 28 inches; 36 inches; 44 inches (2). Cut 8, 39 inches. This specialized form of type III (a) x is marked by a series of rather sudden developmental changes, particularly in the region of 210 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 the socket. The most important of these is the manner in which the socket itself is formed. The open sockets which we have heretofore considered have been straight-sided and perfectly rectangular in shape. Since they were as wide across the top as along the floor, the foreshaft rested lightly in place and was only prevented from slipping out by means of a lashing which passed through a pair of slots and across the open socket. In the present group, the two lashing slots or a single slot with an opposite groove are retained in a number of cases, but in addition to this method of lashing, the socket is cut out in such a manner that it alone will at least partially hold the foreshaft in place. This was accomplished by making the socket walls slant inward ; with the floor of the socket now wider than the opening along the top, the foreshaft remains in place without the aid of a lashing, just as in the case of a completely enclosed socket. The wedgelike appear- ance of such sockets is further increased by having them taper from the base upward, so that in an extreme form, like plate 71, figure 8, the width of the socket at the upper end is only half that of the lower end. The symmetrical lateral spur of type III (a) x is here replaced by one which is made asymmetrical by having the lower end on the socket side beveled off obliquely. A few, like plate 71, figure 5, have retained the older form of spur, uniformly beveled from both sides. The line holes are more variable both as to size and shape than any previously considered. Most of them are rounded and relatively large, others are triangular, either large or small, and still others that were originally round have by secondary cutting taken on a squarish shape. Along with these specific changes of the spur and socket is a change in contour. The majority of the type III (b) x heads differ from those of type III (a) x in their somewhat smaller size and straighter lines ; they have lost completely the graceful shape that in the past was always associated with the type III head. As to material, 19 of the type II (b) x heads from Seklowaghyaget are of ivory; 2, including plate 71, figure 7, are of bone; all are undecorated. It should be mentioned here that one harpoon head of type III (b) x was found at Ievoghiyoq, cut 9, depth 31 inches. However, there is a strong probability that this specimen is intrusive, for it was the only one of the large number of harpoon heads from that site which showed any inclination to the peculiar form of socket characteristic of this type. Open socket type III (b) y—Same as above but with the blade slit at right angles to the line hole. (PI. 71, fig. 10.) The specimen shown in plate 71, figure 10, from cut 2, 18 inches deep, was the only example of this type found at Seklowaghyaget. SL NO. T ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 211 The line hole is large and squarish. Except for the direction of the blade slit, this specimen conforms in every way with type IIT (b) x. Open socket type III (c) x.—Differs from type III (b) x in having a pair of small barbs in addition to an end blade. (PI. 71, figure 11.) The specimen illustrated, from cut 2, depth 22 inches, is the only example of this type. Part of the upper end is broken off, and also a small piece at the socket. Open socket type IV (a).—One lashing slot and an opposite groove ; wedge-shaped socket, as in type III (b) x; single (?) lateral spur ; triangular line hole; two pairs of prominent barbs, at right angles to line hole ; no blade ; undecorated. (PI. 71, fig. 1.) The specimen illustrated, from cut 2, depth 22 inches, was the only one found. It is another western variant of the Thule type 2 head. The triangular line hole, the bladeless tip, and probably the spur are features which correspond with our type IV; the two pairs of barbs are like type IV (a) x. The peculiar form of the socket, however, brings it into relationship with type III (b) x. The spur is damaged, so its exact form must remain in doubt ; however, it was placed later- ally, and very likely was single. The deep, narrow groove above the line hole is like that in plate 70, figure 6. Open socket type V.—Two lashing slots, one lashing slot with an opposite groove, or a groove alone; socket and spur as in Type III (b) x; line hole triangular, rarely round; no blade; undecorated. Pel) 71, figs. 12, 13.) Three examples of this type were found at Seklowaghyaget, the two illustrated, from cuts 2 and 7, depths 18 and 12 inches respectively, and another picked up on the surface. In its general shape and lack of an inserted blade, this type conforms to Mathiassen’s Thule type 1, but the peculiar wedge-shaped socket and beveled spur relate it to our types III (b) x and y. CLosEp Socket Harpoon HEaAps Only seven harpoon heads with closed sockets (exclusive of those used in whaling) were found at Seklowaghyaget. Closed socket type V y—Defined on page 111. Only two of this type were found, an unfinished specimen from cut 3, depth 62 inches, and a smoothed, waterworn specimen from cut 6, depth 28 inches. This was a type which apparently passed out of use soon after Sek- lowaghyaget was established. Closed socket type V (a) y.—Defined on page 207 (pl. 71, fig. 14). There are three specimens of this type, as follows: Cut 3, depth 30 inches ; cut 4, depth 12 inches (pl. 71, fig. 14) ; cut 8, depth 60 inches. i 212 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 Figure 14 bears a phase 3, Punuk style 2 ornamentation of deeply cut lines and short oblique spurs; traces of red pigment remain in the incisions. The other two heads are undecorated. Closed socket type V (b) x.—Defined on page 206. This type is also represented by two specimens. They come from cuts 4 and 7, at depths of 34 and 3 inches respectively. Both are decorated in late Punuk style, with deeply incised lines, spurs, dots, and circles, and one has red pigment remaining in the incisions. SUMMARY We have seen from the above that some of the harpoon heads from Seklowaghyaget are identical with the prevailing types from levoghi- yoq. These are open socket types III (a) x (pl. 71, figs. 2-4) and closed socket types V y, V (a) y, and V (b) x (pl. 71, fig. 14). These forms indicate contemporaneity with Ievoghiyoq and the latest period at Miyowagh, but there are several new types which show that Seklow- aghyaget continued to be occupied after the abandonment of both of the other sites. The most important of these new types is open socket type III (b) x (pl. 71, figs. 5-9), a specialized form of type III (a) x, characterized especially by a wedge-shaped socket which in some cases functions as a closed socket. This peculiar form of socket is also found on the less common open socket types III (b) y, II (c) y, III(c) x, and V; and, as we shall see presently, it is the predominant form of socket at the latest of the midden sites (the old section of Gambell) and at two house ruins of comparatively recent date. Harpoon HEADS FROM OLD SECTION OF GAMBELL AND Houses 8, 9, AND 10 House no. 10, it will be recalled, was the most recent of the 10 house ruins excavated (pl. 64, fig. 2). According to native testimony it was the last of the semisubterranean dwellings at Gambell, having been occupied up to about 50 years ago. The correctness of this state- ment was proven by the material obtained: five harpoon heads of modern type, one of which is shown on plate 71, figure 26; broken pottery lamps and cooking pots of modern type, the former with high wick ledges, the latter with suspension lugs; considerable quanti- ties of hoop iron, copper, etc. ; small ivory bird figures ; flat bone sledge shoes ; and various other modern types of implements. NO: I ARCHEOLOGY OF ST. LAWRENCE ISI-AND—COLLINS 213 House no. 8 (described on p. 187, pl. 63, figs. 2-5) was older, dating probably from about the eighteenth century. Lamp and pot frag- ments were like those from house no. 10, but no other implements of strictly modern types were present. There were nine harpoon heads, six of which are shown on plate 71, figures 15-20. Open socket type III (b) x is represented by five examples, including figures 18, 19, and 20. The last of these is of particular interest, since it represents a very close approach to the modern form. In fact, it is only the wedge-shaped open socket that brings this head into type III (b) x; the rounded contour and the large triangular line hole with deeply beveled lower border are characteristic features of the modern St. Lawrence type. Of the bladeless form, open socket type V, there are three examples, all of bone, two of which are illustrated (pl. 71, figs. 15, 16). These are of smaller size than the previous examples of this type from Seklowaghyaget. Figure 17 is apparently a slightly older form of the bladeless head, for it lacks the wedge-shaped socket. House no. 9 was apparently contemporaneous with house no. 8. Only two harpoon heads were found here, one of open socket type MT (a) x, the other of type III (b) x: The excavation in the shallow midden deposit adjoining the present village of Gambell yielded 39 harpoon heads and fragments large enough for type identification. Among these was a single example of open socket type III (a) x (from cut 1, depth 16 in.), a badly weathered, broken specimen with Punuk ornamentation, probably a relic from an earlier period. The same may be true of the only closed socket head found at this site (cut 1, depth 24 in.). This was a small, much worn specimen of closed socket type V y, which had originally been decorated in Punuk style. Since these two heads, which are typically Punuk both in form and ornamentation, are weathered to a much greater degree than any of the others from the same site, it seems reasonable to suppose that they are older specimens, belonging perhaps to the adjoining site, Seklowaghyaget. There are three examples of the bladeless, open socket type V, all from cut I, at depths of 8, 16, and 32 inches. The first two are shown on plate 71, figures 22, 23; figure 23 and the one not illustrated are of bone. It will also be observed that figure 23 differs from the usual form in having the bladeless upper end flattened transversely, so that the cutting edges lie parallel with the line hole instead of at right angles to it. Of the remaining 34 specimens, all but one belong to open socket type III (b) x, the form with narrow, wedge-shaped socket and 15 214 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 acutely beveled spur which has already been described from Seklow- aghyaget and houses no. 8 and g. The depth distribution of the type III (b) x heads is as follows: Cut 1, depth 8 inches (7) ; 16 inches (5); 24 inches (5); 32 inches (10). Cut 2, depth 10 inches (2) (pl. 71, fig. 25); 20 inches (4) (pl. 71, fig. 24). This type of harpoon head, which is the latest of the prehistoric forms on St. Lawrence Island, affords evidence of an interesting metamorphosis, the last and one of the most striking of the long series of developmental changes which began as far back as the Old Bering Sea period at Miyowagh. This final step is of particular interest in that it shows the precise manner in which the modern St. Lawrence head developed. We have already seen how in the type III (b) x heads the inward inclination of the socket walls produced a socket which though still “open” actually functioned as a closed socket. The present examples illustrate the last stage in this process: Plate 71, figure 24, which has a socket of this kind, differs from the somewhat earlier examples of the type in having a large triangular line hole with deeply beveled lower bor- der which is characteristic of the modern type (pl. 71, figs. 26, 27) ; plate 71, figure 25, is a still closer approach, showing not only the large triangular line hole with the deep, wide groove below, but also the rounded, thickened shape of the modern head. As may readily be seen, this particular specimen and also plate 71, figure 20, (and three others not illustrated), differ from the modern type only in that they have “ open” sockets. Plate 71, figure 26, from house no. 10 (abandoned some 50 years ago), and the iron bladed head, figure 27, collected by E. W. Nelson in 1879, are examples of the modern St. Lawrence form in which a round drilled socket has replaced the wedge-shaped, barely open, cut out socket of type III (b) x. We thus have evidence of a rather un- usual condition, a late reversion or secondary evolution whereby the closed socket harpoon head, which was prominent during the Old Bering Sea period but which seems to have practically disappeared during the late Punuk, appears again as a direct outgrowth of a late form of the open socket head. Although there is clear typological and stratigraphical evidence that this particular metamorphosis actually occurred, it is not to be supposed that the idea of a closed socket per se was anything new to these Eskimos ; for whaling harpoon heads, which had been in common use from the Punuk stage on, had always had closed sockets. Plate 71, figure 29, is another modern St. Lawrence head, collected by Riley D. Moore. It is an example of the rather uncommon modern form in which the blade is at right angles to the line hole. Its prototype NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 215 is open socket type III (b) y, of which we have two examples (pl. 71, figs. 10 and 28). The first of these, from Seklowaghyaget, still retains the general shape and features of type III (b) ; figure 28, from cut 1, old section of Gambell, depth 8 inches, is a later and more specialized form, which differs from its modern descendant (pl. 71, fig. 29) only in having an “open” socket. The spur has lost the acute unilateral beveling which characterized its predecessor (pl. 71, fig. 10) and the closely related heads of type III (b) x. It should also be noted that those examples of the latter type, such as plate 71, figure 25, that have the thickened, rounded contour and the form of line hole peculiar to the modern head, also have the modern symmetrical form of spur; whereas the smaller, flatter examples, e. g., plate 71, figure 24, have the asymmetrical spur produced by an acute beveling on the socket side. . The bone harpoon head shown in plate 71, figure 21, was pur- chased from an Eskimo at Gambell, and its exact provenience is unknown. However, the wedge-shaped socket and the large triangular line hole both attest its recency. It is introduced for the purpose of illustrating two features which none of our harpoon heads happened to show: drilled lashing holes around the socket, and rivet holes at the tip for holding the blade in place. On St. Lawrence Island, as elsewhere in Alaska, it is only the modern harpoon heads which have rivet holes at the tip (e. g., pl. 71, fig. 27, and two others from house no. 10). However, a number of the Thule heads have such rivet holes (e. g., Mathiassen, 1927, plate 1, figure 6), in which the blade is either of stone or of meteoric iron. As for the drilled lashing holes around the socket, these are surprisingly rare at Gambell, even on the latest of the prehistoric heads. Nelson figures a harpoon head which has them (pl. 57, fig. 13—-U.S.N.M. no. 63334) ; the general form of this head is Thule type 2, but it also possesses an iron blade, with rivet holes. From Jenness’ work at Wales and Little Diomede Island we know that drilled lashing holes were common at Bering Strait but here again they were relatively late, the older heads all having had slots for the lashing. In this connection it is of interest to note that drilled lashing holes appear to be mere common at the old site of Kukuliak, to the east of Gambell, than elsewhere on St. Lawrence Island, if we may judge from several small collections from that site obtained by the writer. WHALING Harpoon HEapDs Whaling harpoon heads were scarce as compared with those used for seal and walrus. The excavations at the five Gambell sites yielded 216 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 only 13 more or less complete specimens and three fragments. None were found at the Hillside site, and only two at Miyowagh (pl. 72, figs. I, 2). The first of these is from cut 18, depth 60 inches. It is unusually large, measuring 21 cm in length, 4 cm in breadth, and 2.7 cm in thickness, and has a blade slit 7 cm long. It is much thicker than the modern St. Lawrence form and lacks the deep groove below the line hole. The socket also is large—1.4 cm in diameter and 2.5 cm deep—and the long offset spur is another feature not found on modern whaling heads. The decoration consists of incised lines and small round pits containing plugs of baleen. The simple arrangement of the lines does not conform very closely to either the Old Bering Sea or Punuk patterns, but the two rounded elevations on either side of the line hole, each surmounted by a smail circular pit, are common to both styles. As a whole the decoration is more suggestive of early Punuk than anything else. The second whaling harpoon head from Miyowagh is shown in plate 72, figure 2. Although smaller than the other specimen, it is similar structurally. It comes from cut 12, depth 27 inches. From cut 20, depth 12 inches, there is half of the tip of another, with a Punuk decoration consisting of straight, evenly incised lines and dots. From levoghiyoq there are no complete specimens, and only two fragments, one from a depth of 42 inches, the other purchased from an Eskimo. Both are decorated in Punuk style, with deeply incised lines, spurs, dots, and circles. Plate 72, figure 3, decorated in the same fashion, was excavated by Mr. Chambers at Kitneapalok, an old site about 20 miles south of Gambell. The surface of the ivory is much weathered, but the lines and nucleated circles of the Punuk ornamentation are still visible. The line hole has an oblique slant, but the form of the spur is essenti- ally the same as on modern examples. Five whaling harpoon heads, none of them complete, were exca- vated at Seklowaghyaget. They come from cut 2, depth 12 inches (2). Cut 7, 26 inches; 44 inches. Cut 8, 26 inches; 34 inches. Plate 72, figure 4, which was purchased from an Eskimo, is a better example of the Seklowaghyaget type than the incomplete specimens that we excavated. The deep beveling below the line hole, the curving of the lower part of the body just above the socket, and the two narrow slots along the lower edges of the spur are all features which are characteristic of the modern St. Lawrence form. From the old section of Gambell there are five specimens, two from cut 1, depth 8 inches (pl. 72, fig. 6) and three from cut 1, depth 16 {y . {eV “uV¥ jean | ed | <0ov - ~ ~~ - , «6 x I SRO SOL MAY i tad T wood wEh Yo soltadigiid—.0 gmaaT t } de i aio} todoos abgQ . eat et een nein CUapE LD CRETE BEE | x TED bated eet lial ity rr ey 1 oe ly seb aoa db hal lal as ni Sd has ASL mb. | MAAS BARE | Nak’ dante = i eran Pree raieee gee) web ra, POLE. Nim ldd bigs s ’ j ; ; yo: j | i rt } { le | _ ; “~s . . v} ) * + Peso € hie bok H aay . | ; . ee ; : ; - j : S ; 1 Lae | rae ae | ‘| 2) eee fries “onpce bee | | BETO TM AM conaie of MR de ole nl : ; d } ee lin 1 ©» jad wel . ; | if : a : : | i i dey yin tH / & £ oS in j ; «fs SATO Ne LF Gh. ; D Say} |e) Ge ie eg 1% h ) 4.3 ee Rhee ohm We | ty HE 1 - \ ea we - e« * j NIN ea i, 2 a. fe l, mm ih : j “144 2 / + has - i’ i@ “io ! f ; { t 1 yt é 4 f, pat | . aS | my PA aes oan wet, inehh Wich thetic ee cs ; Tai pi a a ee : ; ies | a i 4 i : =~ + i | I | a. a TABLE 2.—Distribution of Harpoon Types at the Five Gambell Sites Open socket forms Closed socket forms Ix | Ly |I(a)y| 1(b) [IT x|I1y/II (a) x|II (a) y|11 (b) x 11) x|11 (c) {II (c) y|IL(d)|11 (e)} IL(f)x | I1(g)x] III x } II y | ILI (a)x | I11(a)y | IIT(b)x | IIT (b)y |I11(c)x|IV|IV(a)|IV (a) x} V ||L x] Il x | IVy| Vx! Vy] V(a)x| Via)y | V(b)x |VI| Total arene EIN ee ceate Teen ele rT Lal|ly. «| MMMM aie all'ce » ans a0 coe sce ciiees caleweuafinesa|aceclesesclacecclesecalecees Zl a 38 aM IR racers cll wea Netexsealt epee: iNeai| | ebell dese allel ani Th | erates |reatenel| teeta oe 50 Seklowaghyaget: eee a | Ms evs llee ee eles ec clv ees elaceas|(neces|sars[ccusfaceeclescselescasfecees MDE ers 2. «'s 15 I VR |i | kal perf seemed hegeeews| cl Peale oe I Dm |e. -34 eae eee eeere eee mates + «|More eens Selec e euler es elie rcciedes|lecseleenc[eiececlecevslecseals reas 7 ee BaP lPoetrcs chee |la-oo) sil atss\ ares [izunccestene [havcll | soll kasevanes nt eae eS 1A ee I Eo [tei] 16 ney meee ere |. «| MEE ey live «wollen ce «fee caicfemaesilce vie olla siaileesc|ecscsieeccclocecches core eevee. «+ | MMMM 8 | ee koy are all etnias: oll atsilfe woke ataker svelte {tate le ee-a enemas eRe 1 ee Teal neeon tes 4 Ievoghiyoq: oS Bee ee Gee Peed Gee) cae ae I I Mer eater RaneRA Ma Ee fei ceva lie ss.0 avalos la gabe ACD sas. . | MMe 2 rey ye foley ace DNA, «Fc cassva call evel lferel| eects areca V1) esl | ayn 5 BF aller POR aaa ana mmAa eta ated Tatts hae} ov uel etaeety fecclV|(«» = | MEBs + cie|fnio se ofie.e esa wee sllea ces Cm MRR Pet ss |chsraitas Allee. 6 ai|ees colle: «:2he [ls © eave 5 4 " Me | Memeetevellfouesaie«/llfaresaim lam lfotaveslliehas. ocell iva Holl 3 choked eae Te, || OMe aerate cates a] Rete cain te 22 OOO ae ieiecve's + a\fe. v6 pa ie CM fers >| MMMM | ot oe 5) ole setoll clave totale lfe eatacailia.nteis lee mesa. wa c!| ower. 2 UO AES cece: ~ cota] Bape | (Preece bole [exe | Creme sect | (aed (See [ene 1] PRE PR Al b weRe (ea rate ee 7 Hillside. site... «.... 0.0 SD eee |. «| MEME wea lie cals a el'e orsictalletastclaw ooo lle'o wallaces|'ca es a)s ove 3 2 I Ti | ME rel llcasie shes eral emersiite.s. oes alteetilese I tah lerealle, tecieltepencars le waters 16 MOA e «ie este 8 Sees r | 6 |e 2 3 I 1 | I F I I I 2 | 28 14 | 161 14 60 3 To Ohler I Died 2 |30 |26 | 2 8 NV GlAl7 oy baer t » eg aoe. tg lt es rey ke “ ELT GOEL ri T ! i : I : ri Ne we am " , : | ¢ } ' h Re ‘ os j i | I 4 & i. ie | } i oe 3 { eT a S ; ‘a £ sts | th | , I : | i \& g So Si-« : 4 i a ' ; 4 4 * i ; f | ae ‘ ’ - Pe A cum wed | 1 bs ef Bet | | —— ~ ~~ ° _ salt is - | war RI a a eth i : \ ; ' [ee ey ens emer de x(gitd xt (4) 7 k | : xqodwoh, Yo motiadisivid—.¢ gagaT afro) fodaoe abqQ one (Wi) TT eto) 12 on ee ns le ne de a a ema (9) Thy (S) 1 REDE eee ian a | tae wan Aa Soredde "i It j ; itd ; 4 | ; i" PRN hth My A | / i { ee “4 | ; tute t ple ede ‘ : ; \ } i t yew | \ | H ; | ar | | a - ' ; | D re Pe | cad "7? | hy SE.90 8 BA wien he oa eee | rm , oon i ee haere «}s { } | | owe ; ; ; i 7 ie pwls 19.65 20h. ‘ if i +7: 3) it. 4, A BS MO ER SE tie Loe i, i @ wiht yy eee yee |e d L m/ ‘ ; ! i \ Tay Ty OLD BERING SEA- FARLY PUNUK OLD BERING SEA Fic. 24.—Chart showing chronological development of open socket harpoo! heads at Gambell. NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 217 inches (pl. 72, fig. 5). None of them differ essentially from those found at Seklowaghyaget. The whaling harpoon head shown in plate 72, figure 7, was pur- chased from an Eskimo, who had excavated it at Kukuliak, the large old site to the east of Gambell. It is somewhat longer than most of the others, and the edge opposite the spur is straight, wide, and flat as on the whaling harpoon heads from the Arctic coast of Alaska. Some of the Gambell specimens also have this kind of edge, e. g., plate 72, figure 6, but more often it is beveled. The fact that to whaling harpoon heads were found at Seklowaghy- aget and the old section of Gambell, the two sites where the least amount of excavating was done, whereas only two heads and three fragments were found at Miyowagh and Ievoghiyoq, shows con- clusively that it is only in relatively recent times that whaling has become an important occupation with the St. Lawrence Eskimos. The few specimens from Miyowagh and Ievoghiyoq (four of them bear- ing Punuk decoration and one plain) show that whaling had begun to be practiced to a slight extent in Punuk times; on the other hand, the absence not only of whaling harpoon heads, but also of whale bones, at the Hillside site, would seem to indicate that the Old Bering Sea Eskimos had not hunted the whale at all. The baleen at the Hill- side site might have been obtained from animals that had drifted ashore after having been killed by killer whales. In plate 72, figures 8 and 9, are shown two ivory objects which the local Eskimos thought had been used as walrus lance heads. Both of these specimens were purchased ; figure 9 was excavated at Kukuliak, figure 10 probably at Seklowaghyaget. We did not find any of these implements in our excavations at Gambell though I have previously obtained examples from Punuk Island and Cape Kialegak. They seem to occur only in the Punuk period and their exact function remains doubtful. They differ from harpoon heads in the absence of a central line hole, the line having passed through a narrow slot at the lower end. The open socket is very long and deep. The foreshaft was held in place by means of a lashing which passed through two slots, as in figure 9, or ran along a groove as in figure 8. The blade slit is parallel with the socket. Figure 8 is round in cross section; figure 10 is hexagonal. IMPLEMENT TYPES OF THE PUNUK PERIOD Most of the specimens described below are from the pure Punuk site, levoghiyog. Others are from the later, northwestern section of 218 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 Miyowagh, from Seklowaghyaget, and from the old section of Gam- bell ; an occasional specimen is from Punuk Island, off the southeastern end of St. Lawrence Island. Foreshafts—Two Punuk foreshafts are shown in plate 73, figures I, 2 (from Seklowaghyaget, cut 2, depth 28 in., and surface). They are generally similar to those of the Old Bering Sea period, the prin- cipal difference being that the line hole is somewhat wider, usually more triangular in outline, and is placed near the edge instead of at the center as was the case with most of the Old Bering Sea specimens. A modern St. Lawrence foreshaft with two line holes, plate 73, figure 3, is shown for comparison. In all, 34 foreshafts were found at the three later sites, 11 from levoghiyoq, 9 from Seklowaghyaget, and 12 from the old section of Gambell. Socket pieces—tIn contrast to the elaborately decorated socket pieces of the Old Bering Sea period, those of the Punuk are gener- ally plain and of simpler construction (pl. 73, figs. 6, 7, from Sek- lowaghyaget, surface, and house no. 9, respectively). The tang, instead of being conical or bifurcated, is wedge-shaped, like the modern St. Lawrence examples ; the tang may also be perforated for rivets which hold it in place in the shaft, and just above the tang there may be one or more line holes. Plate 73, figure 5, seems to be a toy socket-piece, from [evoghiyogq, cut 6, depth 22 inches. Seven other specimens of the same type come from this site. In addition there are three somewhat larger speci- mens resembling the two from Miyowagh shown on plate 33, figures 28, 29. Ice picks——Punuk ice picks are among those objects which show a clear transition from the Old Bering Sea to modern times. Old Bering Sea types I and 3 (pl. 32, figs. 1, 2, and 6) are each represented by a single example from levoghiyoq ; Old Bering Sea type 2 (pl. 32, figs. 3, 4) is well represented at [evoghiyoq and Seklowaghyaget. At both of these sites, however, the shoulder, which is characteristic of type 2, becomes more pronounced, resulting finally, at the old section of Gambell, in the type of ice pick shown in plate 73, figure 8, which with its straightly scarfed lower end and tang, approaches the modern St. Lawrence form, illustrated by figure 9; it differs from the modern type in not having a lashing slot at the tang. It should be noted also that rather flat ice picks, usually made of walrus penis bones (type 4), are characteristic of the Punuk stage (pl. 73, fig. 4, from Ievoghiyogq, cut 5, depth 24 in.). NO. iI ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 21Q The following table shows the distribution of the 34 ice picks at the three sites according to depth: Type 1 Type 2 Type 3 Type 4 Old section of Gambell: Suntaceitoul ttatsc- os. ee as it aig IP Tish) HOA Geo goon oe on : a Daten ieeto ie teOunl.. 24. r no se \ * Seklowaghyaget: Surmacertare ites -acke ss oe PEt aetna Conseco om cls - 3) es UNS OR BY lesc ooo cone AES S18 S01 ROME 4 Ales Seer ene t el S| levoghiyoq: SUMAacenOrs tease n ss eee. I Dene DalnetO! Sekiteee, ch hn cee Pe COLA ites os /ay-ns = -- Mal tamtinentOrsilitenen te cies: Hw Oo st Finger rests —The few finger rests found at the Punuk sites do not include any of the Old Bering Sea type with flat base and sloping top, but this may be accidental. Plate 73, figure 13 (levoghiyogq, cut 5, depth 18 in.), has a concave base for resting on the shaft, and a narrow slot for the lashing ; it is somewhat similar to one of the Thule types (Mathiassen, 1927, vol. 1, pl. 5, figs. 1, 2). Another of the same type comes from levoghiyoq, cut 1, depth 40 inches. Another finger rest (pl. 73, fig. 14, old section of Gambell, cut 1, depth 16 in.) is short, with a flat base, a large, round line hole, and a blunt end; it is of the same general type as one of the Old Bering Sea forms shown on plate 32, figure 8. Sealing scratcher——tThis is another implement that appears in the Punuk stage. One of these, of wood, is shown in plate 75, figure 17; it was purchased from an Eskimo who excavated it at Ievoghiyogq. It has two prongs to which the seal claws were attached, and a transverse lashing hole below. The bone object shown in plate 81, figure 14 (Ievoghiyoq, cut 6, depth 22 in.), seems also to have been a sealing scratcher, although it lacks the lashing hole; the lower end is rough- ened as if for insertion in a handle. Harpoon shafts—vThe collection also includes sections of har- poon shafts and a few wooden lance foreshafts like those previously described. Throwing board.—Plate 8o, figure 3, is an unfinished throwing board, from the midden on Punuk Island. It is 44 cm long and has a maximum width of 7.4 cm. It is somewhat heavier and wider than 220 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 the Old Bering Sea type, if we may judge from the incomplete speci- mens shown in plate 37, figures I, 2. It seems designed for the same kind of grip, however. It is, on the whole, very similar to the old Birnirk (Point Barrow) throwing board (Mason, 1928, pl. 1), even though it might not have been cut out at the center as was the latter. Spurs for the end of the throwing board (pl. 73, fig. 17, Iev., cut 5, 12 in.).—These differ from those of the Old Bering Sea period in having the upper, free end at a right angle to the tang. In addition to the specimen illustrated are two more from levoghiyoq, 12 and 22 inches deep, and one from Miyowagh, cut 19, 5 inches deep. Plate 73, figure 18 (lIev., cut I, 50 in.), is an ivory peg like the one shown in plate 33, figure 25; it has a shallow pit in the upper end; possibly for the butt end of the dart shaft. Wooden mouthpieces for seal skin floats —These forms, which were common at Miyowagh, were represented by only a single example, a specimen with stopper in place, found at Ievoghiyoq, cut 5, depth 33 inches. Plate 73, figure 16 (lev., cut 9, Ig in.) is a mouthpiece of this type, but made of ivory. Plate 73, figure 19 (lev., cut 5, 31 in.), shows the Pache type of wooden plug for closing holes in the seal skin float. It is very different from the Old Bering Sea type, being longer but of smaller diameter. Three others were found at Ievoghiyogq, at depths of 33 and 39 inches. Mouthpieces for bladder floats —The present evidence suggests that the bladder float for the light harpoon may not have been known to the Old Bering Sea Eskimos. In contrast to the large number of wooden mouthpieces for seal skin floats which were found at Mi- yowagh, there is but a single example from the older section of Mi- yowagh of the tubular ivory mouthpiece or nozzle such as is used for inflating the bladder float. This is a specimen from cut 27, depth 42 inches. It is a straight tube, with the center sunken and roughened, leaving a rounded projecting rim at either end. Since this was the only ivory mouthpiece found in the older section of Miyowagh, and since it comes from a cut which adjoins cut 19 where much early Punuk material was found, it would seem permissible to regard it as of early Punuk age. The four other tubular mouthpieces for blad- der floats found at Miyowagh are definitely of Punuk age, all of them coming from the later, northwestern section of the midden. One of these is the specimen shown on plate 73, figure 11, from cut 4, depth 20 inches. It has the “ cannon ”’ shape of the modern Alaskan examples (Nelson, pl. 56). The forward end, which was inserted in the bladder, is cut down; it is also roughened and has a slight lip INO: I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 221 for holding the lashing in place. Below the tubular body there is a constriction on both sides through which the lashing slot is cut; the base then flares outward, with the under side concave for resting on the harpoon shaft. It was hollowed by drilling from both ends. A similar nozzle not quite finished, comes from cut 20, 39 inches deep, and one which is just blocked out, from cut 2, 24 inches deep. Another, from cut 17, depth 26 inches, lacks the slotted base and has only a suggestion of the elevated rear part of the body. Plate 73, figure 10, from levoghiyoq, cut 6, depth 30 inches, shows an approach to the “ cannon ”’ shape, with the middle part sunken and the bottoms of the rims cut so as to rest evenly on the harpoon shaft. Plate 73, figure 12, from lIevoghiyoq, cut 6, depth 40 inches, has a prominent lip at the front, and a back end with flaring base through which is a drilled hole ; the wooden stopper was found in place. There are two more specimens from Ievoghiyogq, one, similar to figure 11, from cut 7, depth 18 inches, the other, unfinished, similar to figure 12, from cut 9, depth 12 inches. Float bars (pl. 73, fig. 20, Iev., cut 4, depth 53 in.).—These are all of wood; none of the ivory handles of the Old Bering Sea period (pl. 47, figs. 6, 7) were found at the Punuk sites. Plate 73, figure 15, is a heavy ivory tube, probably the mouthpiece for a water bag. It is from Seklowaghyaget, picked up on the surface. Side prongs for bird darts—From levoghiyoq there are 29 side prongs for bird darts, most of them fragmentary. Plate 74, figures 1 and 2 (from Iev., cut 4, 45 in., and cut 1, 28 in.), illustrate the prin- cipal type, which instead of being rounded like the Old Bering Sea type (pl. 33, figs. 1, 2), tends to be either flat or more or less triangular in cross-section, with one surface flat and the other arched ; there are often two small notched elevations opposite the lashing slot and two or more small round pits for decoration. No side prongs were found either at Seklowaghyaget or at the old section of Gambell though this is doubtless due to accident. One specimen was found in house no. 8. Thirteen small, slender points, apparently for the end of the bird dart, were found at Ievoghiyog, and a few others at Seklowaghyaget and the old section of Gambell; three examples are shown in plate 74, figures 3-5; they are from Ievoghiyogq, cut 5, depth 12 inches; cut 8, 29 inches, and cut I, 44 inches. Arrowheads —The arrowheads of the Punuk period are strikingly different from those of the Old Bering Sea, and in contrast to harpoon heads there is no gradation from one stage to the other. Although a few individual features are carried over, the types as a whole are 222 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 quite distinct. Arrowheads are much more common in the Punuk than in the Old Bering Sea; no less than 88 were found at Ievoghi- yoq alone, with smaller numbers from the other two sites. It is also of interest to note that almost half of the arrowheads from Ievog- hiyoq are of ivory, whereas at the two earlier sites bone was used almost exclusively. The most common type by far is that with a single barb (pl. 74, figs. 6-9). Also represented are types with two or more barbs (fig. 10) and no barbs (fig. 14). Figure 15 (lev., cut 1, depth 28 in.) is an uncommon type, of bone, with three pairs of opposed barbs and a bifurcated tang. Many of the ivory arrow- heads, particularly those with one or more barbs on the same side, are triangular in cross-section, thus approaching the modern St. Lawrence form (Nelson, pl. 61 a, figs. 1, 2). None of the above types, with the possible exception of figure 14, were found at the Old Bering Sea sites; on the other hand, some of them resemble rather closely some of the Thule types (Mathiassen, 1927, vol. I, pl. 8, figs. 2-4; pl. 9, figs. 1, 3). Some of the tangs par- ticularly resemble the Thule in that they are provided with a shoulder or conical enlargement. It should also be noted that the large collec- tion of Punuk heads included only one specimen with a blade slit, a feature which was often present on the Old Bering Sea heads. Plate 74, figure I1, is an ivory peg, from Seklowaghyaget, cut 2, depth 12 inches ; it has a sharp slender tang and a blunt, rounded head ; possibly an arrow tip. Pegs of this kind, of both bone and ivory, are characteristic of the Punuk stage. Blunt tips for bird arrows.—These were found neither at the Hill- side site nor at the older section of Miyowagh. We may therefore assume that they were unknown to the Old Bering Sea Eskimos. However, five examples were found in the later, northwestern section of Miyowagh, as follows: Cut 4, depth 43 inches; cut 6, 12 inches; cut 19, 48 inches (the first two are conical in shape, like the example from Seklowaghyaget, cut 8, depth 67 in., shown in pl. 74, fig. 13; the third is similar in shape, but longer, with a large round socket 2 cm deep, and two lateral lashing slots at the lower end). An un- finished specimen similar to the latter comes from cut 20, depth 12 inches. Apparently the straight-based, conical form is older than the ovoid. The fifth specimen from Miyowagh (cut 2, depth 43 in.) is of another type, with a bifurcated tang instead of a round socket and with an enlarged end into which were cut three deep slanting grooves; it is of bone (cf. Nelson, pl. 61, b, fig. 17; ¢, figs. 3-7). Plate 74, figure 13, from Seklowaghyaget, cut 8, depth 67 inches, was the only example of the conical type of blunt arrow tip found at NOs I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 223 the three later sites; the others, like plate 74, figure 12 (Sek., cut 8, 60 in.) are constricted at the lower end, producing the ovoid shape which still prevails on St. Lawrence Island (Nelson, pl. 61, c, fig. 2). Four other specimens of this type come from Seklowaghyaget, depth 44 inches, the old section of Gambell, depth 24 inches (2), and house no. 10. In keeping with the large number of arrowheads from the Punuk sites are two other implements, bone bow braces and sinew twisters, which were found in considerable numbers and which afford definite evidence of the use of the sinew-backed bow during this period. Bow braces.—These are of two kinds: flat, short, narrow strips of bone which were placed at the center of the bow beneath the sinew reinforcement on the outer side; and similar strips, but curved, for strengthening the bow at its most vulnerable point, that is, at the recurved ends. Plate 74, figure 16 (Iev., cut 6, depth 5 in.), is one of the flat braces for the center; others range from 7 to 13.5 cm in length. Figures 17 and 18 are two braces for the backward curving ends ; figure 17 (old section of Gambell, cut 1, 8 in.), the more com- mon type, has a shallow longitudinal groove on the concave side and was placed on the outside, in the bend, beneath the sinew cable; figure 18 (lev., cut 8, 44 in.), with a rather deep groove down its convex side, was fastened to the inner side. All three of the above types are found on modern Alaskan bows. In all 21 bow braces, including the three illustrated, were found at Gambell, as follows: From the Hillside site and Miyowagh, none; from Ievoghiyogq, 13, from the following depths: 5 inches (2), 9 inches, 12 inches, 24 inches (3), 28 inches, 33 inches, 36 inches, 44 inches, unknown (2) ; from Seklowaghyaget, 3, as follows: 18 inches, 19 inches, 22 inches; from the old section of Gambell, 5, as follows: 8 inches (3), 10 inches, 32 inches. Sinew twisters—Plate 74, figures 19, 20, are sinew twisters— slender bone rods with opposite curved ends, used for tightening the sinew cables on the backs of the bows (Murdoch, 1892, pp. 315- 316). They are somewhat thicker and more curved than the modern examples. Sinew twisters were found only at Miyowagh and Ievoghiyog. From Miyowagh there were four, as follows: Cut 4, 25 inches, of bone; 61 inches, of ivory; cut 17, 28 inches, of wood; cut 27, 29 inches, of ivory. From Jevoghiyoq there are I0 examples, all of bone, as follows: 5 inches, g inches, 12 inches (2), 18 inches (pl. 74, fig. 20), 25 inches (2), 28 inches, 31 inches (pl. 74, fig. 19), 42 inches. 224 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 Wrist guards——Another object connected with the bow that ap- peared during the Punuk stage is the wrist guard. Four of these, all of ivory, are shown in plate 74, figures 21-24. Figure 23, from Ievoghiyoq (purchased), is straight across the bottom, rounded at the top, and has a narrow lashing slot on either side. The ornamenta- tion is Punuk, consisting of pairs of lines, some straight, some curving, with shorter pairs of lines attached, and light spurs placed at sparing intervals. Figure 22 is also from Ievoghiyoq, and was purchased; it has a straight base in which there are five slots, and rising from this, two pointed wings; it has a very simple decoration of deeply incised lines. Figure 21, a surface find from the old section of Gambell, is similar in shape but has the upper part undivided. The decoration is neatly applied and consists of bands of horizontal, vertical, and oblique lines, with spurs at regular intervals. Figure 24, purchased from the Eskimos at Savunga, illustrates the modern St. Lawrence wrist guard, which is practically identical in shape with the modern Chukchee type (Bogoras, 1904-09, vol. 1, fig. 73). In its essential features it is also similar to the older St. Lawrence form, plate 74, figure 23. Little can be said regarding relative ages of the different forms, as several of them appear to have been more or less contemporaneous. On the basis of the present rather meager material, however, it would seem that plate 74, figure 23, represents the older type. Six wrist guards have already been described in the sections on art. One is the fragmentary specimen from Miyowagh, cut 27, depth 9 inches, shown on plate 22, figure 10. The others are from Ievoghiyogq, cut I, depths 36, 44, and 36 inches, respectively (pl. 65, figs. 9-11) ; Seklowaghyaget, depth 20 inches (pl. 67, fig. 7) ; and old section of Gambell, depth 8 inches (pl. 67, fig. 5). In plate 75, figures 20-22, are shown three objects which to a certain extent resemble wrist guards, but which apparently were used in some other way, perhaps being worn as ornaments. Figures 21 and 22, of ivory, come from cut 1, depth 12 inches, and cut 17, depth 12 inches, in the later section of Miyowagh. One edge is straight with round holes or small rectangular slots near the ends; the ends are curved and incline inward, and between them, at the center, is a pronged projection. Figure 20, of baleen, was found at Ievoghiyoq, cut I, depth 40 inches. It is bilobed with a small round hole and two notches at the edge of each lobe; opposite are three large round scal- lops, leaving two narrow projections at the end and two similar but notched projections at the center. Plate armor.—This is another important element introduced during the Punuk period. It is interesting to note that on St. Lawrence Island NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 225 it was made entirely of bone, whereas at Bering Strait both bone and ivory were used (Hough, 1895, pp. 632-3). Twenty-four of the bone plates were found at levoghiyoq, nineteen at Seklowaghyaget, and eight at the old section of Gambell; none were found at the Hillside site or Miyowagh. The greater number of them were long narrow strips, like plate 76, figures 22, 23, with three pairs of holes, either round or rectangular, for the thongs which held them together. From levoghiyoq there are several plates which differ in size from the usual type, being both wider and longer (pl. 76, figs. 20, 21). Figure 24 was probably one of the pieces of a helmet such as used by the Chuk- chee, Koryak, and Gilyak (Bogoras, 1904-09, vol. 1, figs. 84, 89, and go). The 24 pieces of plate armor at Ievoghiyoq were found for the most part in the upper levels of the midden, as follows: 8 inches; 12 inches (4); 16 inches (2) ; 20 inches (5) ; 22 inches (2) ; 24 inches fe; 20 inches (2));) 30 imches’ (2) ; 37 inches; 42 inches. The 19 pieces of armor from Seklowaghyaget occurred at deeper levels, as follows : 20 inches ; 22 inches ; 30 inches (2) ; 34 inches (3) ; 38 inches (11) ; 60 inches. Those from the old section of Gambell were all found at depths of 24 and 32 inches. Fish spears—The three small ivory points (pl. 75, figs. 1-3) are probably prongs for the end of a light fish spear such as Nelson describes from western Alaska (Nelson, 1899, pp. 194-5; pl. 67, figs. Hee?) Barbs for salmon spears.—Plate 75, figure 19, is an ivory barb for a salmon spear from the midden at Cape Kialegak, southeastern end of St. Lawrence Island. Several similar specimens were found on Punuk Island, but not at any of the Gambell sites. Their absence there may be accidental, but on the other hand, the salmon spear seems never to have been used very extensively on St. Lawrence Island. The above specimen is of the modern Alaskan type (Nelson, fig. 42, 1) which is practically identical with the Thule type (Mathias- pemeno27, vol. i, pl. 12, figs, 10, 11; pl. 43, figs. 1, 3,4, 55 pl. 71, figs. 2, 3; pl. 77, fig. 13). Compound fishhook.—This form makes its first definite appearance in the Punuk stage. Plate 75, figures 4, 5 (old section of Gambell, 10 and 24 in. deep) are the bone shanks of two such fishhooks, used no doubt for catching tom cod. The lower end is enlarged, with three or four deep vertical grooves in which the prongs were set ; the upper end has a perforation for the line. From Ievoghiyogq there is a simi- lar specimen of ivory and six others of wood; from Seklowaghyaget there are two specimens, one bone, one ivory. The prongs for these 226 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 hooks are of ivory, as a rule slightly curved, with the tang beveled on the inner side so as to fit into the groove, and with the upper end either barbed or plain (pl. 75, figs. 6-9, from Iev., depth 12 in., and old section of Gambell, 8, 24, and 32 in. deep). This type of fishhook has continued in use on St. Lawrence Island down to modern times (Nelson, 1899, pl. 68, fig. 20; pl. 60, fig. 28). The fishhook in one piece, plate 75, figure 10 (old section of Gambell, depth ro in.), is later and seems to represent a development from the compound type. The prongs, four in number, project upward from the base in the same way as the detachable prongs of the earlier type. This specimen, from the old section of Gambell, is in general similar to some of the modern fishhooks of the Siberian Eskimo (Nel- son, 1899, pl. 69, fig. 2) and of the Chukchee (Bogoras, 1904-09, fig. 67, €). Fishline sinkers——The fishline sinker undergoes an interesting de- velopment during the Punuk period. Twenty-six ivory sinkers were found at Ievoghiyoq, about half of them with the two line holes parallel and the other half with the line holes at right angles. The first type, illustrated by plate 75, figure 12, is practically identical with one of the Old Bering Sea types, but it is somewhat less sym- metrical and in most cases is less carefully made. The second type, figure 13, with the line holes at right angles, also resembles one of the Old Bering Sea types in that the lower end is thicker than the upper. Here again, however, there is a difference in technique; the Punuk examples are of poorer workmanship and lack particularly the studied asymmetry of the earlier forms which was produced by the prominent side bulge below the center. At Seklowaghyaget only five fishline sinkers were found. Three of these are like the types from Ievoghiyog, but the other two, figures 14 and 15, are modern types. Figure 14 has two small slots near the lower end to which lines, with hooks attached, were tied. Sinkers of this type from St. Lawrence Island, but usually more crudely made, are described by Nelson (pl. 69, figs. 28, 31, 33). The other sinker, figure 15, has seven cylindrical oblique holes in the sides in which barbs were set, thus being a fish- hook and sinker in one; there are also perforations in which lines, with hooks at the end, were fastened. A modern fishhook-sinker of this type from Cape Nome is described by Nelson (pl. 69, fig. 4). Net sinkers—Heavy ivory net sinkers are also characteristic of the Punuk stage (pl. 75, fig. 16, from Iev., cut 5, depth 28 in.). These are sections of tusks, with only the ends worked and perforated; the present specimen had been used originally as a pick. These net sinkers are larger and heavier than those described by Nelson (p. 189). Three NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 227 examples were found at Miyowagh, from cut 1, depth 42 inches; cut 4, 15 inches; and cut 8, 12 inches. From Ievoghiyoq there are two more, one from cut 2, depth 52 inches, the other depth unknown. The only other sinker of this type comes from house no. 8. Stone sinkers.—Plate 80, figure 4, is a stone sinker for a tom cod net, found at the old section of Gambell, depth 8 inches. Just above the center is an encircling groove, joined by two vertical grooves from the lower end. Tom cod nets of baleen with grooved sinkers at the bottom are described from the Chukchee and Siberian Eskimo (Bo- goras, 1904-09, vol. I, p. 149, fig. 65; Nelson, 1899, p. 187, pl. 70, fig. 12). They are also used on St. Lawrence Island at the present time. Five other sinkers similar to this were found at the old section of Gambell and one at Seklowaghyaget. The tom cod net was known earlier, however, for in the later section of Miyowagh we found the bottom of one of the baleen nets and several small sections of netting, but none of the sinkers. Meat hooks——Only two meat hooks were found at the later sites. One of these, from Ievoghiyoq, was of the type shown on plate 35, figure 8; the other, from Seklowaghyaget, depth 61 inches, is shown on plate 75, figure 18. It is of bone, with a concave base and a slender pointed prong; there is no slot, the lashing passed around the top, which was roughened to prevent it from slipping. Wound plugs —Wedge-shaped wooden wound plugs were also rare. Plate 75, figure 11 (lev., cut 1, depth 31 in.) illustrates one of the three which were found, all at levoghiyoq; the other two come from cut I, depth 50 inches, and cut 2, 26 inches. They are smaller and more pointed than the earlier examples. Ice creepers——Only one ice creeper was found at the three later sites, the specimen shown on plate 75, figure 24, from Seklowaghy- aget, purchased. However, two of those from the later section of Miyowagh may also be regarded as of Punuk age. One of these, plate 75, figure 23 (cut 19, 24 in.), is of ivory with the three spikes carved out in relief ; it is convex on the opposite side. Plate 75, figure 25, of bone, from cut 19, depth 5 inches, approaches the modern ice creeper in shape (Nelson, p. 216), although it has the three widely spaced ivory pegs characteristic of the Old Bering Sea type. Plate 75, figure 26 is a modern ice creeper from St. Lawrence, shown for comparison. Bird bolas —The bird bola is another important element that makes its appearance in the Punuk stage. No bola weights were found at the Hillside site. At Miyowagh, 37 were found, all of them confined to the upper levels of the later, northwestern section, as follows: Cut 228 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 I, 12 inches (6) ; 38 inches. Cut 2, 12 inches; 20 inches. Cut 4, 15 inches (2); 31 inches (2); 43 inches. Cut 17, 20 inches (5); 26 inches (2); 27 inches; 28 inches (4). Cut 20, 12 inches (2); 22 inches (6) ; 26 inches (2). Cut 19, 17 inches, 32 inches. At Ievoghi- yoq, there were 167, found at all levels; at Seklowaghyaget, 38; and at the old section of Gambell, 14. Thus, in all there are 256 bola weights from the four sites. They may be divided into three general types: (1) Walrus teeth, unworked except for the beveled end— usually the upper end—where the line hole is cut through (pl. 76, figs. I-4, 9). (2) Pieces of ivory or bone—rarely the latter—of ir- regular shape; usually unworked except for the line hole but some- times with sides and lower end slightly worked (pl. 76, figs. 5-7). (3) Well made, with lower end conical, rounded or flattened (pl. 76, figs. 8, 10-19). The distribution of the three types is as follows: Miyowagh Tevoghiyoq Seklowaghyaget Cee Dy pe Petes 19 61 18 9 INN Abaodes< 2 37 I 3 Typesi eee’ 16 68 19 2 Plate 76, figure 19, it should be noted, has the line holes sunk into the top, which is the prevailing method north and east of Bering Strait. It is probably an importation, as the type is represented in the present extensive collection by only three specimens, all from Seklowaghy- aget, at depths of 8, 12, and 18 inches. The bola weights of the Thule culture resemble in general those of north Alaska, being small and rounded or squarish in outline, in contrast to the prevailing St. Law- rence types, which are longer and tapering at the upper end. Sledge runners.—These underwent considerable modification dur- ing the Punuk period. In order to bring this out more clearly, we will consider the several finds separately. From Ievoghiyoq there are 25 complete and fragmentary specimens, all of ivory. Two of these are of the Old Bering Sea type 1a, and one of the Old Bering Sea type 2. The remaining 22 are quite different. They fall into three rather distinct but related groups, which developed not from any of the Old Bering Sea types of sledge runners but on the contrary from the Old Bering Sea type of shoe illustrated in plate 45, figures 4-9. (See also table 3, p. 230.) The first of these, which will be designated as Punuk type 1, is represented by three examples, two of which are shown in plate 77, figures 1 and 2. All of these are ends, so that we do not know the original length; the average width at the holes is 4.2 cm and the NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 229 thickness—or height—2.9 cm. A short distance from the end is a transverse ridge, which was left when the top surface was cut down, and beneath this ridge passes the lashing slot—two holes drilled obliquely and meeting at the center. Type 1a (pl. 77, fig. 3), of which there are five examples, is flat on the upper surface, like type 1 and seems to have been cf the same general shape. The lashing arrangement, however, consists of two holes, either at the tip or a short distance back, drilled completely through and connected on the under side by a sunken groove to pro- tect the lashing. This groove, connecting the two holes, is visible in plate 77, figure 3. Four of this type were also found at Miyowagh in the later, northwestern section of the midden. Type 2 (pl. 77, figs. 4, 5), 14 examples, differs from type Ia in having a V-shaped concavity extending the entire length of the upper surface. Figures 4 and 5 show the upper and lower sides, respectively, of this type of runner; the sides are straight. The lashing slots either extend completely through, as in type 1a, or are drilled in obliquely, beneath a transverse ridge as in type 1. The average width, at center, is 5 cm; thickness—or height—3 cm. The above three types, whether they be considered runners or shoes, are clearly derivatives of the flatter, Old Bering Sea shoes illustrated in plate 45, figures 4-9. Thus the Punuk type 1, with its transverse ridge beneath which the lash- ing holes are drilled, has its prototype in the Old Bering Sea shoe types 2 and 3 (pl. 45, figs. 6-9), which though flatter and wider, also had the lashing holes passing under a knoblike elevation ; whereas the Punuk types 1a and 2, which lack this ridge, are descended from the Old Bering Sea shoe (type 1, pl. 45, figs. 4, 5), which was flatter but had a similar arrangement of the lashing holes. In view of their height, which increases still further in the late Punuk and modern specimens described below, it seems better to call them runners, or shoe runners. From the next site, Seklowaghyaget, we have 11 sledge runners and fragments. Old Bering Sea type 1 a (which was illustrated on pl. 45, figs. 1, 2) is represented by two specimens, which ditfer however in having two drilled holes, instead of a rectangular cut-out hole, beneath the notch on the top side. Punuk type I a is represented by one specimen and type 2 by two. One of the type 2 specimens, however, differs from the previous ex- amples (such as pl. 77, figs. 4, 5) in being much narrower and some- what higher, with the sides inclining downward to the bottom. The remaining six runners from Seklowaghyaget belong to what will be called type 2a (pl. 77, fig. 6), a further specialization of type 16 230 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 2, but with the lashing arrangement of type 1 at the front end; to- ward the back end there are several additional oblique holes drilled through from the top to the sides. The upper surface has the usual V-shaped concavity, but the sides are made differently, first sloping sharply inward from the rim, and then descending yertically to the bottom, making it Y-shaped in cross-section, like the modern St. Lawrence examples. Plate 77, figure 6, is 36.8 cm long, 2.9 to 3.5 cm wide, and 3.4 cm high; the three fragmentary specimens average 3.8 cm in width and 4 cm in height. From the last site, the old section of Gambell, there are five sledge runners. One of them is an example of: Old Bering Sea type 2. Another, shown in plate 77, figure 7, is like the preceding Punuk type 2a, but still higher and narrower, with a height of 4.7 cm and a width of 2.7 cm across the flaring top and 1.3 cm across the constricted bottom. Of the remaining three, two are too fragmentary for classi- fication although their shape is that of type 2a. The other, shown in plate 77, figure 8, possesses two features which set it somewhat apart, so that we may call it type 2b. The shape is that of 2a, but the lashing arrangement at the end consists of two oblique holes drilled through the flaring edge; in addition, there were two long transverse holes, rectangular or oval in shape, like those on the old Bering Sea types I and Ia. The distribution of the various types of sledge runners and shoes at the five sites is given in table 3. Summarizing the distribu- TABLE 3. Old Bering Sea Old Bering Sea Punuk sledge runners sledge shoes sledge shoe-runners Type | Type | Type | Type | Type | Type | Type | Type | Type | Type | Type 1 la 2 1 2 3 1 la 2 2a 2 Old section of Gambell..... ie I se Ee: I I Seklowaghyaget B at I 2 6 levoghiyoq....] .. 2 I 5 3 S|) ae! Alt Miyowagh.....] 2 6 | uy Ay | aig) 4° Hillside site....1 3 Be 3 oe $. 8 All four of these are from the later, northwestern section of the midden. tion shown therein, we see that the heavy sledge runners made from entire walrus tusks are of Old Bering Sea age, being found only at the Hillside site and Miyowagh; but that in a modified form (type 1a) it continues into the Punuk stage. Type 2, on the other hand, may have continued in use until modern times, although NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 231 it was in large part superseded by later types, which as was pointed out above, developed during the Punuk period from the Old Bering Sea sledge shoes. The transition from shoe to runner is brought about mainly by an increase in height. Thus, the Punuk shoe-runners of types I and I a are only modified forms of the Old Bering Sea shoes with higher sides, and—in type I—a more prominent transverse ridge over the lashing hole. The next type, 2, differs in having the upper side con- cave, and in the later types, 2a and 2b, this V-shaped concavity be- comes more pronounced, developing into flaring rims as the runner becomes higher and narrower and the form in cross-section changes from a straight-sided M to a wide-bottomed Y. Sledge cross bars, made from walrus penis bones, were also found at the Punuk sites. Bone and ivory knife handles—We find that the compound form, which was the most common type of the Old Bering Sea period (pl. 38, figs. 5-7) is also well represented in the Punuk finds. Twenty-six handles of this type were found at the three Punuk sites; two of these are shown on plate 78, figures 4, 5. In two respects, however, they differ from the Old Bering Sea forms: first, bone (usually a piece of walrus rib) has largely replaced ivory as the material, 23 being of bone, 3 of ivory; second, the face, instead of being straight. as in the great majority of the Old Bering Sea handles, was sharply beveled at the end, so that in order to complete the handle only a short section, about 3 cm long, had to be lashed on. In addition to the compound handle there were a few specimens with deep, wide sockets, somewhat like the Old Bering Sea type shown in plate 38, figure 8, and a few others with narrow slot for an end blade, like plate 38, figures 9 and Io. Of more importance is the fact that in the Punuk we have the first appearance of the “ crooked knife’, with a small side blade near the tip, a type which in its later and more developed form is now wide- spread among the Eskimo, particularly in Alaska. Fifteen of these were found at the three Punuk sites, all of them made from walrus ribs. Most of them are short and only slightly curved, like the two shown in plate 78, figures 1 and 2. These are rather far removed from the later and more specialized Alaskan forms which are longer, more sharply curved and flattened, and which often have the iron blades riveted to the outside; these Punuk handles are, in fact, much closer to some of the Thule types (Mathiassen, 1927, vol. 1, pl. 18, figs. 2, 4; pl. 68, fig. 4; pl. 83, fig. 10). Plate 78, figure 3, is an example of the more curved Punuk type. Most of the handles have no hole 232 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 for a suspension cord. The 15 knife handles of this type were found as follows: Ievoghiyoq, depth 8 inches (pl. 78, fig. 3), 12 inches, 16 inches, 33 inches, 39 inches, 44 inches (2). Seklowaghyaget, 4 inches (pl. 78, fig. 1), 18 inches, 20 inches, 38 inches. Old section of Gambell, 8 inches (pl. 78, fig. 2), 10 inches, 24 inches (2). Wooden knife handles—There are 15 wooden knife handles, all but one of them from Ievoghiyog. They all have slits for end blades, like one of the three Old Bering Sea types. Plate 78, figures 6 and 8 (lev., depths 33 and 39 in.), are in all essential respects the same as the Old Bering Sea handle shown in plate 38, figure 1. Plate 78, figure 7, however, from Ievoghiyoq, depth 37 inches, represents a specialization of this basic type and apparently one which is confined to the Punuk stage. The end is so narrowed that the lower part of the blade would have projected at the sides; this being the case, there must have been two drilled holes at about the center of the blade for the lashing which was fastened to the handle around the narrow grooved end. Three others of this type were found at depths of 24, 44, and 45 inches, at Ievoghiyoq. Eight others similar to plate 78, figures 6 and 8, were found at Ievoghiyoq, at depths of 30 inches (3), 36 inches, 37 inches, 39 inches, 42 inches, and 60 inches. There 1s also a single example of this type from Seklowaghyaget, depth 58 inches. Knife sharpeners——These, made of young walrus tusks, also occur in the Punuk finds. None of them are carved to represent a whale like the Old Bering Sea specimens shown in plate 38, figures 12, 13. Stone implements.—Implements of rubbed slate were as abundant at the Punuk sites as at Miyowagh and the Hillside site, and the types were generally the same. Adzlike scrapers, like plate 42, figures 12-14, also occurred, but infrequently. With regard to chipped stone implements, conditions were entirely different. Only a very few im- plements of chipped stone—slate or any other kind—were found at the three later sites, although these were abundant at the Hillside site and fairly common at Miyowagh. The small slate implements with rubbed edges (pl. 39, figs. 19-22) and the columnar rubbing stones (pl. 43, figs. 1-5) were also absent at the Punuk sites. Ulu handles.—There are 17 wood and 4 ivory ulu handles from the Punuk sites. They show less variation than the Old Bering Sea forms, most of them being rather straight and thick like the one with part of the slate blade still in place, from Seklowaghyaget, depth 48 inches, shown on plate 78, figure 10; a few are crescent shaped. Plate 78, figures 9, 11, 12, and 13 are of ivory. Figure 9 (lev., depth 5 in.) is only 4.9 cm long and has four dots for decoration. Figure 12 has NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 233 the characteristic late Punuk ornamentation of pairs of deeply cut horizontal and vertical lines with occasional gouged oblique spurs ; it is from levoghiyoq, purchased. Figure 13 is decorated with two rows of nucleated circles enclosed by a narrow bordering band of two lines with inward pointing alternating spurs attached. The circles in the upper row are separated by pairs of short vertical lines, those in the lower row have large Y-shaped figures attached; the circles are all compass-made. This is an example of disintegrated Punuk art ; all continuity of design has been lost and decorative effect is accom- plished only through the repeated use of detached elements. It was excavated at Seklowaghyaget, and purchased from an Eskimo. Figure II is a late Punuk type, also purchased, and probably from Seklow- aghyaget; it is pointed at one end, squared off at the other with a large opening which originally held an ornamental link or pendant like that shown in plate 82, figure 30. Fat scrapers ——Two fragments from levoghiyog (depths 12 and 42 in.) indicate that the straight-sided scraper for removing fat from intestines and skins, like plate 13, figure 8, was also used during the Punuk period. However, another though related form, was more common (pl. 78, fig. 14, Iev., depth 26 in.). It was made of bone and was somewhat curved, with the inner edge rounded for holding in the hand and the outer edge sharpened for scraping. There are three other specimens like this, two from Ievoghiyoq, depths 12 and 19 inches, and one from house no. g. Figure 15 is a broken scraper made from a caribou or reindeer cannon bone, from levoghiyog, depth 12 inches. The modern St. Lawrence scraper, which is of ivory, oval in outline with rounded base ( Nelson, 1899, pl. 50, fig. 10), was not found at the Punuk sites, but this was probably due to accident. At any rate we have a prototype for it in such scrapers as those shown in plate 51, figures 9, IT. Spoons.—The spoons of the Punuk period are of bone and have long handles. One of these, from Ievoghiyvog, depth 12 inches, is shown in plate 78, figure 16. It has a shallow rounded bowl and is constricted at the neck. Three similar specimens from Ievoghiyog come from depths of 5, 12, and 36 inches. Awls.—Bone and ivory awls are abundant, just as they were at the earlier sites. Needles—The absence of needles is undoubtedly due to accident. Adz heads—Certain developmental tendencies may be observed when comparing Punuk adz heads with those of the Old Bering Sea period, but the picture is somewhat obscured by the fact that several types were in use at the same time. First, we must note the absence 234 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 on the Punuk specimens of any “ears” for the lashing thongs, such as are present on the two earlier adz heads shown in plate 46, figure 3, and plate 60, figure 2. Most of the Punuk adz heads of ivory (pl. 78, figs. 17, 18, from Ievoghiyoq, depth 50 in., and Seklowaghy- aget, depth 12 in.) are also more slender in outline and have a prominent projection or knob at the upper end. The “ shoe-shaped ” ivory adz head—figure 20—is from Punuk Island, 4 miles off the east end of St. Lawrence Island; it is shown lying on its side; the socket for the handle is a narrow rectangular pit, above which is a slot through which the lashing passed. The knobbed projection ex- tends forward, in the plane of the socket, giving it the appearance of a shoe heel, and the lower part, to which the blade was fastened, has an oblique downward slope. This type of adz head represents a late Punuk specialization of the straighter type, figures 17, 18. That it was not found at Gambell is probably due to accident, for figure 21, from house no. 8, is clearly a derivative form, one which is still fur- ther specialized in the increased height of the upper part where the handle is attached. In this case there is no socket for the handle, but a scarfed surface with two round holes for the lashing thongs ; there is no knob at the upper end. Six more adz heads of the same general type as the two shown on plate 78, figures 17, 18, were found, all at levoghiyogq, at the fol- lowing depths: 22 inches (2), 28 inches, 36 inches, 42 inches, 48 inches. None of the Punuk adz heads had a transverse hole for the handle, like the earlier one shown on plate 46, figure 6. The type with large socket in the end for a stone blade (pl. 46, figs. 6, 7) seems not to have been carried over into the Punuk, as the later type of end socket (pl. 78, fig. 19, Sek., depth 36 in.) is very narrow, for an iron blade. Three other adz heads of this type were found at Seklowaghyaget, depth 28 inches, and houses no. 9 and 10; they are all of bone, with the upper end tapering, and with a flattened surface instead of a pit for the attachment of the handle. Adz handles.—Plate 79, figure 3, is a well-made wooden adz handle from house no. 7, at levoghiyoq. The lower end is flattened and some- what spatulate, with a deep groove and shoulder for the index finger. The upper part is narrower, 3.5 cm high and 2 cm wide; at the end is a small rectangular-oval projection which was inserted in the socket of the adz head, and a narrow slot for the lashing which held the head and handle together. Another wooden adz handle comes from Sek- lowaghyaget, depth 34 inches. Pick handles—Two pick handles are shown in plate 79, figures I and 2. The first, of wood, is from Seklowaghyaget, depth 52 inches; NO. ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 235 it is 19 cm long, with a smooth-rounded grip and a slight groove for the index finger; the pick, which was probably of ivory, rested in a groove at the front end and was held on by thongs passing through a large triangular notch. Figure 2, a surface find from Seklowaghy- aget, is of ivory, and has the same arrangement for the attachment of the pick; the grip, however, is provided with three broad notches for the fingers. A similar specimen comes from Miyowagh, cut 20, depth 39 inches. Picks and wedges.—Ivory picks and wedges were plentiful and dif- fer in no way from those of the Old Bering Sea period, already described. Mattock blade.—Plate 8o, figure 2, is a mattock blade, 38 cm long, made from a whale rib. It is from levoghiyoq, depth 34 inches. At the upper end is a hole for the handle, and somewhat below the middle two notches for the thongs which helped to hold the handle taut. Snow shovels—Snow shovels made of walrus scapulae, like those of the Old Bering Sea period, were found at all of the Punuk sites. Bone knife—The bone knife, or dagger, is another characteristic Punuk element which seems to have been lacking in the Old Bering Sea period. Thirty-four were found, twenty-four of bone, ten of ivory. Three of them are illustrated in plate 79, figures 4-6. The pre- vailing type at levoghiyoq and the northwestern section of Miyowagh is illustrated by the bone knife, figure 5. The average length is a little more than 20 cm, and the width around 3 cm; one side is flat, the other is arched and has a distinct median ridge; the end is pointed and the edges fairly sharp; there is usually a perforation at the end of the handle for a suspension cord, and occasionally a narrow lateral slot through the upper edge of one blade. Seven examples were found at Miyowagh, as follows: Cut 1, depth 42 inches; cut 3, 54 inches; cut 9, 25 inches; cut 12, 20 inches; cut 19, 51 inches; cut 22, 18 inches; and from cache in the floor of house no. 3 (pl. 79, fig. 5). Eight simi- lar specimens come from Ievoghiyoq, at the following depths: 5 inches; 12 inches (3); 19 inches; 25 inches; 31 inches; 42 inches. From Seklowaghyaget there are three, depth 12 inches, 25 inches, and depth unknown. From the old section of Gambell there are four, depths 16, 24, and 32 inches (2). There is also a single specimen from house no. 8. Plate 79, figure 6, of ivory, levoghiyoq, depth 39 inches, along with three others, differs from the prevailing type in having a median ridge on both sides; it is the only one that is decorated. Plate 79, figure 4, of ivory, represents what seems to be a later type found only at Seklowaghyaget and the old section of Gambell. Like 236 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL, 96 figure 5, the lower side is usually flat and the upper side convex with a median ridge; occasionally there is a median ridge on both sides. These knives, however, are smaller, averaging 16 cm or less in length and 2 cm in width; the handle is short and perforated at the end, and there is usually a small narrow slot through the edge of the blade just below the handle. Knives of this smaller type were found as follows: Seklowaghyaget, depths 12 inches (pl. 79, fig. 4), 18 inches, 20 inches, 26 inches, 28 inches (2), 60 inches; old section of Gambell, depths 24 inches (2), and 32 inches. Bone knives are no longer used on St. Lawrence Island, so there must be some doubt as to the exact function of the present specimens. It is possible that they were used for scraping snow from the fur clothing, but the pointed end and straight, sharp edges make it appear more probable that they were daggers used in hunting or in war. Similar knives are known from the Thule culture (Mathiassen, 1927, vol. 1, p. 48, pl. 13, fig. 15). The smaller type (pl. 79, fig. 4), is more doubtful, as it could hardly have been very effective as a weapon. Mouthpieces —Mouthpieces for bow drills are somewhat more com- mon in the Punuk than in the Old Bering Sea finds. Almost all of them are small, like the three of ivory shown in plate 82, figures 37- 39, the average length being between 4 and 5 cm. Figure 37, from Miyowagh, cut 20, depth 12 inches, has the drill hole on the convex end; the upper end was cut down, leaving a ledge which was held between the teeth; through this ledge is a round hole, probably for a suspension cord. Figure 38, from Miyowagh, cut 6, 12 inches, is similar except that the ledge extends completely across. In figure 39 (lev., cut 5, depth 33 in.) the hole for the end of the drill shaft is on the straight, under side, the convex upper part being held in the mouth; there is a groove on either side to allow a firm grip with the teeth, and at the center of this a narrow slot for a suspension cord. Figure 40 is the largest mouthpiece found, 7.3 cm long and 2.1 cm thick. It comes from Mirrukta, an old site 3 miles east of Gambell. The drill shaft was held in a large circular hole on the under side 2 cm in diameter and 1.8 cm deep. Two pointed wings extend upward and between them is the mouth grip, an oval projection with a concave face and a suspension slot through one edge. Drill rests —Plate 8o, figure 5, is a drill rest made from a whale’s ear bone, from levoghiyoq, cut 6, depth 22 inches. It has three deep pits on either side. Ivory drill rests like those from Miyowagh and the Hillside site were also common at the Punuk sites. The following objects connected with the bow drill were found in large numbers at the Punuk sites, but differ in no significant respect INO] I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 237 from those already described: shafts of wood, some painted red ; bows of bone and wood; points of bone. On plate 84, figure 1, is shown a wooden fire hearth from house no. 7, at levoghiyoq; it is a large, flat piece of wood with numerous charred pits made by the shaft of the fire drill. Hand drills and reamers—Hand drills and reamers of bone and ivory were equally abundant, but likewise similar to those of the Old Bering Sea period. Plate 81, figure 21, is one of these, from Mi- yowagh, cut 19, depth 24 inches, made from the distal end of a dog humerus. The point is unusually fine, being 1.5 mm in diameter, with a sharp edge beveled from both sides. Engraving tools.—Plate 81, figures 17-20, are four engraving tools similar to those from the later section of Miyowagh shown in plate 60, figures 10, 11. The first three are from Ievoghiyoq, depths 22, 12, and 31 inches; figure 20 is from the old section of Gambell, depth 16 inches. Figure 1g is made of wood, the others of ivory. Another wooden specimen comes from levoghiyoq, cut 1, depth 60 inches. The later Punuk type of engraving tool is of simpler construction than those of the early Punuk shown in plate 60; those from Ievoghiyoq all have a buttonlike head, whereas figure 20 has no enlargement at all. On the last-named specimen a part of the iron point still adheres to the lower end. Rubbing tools and stone fakers.—Rubbing tools are as abundant in the Punuk as in the Old Bering Sea, and stone flakers of bone are also found. Plate 79, figure 7, is an ivory object of unknown use, from Mi- yowagh, cut 2, depth 28 inches. It is 22.3 cm long, pronged at one end, rounded at the other; the under side is flat, the upper somewhat con- vex ; it gives the impression of being unfinished. Another pronged ivory object of unknown use, from the old section of Gambell, depth 8 inches, is shown on plate 79, figure g. It has some- what the appearance of a “ sealing scratcher ”, even to the slot below the prongs; but the small size, the decoration, and the wide lashing groove around the lower end are features which are not found on any of the sealing scratchers known at present. Bone hook.—Plate 79, figure 8, is a bone hook peculiar to the Punuk culture. It is made from a dog humerus which is perforated near the proximal end for the insertion of a sharp sliver of bone. The exact function of these hooks is uncertain ; but the fact that they are invari- ably made of dog humeri which have a natural foramen at the distal end suggests that the latter was utilized as a suspension hole. In that case it could have been attached to a line, or more likely used inside 238 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 the house as a wall hook. The present specimen is from Seklowaghy- aget, depth 3 inches. ‘Two others come from the same site at depths of 28 and 64 inches. There are also three from Ievoghiyoq, two from depths of 26 inches and one 24 inches. Harpoon rest.—Plate 79, figure 10, from Seklowaghyaget, depth 34 inches, is a harpoon rest for the umiak, made from two pieces of walrus ribs. According to Nelson such rests were “ lashed to the bows of umiaks, just inside and between the front ends of the rails; in these the ends of the lances and spears rest, and through them the lines run out”? (p. 226). The present specimen, which was the only one found, is much cruder than those of ivory which Nelson illustrates from the mainland (pl. 78, figs. 33, 37). Boat hooks.—Plate 79, figures 12 and 13, are two large boat hooks of ivory, from levoghiyoq, depth 12 inches and depth unknown (pur- chased). They both have a slightly concave base, a body constricted at the middle, and a straight projecting prong. Figure 12, which is badly weathered, had the upper surface covered with a typical Punuk ornamentation consisting of small compass-made, nucleated circles, from which stream single and double lines. Figure 13 has a median ridge along the top, and a backward-projecting end. Plate 84, figure 2, is a well-preserved wooden paddle probably for a kayak, found in house no. 7 at levoghiyog. The blade—tip broken— is 9 cm wide and appears to have been about 39 cm long. It may have been a double paddle, for it was made in two sections, lashed together by means of a simple scarfed joint, with wooden pegs at the ends and wrappings of some kind to afford additional strength. Pottery.—Pottery was no less abundant in the Punuk than in the Old Bering Sea period. Just as at the two earlier sites, potsherds were found in great numbers at Ievoghiyoq, and the ware differed in only two respects from that previously described: only a relatively small number of sherds had the roughened or corrugated exterior which was so common during the earlier period, and none were grass-tempered. The shapes of the vessels apparently remained unchanged. The lamps were shallow, circular, and thick-walled, with conical bases and no wick ledges. Plate 84, figure 3, from levoghiyog, illustrates the type. Plate 53, b, is a complete lamp of this kind, excavated by the writer from an old grave at Kowieruk, Seward Peninsula. The cooking pots at Ievoghiyoq, to judge from the fragments found were round, steep-sided, thin-walled, with rather heavy rounded bases. These same types were also found at Seklowaghyaget, but along with them were fragments of the pottery which is so characteristic of St. Lawrence Island today (Hough, 1898, pl. 15, fig. 16; Nelson, NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 239 1899, pl. 28, figs. 1, 4, 5, 7-10). At the most recent site, the old section of Gambell, this modern type of pottery was found exclusively (pl. 84, figs. 4, 5). ! Baleen vessels——Baleen vessels with wooden bottoms are as char- acteristic of the Punuk as of the Old Bering Sea period. Wooden vessels—Long, narrow wooden vessels also continued to be used. Plate 80, figure 1, is one of these, a food platter from levoghi- yoq, cut I, depth 41 inches. It is 44 cm long, 6.2 cm deep, and 13 and 6.4 cm wide at the ends. Drum.—The drum occupied a much more prominent place in the Punuk than in the Old Bering Sea culture. Five wooden and seven ivory handles were obtained from the three Punuk sites, in addition to a considerable number of rim fragments. Five of the handles are illustrated in plate 81, figures 2-6; the first three are from Seklow- aghyaget, purchased ; figure 6 is from house no. 7, levoghiyoq. Figure 2 was originally about 17.8 cm long, with a deep square notch for a thick rim and a narrow lashing slot below it. At the front end is an oval projection bearing a Punuk ornamentation of spurred lines radiat- ing from rim to center. Originally there was a perforation at the lower end. Figure 3 is very similar, but only 9.3 cm long; the flange- like projection at the upper end is proportionately larger and slightly concave. Figure 4 is provided with four small notches as finger grips, and a rather large hole at the lower end where a link ornament was probably attached. Figure 5 lacks the flange at the end; it bears a late Punuk ornamentation consisting of two bands of four and five lines connected by a pair of straight longitudinal lines having two very oblique spurs at the center. The wooden drum handle, figure 6, lacks ornamentation of any kind. The drum handles not illustrated come from levoghiyoq, depths 12 inches (2), 26 inches, 30 inches, 31 inches ; and Seklowaghyaget, depths 12 inches and 60 inches. Objects of wood and baleen—These are not nearly as abundant at the three later sites as at Miyowagh or the Hillside site. This does not mean that these materials were less used but that the conditions for their preservation were less favorable at the three later sites. There was much more gravel in the levoghiyoq and Seklowaghyaget middens (especially the latter) than at the sections of Miyowagh where the greatest amount of digging was done, and as was pointed out previously, cultural material is always less abundant at such places. This is not only because the gravel itself is often an intrusive fill, and therefore more or less sterile, but also because cultural material, when included is not likely to be permanently preserved due to the fact that the gravel thaws rapidly and to a considerable depth. At 240 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 the sections of levoghiyoq where conditions were comparavle to those at Miyowagh, baleen and wood were plentiful. In addition to those previously mentioned, the following classes of objects were found, none of which, so far as could be determined, differed essentially from those described from the Old Bering Sea sites : pieces of wooden arrow shafts, small wooden cylinders, painted red (like pl. 57, figs. 18, 19), wooden objects with scoop end (like plate 47, fig. 11), pail handles, drill handles, drying racks, toy boats of wood and bark, wooden shafts of various kinds, painted red; baleen ice scoop, toy bows, knots, toboggan cross pieces, plaited baleen objects like plate 56, figures 8, 9. Miscellaneous objects of wory and bone.—Plate 81, figure 1, is a bone tube, from Ievoghiyog, depth 24 inches, made from a swan’s wing bone with both ends cut off. There are several others from the same site made from smaller bones. They may have been drinking tubes, although such an interpretation would be conjectural. Plate 81, figure 12 (lIev., depth 12 in.) is the canine tooth of a polar bear, with the upper end perforated for suspension. It may have been used as a charm against sickness, a custom still in vogue on St. Lawrence Island. Plate 81, figure 13 (from Sek., purchased) is an ivory object with a smooth flat top 8.9 cm long and 1.9 cm wide, and two cylindrical “legs”? 1.3 cm in diameter. The flat upper surface is crudely orna- mented with a crooked longitudinal line and a series of irregularly spaced transverse lines. The only objects comparable to this that I have seen are some of bone excavated by Dr. Hrdlicka on Kodiak Island. Another characteristic Punuk element is a small bird bone inserted in another. Two of these, from Ievoghiyog, depth 12 inches, are shown in plate 81, figures 15, 16. There are four others, all from Iev- oghiyoq at depths of 5, 12, 18 and 30 inches. Their use, if any, is problematical ; none of them show any signs of having been worked, the ends of both bones being invariably broken off. Such bones, but with one or both ends sharpened, are known from the Thule culture and are regarded by Mathiassen as gull hooks and bodkins ( Mathias- sen, 1927, vol. i,.pl. 435 fiss10; “pl. 52; fig. 10) Plate 82, figures 19 and 20, are two ivory objects, from Ievoghiyogq, depths 22 and 30 inches, the function of which is unknown. They have a short neck or shaft, through which is a transverse perforation with deep grooves leading down to the end. The head is enlarged, and somewhat conical; that of figure 20 has been carved to represent a seal with prominent eyes made of round wooden plugs set in deep pits, the pupils represented by small ivory insets at the center of the plugs. NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 241 A similar object has been described from Point Hope (Mathiassen, mgsod, pl. 13, fig. 9). Plate 82, figure 21, is an ivory plug from the old section of Gambell, depth 20 inches. It is probably an ornament of some kind, and is similar to one shown in plate 60, figure 7. Plate 82, figure 23, is an ivory “ thimble guard ” from house no. 8. It has the shape of a double crescent, like one described by Nelson (pl. 44, fig. 24, p. 110) ; similar forms are also known from the Thule culture (Mathiassen, 1927, vol. I, pl. 52, fig. 17; pl. 63, fig. 6: Mathias- sen, 1930 b, pl. 14, fig. 7). Plate 82, figure 24, from levoghiyogq, depth 24 inches, may have had a similar function, although this is rather un- certain. It is a thin ivory bar 7.4 cm long, perforated at the middle. Plate 82, figures 25 and 26, from Seklowaghyaget (purchased) and levoghiyoq, depth unknown, respectively, seem to be handles for either drill cords or drag lines, although I do not know that any of this exact form have been described elsewhere. A distinguishing feature of the Punuk culture in its later phase is the use of ornamental links of ivory. These are found on ulu and drum handles, combs, and other implements, as a part of the same piece of ivory from which the object itself is carved. An example is shown on plate 82, figure 30 (Sek., purchased), which is an ivory ulu handle, of the same type as shown in plate 78, figure 11. The slit for the blade is between 2.5 and 3 mm wide. Plate 82, figure 31, also from Seklowaghyaget (purchased), is a larger link of the same kind, and no doubt used in a similar fashion. Plate 82, figures 32 and 33, are two smaller links, the latter carved to represent a walrus head ; they are from Cape Kialegak and Punuk Island. Figure 34 is a rather crudely carved link pendant with part of another link still remaining ; it comes from the old section of Gambell, depth 8 inches. Plate 82, figure 29 (from Sek., purchased) is an ivory object of unknown use, with a double link at either end. In the base of the object is a narrow lozenge-shaped socket, which tapers upward and emerges as a narrow rectangular slot on the upper side. The collection contains four other link ornaments, one from Ievoghiyoq, depth 16 inches, the others from Seklowaghyaget, two purchased and one from 12 inches deep. Plate 82, figures 27 and 28, are two objects with the same general contour and the same kind of socket as figure 29, but lacking the links at the ends. In a general way these objects resemble the guards at the ends of sword hilts from the Scandinavian iron age (Nerman, 1929, p. 65 ff.), although there is nothing else to suggest that they were used in this way. 242 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 Small flat square pieces of bone, like the one illustrated in plate 58, figure 15, were also found at the Punuk sites. From Ievoghiyog there is a bone ring, made by cutting out the center of the epiphysis of a whale vertebra. According to the Eskimos, it is a rim to the ventilation hole in the front wall of the inner sleeping room. Plate 82, figure 35, from the old section of Gambell, depth 24 inches, is an ivory trace buckle for a dog harness. It is of the usual modern Alaskan form, with one hole through the lower, somewhat square end, and another, at right angles to it, through the smaller, notched end (Nelson, 1899, pl. 76, figs. 3-5). This is the only example of a trace buckle from any of the old sites at Gambell. Furthermore, the other elements which are invariably associated with dog traction in its known form—swivels, ferrules for the whip handle, and flat bone sled shoes— are likewise missing. Not one of these objects was found at the four older sites which all together yielded over 7,000 specimens. On the other hand, a considerable number of such objects are included in the much smaller quantity of material excavated from recent house ruins and midden deposits of comparable age at Gambell, Punuk Island, and Cape Kialegak. The conclusion is unmistakable that dog traction, insofar as its presence is indicated by the above classes of objects, was unknown to both the Old Bering Sea and Punuk Eskimos, and that it was introduced on St. Lawrence Island only in very recent times, probably not until near the beginning of the nineteenth century. The small ivory object shown in plate 82, figure 36 (lev., depth 24 in.), appears to be a harpoon or dart rest for a kayak (cf. Nelson, p. 226, pl. 78, fig. 2). The base is slightly concave and down the center is a longitudinal groove 4 mm wide which opens through a small slot at the back end and through a similar slot at the junction of the curving prong and the base. By means of this slotted groove it was attached to one of the cross lines which extended over the deck of the kayak ' Objects of personal adornment.—These are considerably more com- mon in the Punuk than in the Old Bering Sea culture, although the categories are still somewhat limited. Ivory pendants are found in considerable numbers and in a wide variety of forms. These are illustrated in plate 82, figures 1-13. Some are long, and either straight-sided or thickened at the end (figs. 1-4) ; some are squarish with a rounded end (fig. 5); some are conical (figs. 6-9) ; and some deeply grooved (fig. 10). Others are suspended from the middle, and are bilobed (fig. 11), four-lobed (fig. 12), or knobbed at the ends (fig. 13). Provenience of the illustrated specimens is as follows: Plate 82, figure 1, Miyowagh, cut 20, depth 12 inches; NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 243 figure 2, levoghiyoq, depth 12 inches; figure 3, Miyowagh, cut 1, 12 inches; figure 4, Ievoghiyoq, 28 inches; figure 5, Ievoghiyoq, 22 inches; figure 6, Punuk Island; figure 7, Punuk Island; figure 8, Seklowaghyaget, purchased ; figure 9, Miyowagh, cut 20, 12 inches; figure 10, unknown, purchased at Gambell; figure 11, Punuk Island; figure 12, Ievoghiyoq, 24 inches; figure 13, unknown, purchased at Gambell. In addition to these there are two more from Ievoghiyogq, depths 22 and 60 inches. Plate 82, figures 14 and 15, are two dog tooth pendants perforated at the upper end. They were excavated from the old section of Gam- bell, depth 32 inches, along with several others. Since none were found at the four older sites, they must be regarded as rather recent on St. Lawrence Island. It should also be noted that perforated bear teeth, which were very scare at the Old Bering Sea sites, were fairly common later. Ivory brow bands and buttons also occur at the Punuk sites. Plate 82, figure 16, is an ivory ear ornament inset with small lumps of iron pyrites, held in place by some black sticky substance, now hardened. Of the original eight insets, four remain. At the upper end is a small slot for the suspension cord. This ear ornament comes from the midden on Punuk Island. Plate 82, figure 17, also from the Punuk midden, is an unfinished ear ornament of the same type. The view is of the back side and shows the high ridge from which the hook would have been carved. These two specimens are of particular interest because they represent a type which is common in southwest Alaska, especially on Nunivak Island, but which has not been reported previously from north of the Yukon. Otto W. Geist has also excavated an ear ornament of this kind at Kukuliak, on the north side of St. Lawrence Island (Bunnell, 1934, p. 5). Plate 82, figures 22 and 18 (from Seklowaghyaget, 22 in., and the old section of Gambell, 16 in., respectively), are probably also ear ornaments, of a different type. ‘ Combs are rather prominent in the Punuk finds, being represented by 13 examples, all of them of ivory (pl. 81, figs. 7-11). There is considerable variation in shape, some with wide handles, like figure 8 others with smaller, thinner handles, like figure 10. The teeth are usually about 2.3 cm long. Figure 8 has unusually short teeth and a crude ornamentation on the handle consisting of straight lines and long alternating spurs. Figure 10 has an ornamental link attached to the handle. The combs were found as follows: Miyowagh, cut 6, depth 20 inches. Ievoghiyoq, depth 12 inches (2) (pl. 81, fig. 10) ; 22 inches (pl. 81, fig. 9) ; 28 inches; 30 inches ; 34 inches; 36 inches, 244 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 purchased (pl. 81, fig. 8). Seklowaghyaget, depth 12 inches, picked up on surface; purchased (pl. 81, fig. 11). Old section of Gambell, 12 inches (pl. 81, fig. 7) ; 32 inches. Plate 79, figure I1, is a pair of crude snow goggles from Seklow- aghyaget, made from a section of walrus penis bone. Toys.—F rom the Punuk sites we have a number of toys, mostly of ivory, but there are also wooden dolls, bows and kayaks, and (from Tevoghiyoq) the small bark figures, single or double, representing whales—all similar to the Old Bering Sea types shown in plate 59, hguresi2) ©; 751314 waO S17 20 27. Other Punuk toys of ivory are shown in plate 83. We have here the first appearance of the small ivory bird figures (pl. 83, figs. 7-11) which today are so common on St. Lawrence Island, in northeastern Siberia, and at Bering Strait. They are used by the children in play- ing a kind of dice game just as among the central Eskimo (Boas, 1888 a, pp. 567-8). They are also very common in the Thule culture (Mathiassen, 1927, vol. 2, pp. 117-118). At Gambell they appear to be restricted to the two latest sites, none having been found at Ievoghi- yoq or Miyowagh; the two carved figures of birds found at the Hill- side site had evidently a different function. The great numbers of small dogs, bears, foxes, and human figures carved out of ivory, which today are used as toys by the children on St. Lawrence Island, may be somewhat later than the bird figures, as none were excavated at Gambell ; however, this might be because of the relatively small amount of work done at the two latest sites. Plate 83, figures 8, 9, and 11 come from Seklowaghyaget, depth 12 inches; the other two were pur- chased from Eskimos and are said to have been excavated at Seklow- aghyaget. A sixth specimen comes from house no. 10. The three ivory whale figures, plate 83, figures 16-18, are probably not toys but objects of some religious significance. The first is from Tevoghiyoq, depth 8 inches, the other two from houses nos. 9 and 8, respectively. Figures 16 and 17 are carved realistically, but in figure 18 only the tail is represented naturally; it also has a long hollow on the under side; both figures 17 and 18 have a slot just below the tail. There are also several whale figures of wood, carved rather crudely and larger than these. Two toy sledges and a sledge runner are shown in plate 83, figures 1-3. The runner, from Ievoghiyoq, depth 31 inches, is of the form described above as Punuk type 2. The sledges, both of which were purchased from the Eskimos, have the sides and runners carved in one piece ; otherwise they are identical with one from St. Lawrence Island figured by Nelson (fig. 132). Figure 2 is 14.7 cm long; it has round NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 245 perforations at several places, and rows of small pits, for decoration, on the sides. The three ivory kayaks, plate 83, figures 4-6, present several features of interest. In the first, from Seklowaghyaget (purchased), the framework at the bow is open and visible, an arrangement which, if it represents the actual conditions, is unique among the Eskimo. Figure 5, from Ievoghiyoq, depth 31 inches, is 6 cm long and shows the rider sitting in the manhole. Back of him is a double seal skin float (cf. the single float on the specimen figured by Nelson—fig. 134). Although the figure of the man is very schematic, it may be that the projection at the head represents a hunting helmet. The base of the kayak is slotted, indicating that it was attached to something as an ornament, rather than used as a child’s toy. Figure 6 is from the old section of Gambell, depth 16 inches. It has two pairs of sealskin floats, and a small pit at the center in which the figure of the rider may have been set. Plate 83, figure 15, is an ivory top from Seklowaghyaget, depth 36 inches; it is oval, with a flat base and low convex upper side. Two ivory weights for spindle buzzes like plate 46, figure 10, were found at Ievoghiyoq, depth 22 inches, and Seklowaghyaget, depth 48 inches. Plate 83, figure 12, is an ivory doll, and figures 13 and 14 are the heads of two similar dolls. The larger specimen, 19.7 cm long, is armless and legless, with the lower part of the body brought to a blunt point. The features of the face consist of narrow slits for the eyes and mouth; it is from Cape Kialegak. Figure 14 is from Punuk Island and has the features more clearly indicated. Three parallel lines be- neath the eyes represent tattooing and the hair is shown knotted at the top of the head. Figure 13, from Seklowaghyaget, purchased, is round-faced ; the eyes and mouth are represented by small pits and the nose is carved in low relief. Textile—The piece of textile found at a depth of 26 inches in cut 4, northwestern section of Miyowagh, was submitted for examination to Dr. Gene Weltfish, of Columbia University, who makes the fol- lowing comment : The textile fragment taken from an Eskimo kitchen-midden on St. Lawrence Island, Alaska, is clearly not intrusive on two counts: On the one hand I know of no modern Eskimo work which is exactly of this kind, though most of it is similar; therefore, while technically congruent with modern Alaskan Eskimo style it is not identical. On the other hand it is not intrusive Russian since it is sufficiently like the native twined basketry and bag work of the whole north west coast of the continent to be indigenous. The technique of the textile is plain twined, with stitches trending upward to the right, on a double twisted warp. The material appears to be a grass 17 246 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 used for both warp and weft and the texture was semi-stiff very much like that of bags of the Klamath-Modoc of northeast California and Oregon, made of tule rush. It is like the Klamath-Modoc twining also in the fact that the warp is double-twisted, a characteristic not found either in the rest of modern work along the coast or in modern Eskimo twining. There are a few horizontal rows of decorative stitching of an elementary kind, involving the introduction of one dark-colored weft thread instead of the natural. It is interesting to notice by way of comparison that O. T. Mason, “ Aboriginal American Basketry”, pp. 397-8, comments on Asiatic twined bags. While the Chukchi example he mentions is like the Bering Strait modern Eskimo work, the wallet from Kamchatka resembles your specimen according to his comment in that it has a coarse hemp cord warp (which would undoubtedly refer to the double-twisted type of warp). SKELETAL REMAINS Although careful search was made for early burials on the top and along the slopes of the mountain, no traces of these were found. There are large numbers of recent burials among the rocks on the mountain slope, and it was here for the most part that Dr. Riley D. Moore ob- tained his extensive collection of skulls and skeletons in 1912. How- ever, neither at Gambell nor elsewhere on St. Lawrence Island have I ever found burials which from the accompanying grave offerings could be identified as of Punuk or Old Bering Sea age. If, as seems probable, the prehistoric St. Lawrence Eskimos followed the usual practice of simply placing the body among the rocks, either on a flat surface or in a natural crevice, it is easy to understand why no skeletal remains of any antiquity have been found, since in this method of burial the bones very soon become scattered and broken. Fortunately, a sufficient number of crania from houses and middens of Punuk age, both at Gambell and the eastern end of the island, have been collected to indicate with some degree of certainty the physical type that was associated with the Punuk culture. The measurements will not be given here as they are to be included in the next issue of Dr. Hrd- licka’s Catalog of Crania. However, the essential point for the present is that in general they conform to the modern St. Lawrence type, which is mesocephalic. Unfortunately, the excavations at the oldest sites have yielded very little skeletal material, only two incomplete skeletons and a detached skull found at Miyowagh. Of these, the skeleton found in cut ga, back of house no. 3, had a skull that was mesocephalic. The other two skulls, found in the Old Bering Sea house at Miyowagh (house no. 4), were extremely dolichocephalic, falling in this respect beyond NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 247 the range of normal variation for modern St. Lawrence crania. Though it would not be proper to draw any far-reaching conclusions on the basis of these two skulls alone, they represent nevertheless the most ancient Eskimo skeletal remains that have yet been found, and the fact that they are markedly long-headed is something to be con- sidered by any theory which seeks to trace the origin and spread of the Eskimo on the basis of physical characters. ANIMAL REMAINS FROM THE FIVE SITES Animal bones were found in such abundance in the middéns and house ruins that it was impossible to preserve all of them. Conse- quently, only representative samples from each cut were collected. The order of occurrence of the more easily recognizable forms, such as whale, walrus, seal and dog bones, were observed at the time and samples preserved, but the smaller and more questionable mammal bones were collected as found. Dr. Remington Kellogg, assistant curator of mammals, United States National Museum, has kindly examined and identified the mammal bones as follows: From the Hillside site: Ringed seal (Phoca hispida) ; harbor or hair seal (Phoca richardu pribilofensis) ; bearded seal (Erignathus barbatus nauticus) ; walrus (Odobenus divergens) ; dog (Canis fam- iliaris) ; polar bear (Thalarctos maritimus) ; fox (Vulpes alascensis ) ; beluga (Delphinopterus leucas) ; and whale (represented by baleen). Of the above the seal, walrus, and dog bones are greatly in the majority, the others being relatively scarce. At Miyowagh all of the above forms are represented, and in the same relative proportions, but in addition there are bones of the bow- head whale (Balaena mysticetus) ; the humpback whale (Megaptera novae-angliae ) ; the common finback whale (Balaenoptera physalus ) ; the ground squirrel (Citellus lyratus), and the field mouse (Microtus innuitus innuitus), both probably intrusive; the varying hare (Lepus americanus dalli), rare; a few reindeer or caribou bones and pieces of antler; and the horn of a mountain sheep (Ovis, sp). The hare, caribou, and mountain sheep are not indigenous on St. Lawrence Island and hence were no doubt brought over from Siberia as meat used for food on the voyage. From levoghiyoq we have the dog and the same species of marine mammals, in abundance, in addition to the fox. The absence of other forms is probably due to accidental selection. The Seklowaghyaget midden contained more gravel and less refuse than the other middens, and as a result of this condition and the rela- 248 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 tively small amount of digging done there, not many mammal bones were collected. Only the following species are recorded: Walrus, ringed seal, harbor seal, dog and whale. The old section of Gambell and the more recent house ruins (no. 8, g, and 10) yielded walrus, seal, dog, whale, fox, and polar bear bones. From the most recent house ruin (no. 10) came the cervical vertebra of a horse. . Comparison of the mammal remains from the various sites shows that the Old Bering Sea and Punuk Eskimos hunted the same species as do the present day Eskimos. A possible exception is the right whale, for the bones of this mammal were not found at the Hillside site, al- though baleen was present. The abundance of walrus bones in all of the middens and the great number of implements made of ivory show that the walrus has always held a position of prime importance in the economy of the St. Lawrence Eskimos. Considering that the walrus was the principal source of food, of oil for heating, cooking, and il- lumination, and of the skins used for lines, thongs, and boat cover- ings, it might seem that this animal, if any, would have been the object of ceremonial observances designed to propitiate the spirits of the dead animals in order to insure good hunting in the future. We have, of course, no means of knowing definitely whether or not there were ceremonial practices in connection with the walrus in prehistoric times. It seems significant, however, that walrus skulls were found in large numbers at all of the old sites, whereas seal skulls were exceedingly rare, evidently having been disposed of in some particular manner— no doubt thrown into the sea—in accordance with a ceremonial cus- tom still observed by the St. Lawrence and other Eskimos. It is evident from the dried seal muzzles tied together with baleen thongs and the bearded seal skull with suspension cord of baleen found in cuts 16 and 6 at Miyowagh that ceremonial practices, much like those of today, were observed by the Old Bering Sea Eskimos in connection with the disposal of seal remains. On the whole, it would seem that just as today, the walrus, in spite of its economic importance, was accorded scant respect as compared with the seal. Seal bones, however, were found in large numbers at all of the sites, showing that as a food animal it was second only to the walrus. We found no evidence—such as burned or cracked scapulae—to indicate that scapulimancy had ever been practiced. In connection with the dog bones, which were also abundant, it is of interest to observe that Gerrit S. Miller, Jr. and Dr. Kellogg think that the skulls from the Hillside site and the southeastern section of Miyowagh represent a different breed from those found at the NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 249 Punuk and later sites. The older skulls show a remarkable uniformity, being somewhat smaller than the later ones and possessing a charac- teristic beveling or deep longitudinal depression along the frontal sinus. The skulls from Ievoghiyoq (also those from Punuk Island and Cape Kialegak), in addition to being larger, are more variable, with a relatively smooth frontal region more like that of the modern St. Lawrence dogs. It should also be noted that many of the dog skulls, particularly those from the Old Bering Sea levels, had a large opening in the parietal region, evidently made for the removal of the brain, which must have been eaten. The several thousand bird bones from the Gambell sites have been carefully studied by Dr. Herbert Friedmann, curator of birds, United States National Museum, who has published a separate report upon them (Friedmann, 1934). The collection was found to include the bones of 32 species, two of which, the red-legged kittiwake (Rissa brevirostris) and Kittlitz’s murrelet (Brachyrhamphus brevirostris), represented additions to the known avifauna of the island. As to the relative abundance of the various species, Dr. Friedmann reports as follows: The species most abundantly represented in the collection is Pallas’s murre. It is obviously the most important single bird species to the Eskimo, and it is obvious from the enormous number of bones, that the species was as abundant in the past as it is in the present. The other birds commonly used for food include the crested and the paroquet auklets, the Pacific and king eiders, and, strangely enough, the pelagic cor- morant. One of the surprises was the paucity of goose bones, especially of the emperor goose. Pigeon guillemot, old-squaw, long-tailed jaeger, red-faced cormorant, and short-tailed albatross come next in descending order of frequency, and after them come a large number of species, present in varying quantities. [Friedmann, 1934, p. 84.] Table 4, compiled from Dr. Friedmann’s paper, shows the occur- rence of the different species according to site. The relative abundance of the various species, as shown on the table, may be taken as a fair index of the Eskimos’ selection for food purposes of the birds most desired—and most readily obtained—from among those frequenting the western end of the island. It is of interest in this connection to observe that the bird bone collection from Cape Kialegak, at the op- posite end of the island, though not as large as that obtained at Gam- bell, included the following species which were not found at Gambell : the whistling swan, cackling goose, black brant, American pintail, spectacled eider, white-winged scoter, surf scoter, American scoter, merganser, red-breasted merganser, pomarine jaeger. Furthermore, the fact that the emperor and the white fronted geese, the old squaw, 250 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 the Steller, Pacific, and King eiders, and the paroquet auklet were all more abundantly represented at Kialegak than at Gambell would indicate that these species also were more common at the eastern than at the western end of the island. It should be recognized that the accidental selection involved in collecting the bones might be responsible in some degree for the con- ditions shown in table 4; and that many of the gaps would certainly be filled if a still larger collection of bones were available. This would apply particularly to Seklowaghyaget, where material of all kinds was rather scarce and where relatively little excavating was done. How- ever, there is no reason to believe that the relative occurrence of the various forms would be materially affected. One of the most striking facts to appear from the table is the large number of species and the quantity of bones from Ievoghiyog, indi- cating that birds were a more important item of food during the Punuk period than at any other time. It will also be observed from table 4 that there was always a strong preference for sea birds. This is shown by the fact that the murre, the tufted puffin, the pelagic cormorant, the long-tailed jaeger, and even such a strong flying bird as the alba- tross were all obtained in greater numbers than either the loons or the geese. The waders, the small land birds, the raven, and the owl are among the local forms which seem not to have been utilized at all for food. Samples of fish bones collected from the five sites were all identified by E. D. Reid, of the division of fishes, United States National Museum, as those of the cod, genus Gadus. Barnacle shells collected from all of the sites, except Seklowaghy- aget, were identified by Dr. H. A. Pilsbry, of the Philadelphia Academy of Sciences, as belonging to the species Coronula diadema (L.), the whale barnacle, and Balanus balanus (L.), a north Atlantic form which is also found in Bering Sea and which commonly grows on molluscan shells. Harald A. Rehder, assistant curator, division of mollusks, United States National Museum, has identified the molluscan remains as fol- lows: Thais saxicola Val., theWest Coast rock purple ; V olutopstis sp. ; Mytilus edulis L., the edible mussel; Pecten islandicus Mull., the northern scallop; Serripes laperousei Desh., one of the cockles; Macoma middendorffi Dall; Spisula polynyma alaskana Dall, the “Alaskan” surf clam; and Savicava pholadis L., one of the rock- borers. Most of the forms were represented at all of the sites, except Seklowaghyaget, but in view of the relatively small size of the sample, the absence of a particular species is of no significance. NOS T ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 251 TABLE 4 Species Hillside Miyowagh|levoghiyoq| Seklowaghyaget Pieection Mellow-billed loon... ..:...| =... x x eee eee PTAC OOM. ¢s.0 2 2 ye ees es oer viens x ae >< Red-throated loon........ sheet: x x x No se Short-tailed albatross..... x< eat: Daag Long-tailed jaeger........ x xX X xX X xX X Glaucous gull.. rs x x x Ae Glaucous- winged gull 5 eee Rais x< x x Short-billed gull.......... rey sy SMITHSONIAN MISCELLANEOUS COLLECTIONS PLATE 30 Hillside site, Old Bering Sea culture VOL. 96 Drill shaft, wood, below floor stones, house no. 1, U.S.N.M. no. 352,664. Drill mouthpiece, ivory, house no. 1, U.S.N.M. no. 352,616. “ “ Ivory tube, second level of midden, U.S.N.M. no. 352,740. Fat scraper, ivory, house no. I, “ “ “ “ ’ “ “ “ , “ , third level of midden, U.S.N. “ “ 352,602. 352,600. “ “ , walrus tooth, among rocks, U.S.N.M. no. 352,868. M. no. 352,745. Ivory object, below floor stones, house no. 2, U.S.N.M. no. 352,545. Stone flaker, pone: gene nae, U.S.N.M. no. 352,587. “ “ 1) “ “ “ 352,890. ; ivory. third level of midden, U.S.N.M. no. 352,766. Scraper of dog femur, first (upper) level of midden, U.S.N.M. no. 352,716. “ “ “ “ “ “ “ “ , fourth (lowest) level of “ , house no. 1, U.S.N.M. no. 352,639. “ Reamer, ivory, house no. 1, U.S.N.M. no. 352,837. Bone awl, among rocks, U.S.N.M. no. 352,578. Ivory awl, house no. 1, Ivory object, pe “ “ “ “ “ “ ’ ’ , house no. I, Toy kayak, woud Ses floor stones of house no. 2, U.S “ , house no. 2, U.S.N.M. no. 352,507. Toy umiak, ’ “ “ “ “ “ 352,834. 352,480. , between floor stones of house no. 1, U.S.N.M. no. 352,660. « “ « I, ; arene foes) U.S.N.M. no. 352,589. = 352,070. “352,839. “ 352,782. « "352:660 a, .N.M. no. 352,552. Piece of brow band, ivory, third level of midden, U.S.N.M. no. 352,746. Ivory object, fourth (lowest) level of midden, U.S.N.M. no. 352,781. «“ “ “ “ , house no. 1, U.S.N.M. no. 352,640. , third level of midden, U.S.N.M. no. 352,757. PLATE 31 Miyowagh, Old Bering Sea culture Harpoon foreshaft, ivory, cut 9, U.S.N.M. no. 354,400. “ “ “ “ , 9, 354,400 a. Broken harpoon foreshate, ivory, cut 27, depth 26 in., U.S.N.M. no. 371,043. End of harpoon “ ee ahnS. Sa gaan & “ 369,728. “ “ “ “ 3 “ 4, “ 40 in., “ “ 353, 383. “ “ “ “ ; “ 19, “ 51 in “ “ 371,915. Harpoon SOLES CO ivory, ‘ext 7, depth 75 in., U.S. N. M. no. 353,778. “« : tom 9, 24 in. « « 354,069. “ “ “ 24, 10a ines “ “ 370, les End of harpoon foreshalt: ivory, cut 7, depth 51 in., U.S.N.M. no. 353,735. Harpoon socket piece, ivory, cut 16, depth 26 in., U. S.N.M..no. 354,229. End of harpoon socket piece, ivory, cut 27, depth 52 in., U.S.N.M. no. 371,161. “ “ “ “ “ “ “ “ “ “ Tang of “ “ “ ’ , “ 19, 79 in., 27S e eO nines 24k ES SS Abie “ “ 370,275. 370,989. 371,993. Ce a ee Se Saisie aoe a ON rag eames oO A reise oN oN NH Oo RoR Tea e A I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 403 Harpoon socket piece, ivory, cut 17, depth 27 in., U.S.N.M. no. 369,638. Broken harpoon socket piece, ivory, cut 9, depth 24 in., U.S.N.M. no. 354,446. Harpoon socket piece, ivory, cut 13, depth 14 in., Tang of harpoon socket piece, ivory, cut 23, depth 25 in., PU AnE Es 2 Miyowagh, Old Bering Sea culture U.S.N.M. no. 353,938. U.S.N.M. no. 370,567. Epoontc ice eres iv Ory, EuE 21, eens 56 in., U.S.N.M. no. 370,464. ; 3 he 80 in., : CS Bisehpellisn “ “ “ ; “ 5 “ 275 “ 22 nes “ “ 371,023. “ “ “ : “ ; “ DB “ 64 in., “ “ 370,640. « « “ ‘ bone, “ I, “ 63 in., “ « 353,064. “« “ “ : ivory, “ gy. “ 54 ilo. “ “ 353,970. Finger rest for harpoon shaft, ivory, cut 22, depth 46 in., U.S.N.M. no. 370,525. “ “ “ “ “ “ j “ I 8, “ 32 in 7 “ “ 369,732. “ “ “ “ “ “ “ a “ 62 in. “ “ 353 759. “ “ “ «“ “ F “ ; “ 2 oh “ 38 in na “ “ 37 I ,096. “ “ “ “ “ . “ : “ We “ By, in., “ «“ 353,688. “ “ “ “ “ “ “ Bi, “ I 6 in SF “ “ afi ,009. Float mouthpiece, wood, cut ie depen 44 in., U.S.N.M. no. 369,844. “ “ “ “ 18, “ 48 in. “ “ 369 822. “ «“ «“ ; « 25, « 55 in., «“ « 370,978. SNe ping, i iv Ory; seit 9, depth 48 in., U.S.N.M. no. 354,089. ? oun 5 Tone. € “ 353,082. “ “ «“ «“ 22 “ 41 in “ “ 370 Rin. « « : a 25, « Teens: «“ « 370,837. Float bar, ee Meee Meer DSTA : esean3: Stopper for float mouthpiece, wood, cut 18, depth 41 in., “ “ «“ “ “ “ 19, “ 56 in., ’ ’ “ PLATE 33 Miyowagh, Old Bering Sea and Punuk-cultures Side prong for bird dart, ivory, cut 23, depth 8 in., “ “ “ “ “ ; “ , “ 2 7 ; “ 29 i n. . “ “ “ “ “ “ “ : “ F “ ie “ 75 in oe “ “ “ “ “ “ “ L “ F “ I 8, “ 66 in 4 “ “ “ “ “ “ “ ‘ “ 5 “ 20, “ I 2 in. ; “ “ “ “ “ “ “ ‘ “ ‘ “ 2 : “ 2 4 in Zs “ “ “ “ “ “ “ , “ ; “ 2). “ 24 in on if “ re e Sabie “', “ , purchased, U.S.N.M. no. 354,466. . Sere Vek | CULO: Bente 24 in., U.S.N.M. no. End point for bird dart, “ , “ 19, 34 in., = * “ “ “ “ “ ; « A “ 9, “ 24 in., “ “ “ «“ “ “ “ : “ : “ 20, “ 30 in ce “ “ “ “ “ “ “ : “ , “ I 8 “ I 8 in * “ “ “ “ “ «“ “ ; “ , “ 9, “ 24 in., “ “ “ “ “ “ “ ; “ 2 “ I 9, “ 38 in “y “ “ U.S.N.M. no. U.S.N.M. no. 369,823. 270 Os: 370,536. 371,069. 353,783. 369,914. 370,315 (Punuk). 353,099 “3 353,100 i. 370,002. 370,040. 354,402. 370,367. 369,687. 354,402. 370,058. 404 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 16. Point for fish spear, ivory, cut 19, depth 12 in., U.S.N.M. no. 369,961. 17. « « «“ ies rae 19, “« 51 in., « « 370,135. 18. « « « «“ . « : “ it «“ 30 in., « «“ 352,957. 19. « « « na Be 20, «“ 22 in., « «“ 370,341. 20. “ “ «“ « : “ : « 17 « 26 in., « “ 254;a05: oI. «“ «“ «“ « E « : « Te « 18 in. : «“ « 352,939- 22. «“ « “ « « « 27, “ 32 in. : “ « 371,084. 23. Peg for end of chrowine hoard, ivory, cut 24, depth 39 in., U.S.N.M. no. 370,762. 24. « « é 3K « «“ ee Tr, “ 24 in., « « 352,949. 25. Peg for butt end of dart, Pony: cut 27, depth 16 in., U.S.N.M. no. 371,003. 26. End of dart socket piece, “ , “ 19, “| 12 in., S “ 369,952. Def “ “ “ “ “ “ y “ 19, “ 24 in. “ “ 369, 996. 28. Dart Bore pices ivory, “out 7 depth 37 ine Wes: N. M. no. 353,683. 20. “ 3 “ : “ 10, “ 40 in. “ “ 354, ool. PLATE 34 Miyowagh, Old Bering Sea and Punuk cultures 1. Arrowhead, bone, cut 27, depth 26 in., U.S.N.M. no. 372,020. 2. “ : “ 5 “ 18, “ 36 in., “ « 369,751. Qe - fe EE TR wen ee he AA < NESTLE S3Se 4. a yo, 8 8 Ss Coosinis is “ 369,897. 5. ie SAD cen ser erie aD COME i. Sar KO2 Se 6. s je ed Rt RTO Sas AMinies . *" 370)040: 7 “ ; “ , «“ 75 “ 20 in., “ “ 354,292 (Punuk). 8. M We eel Os, Matra Outi. $ “ 370,079. Ou" ‘ ,AMOrYer |) 20, 0 2 9 12m, ‘ “ 370,326 (Punuk). 10. e 5 lone, “iy, 1 hale is “ 369,963. Tie ., Se gom £25) SOT ants, # 372,012: 12s - Ce ee ee eee Pes Onn Si “ 370,489. 13s a , Ivory, in wooden shaft, cut 9, depth 24 in., U.S.N.M. no. 354,451. PLATE 35 Miyowagh, Old Bering Sea and Punuk cultures 1. Meat hook, ivory, cut 7, depth 37 in., U.S.N.M. no. 353,684. 2 ss See ee ee Foye, eee Oy sinnee e ) 2535n52: 3 ff Sah ty th! ROE ATO: ee Reema Onde : >) 370101 . 4: «“ «“ § “ ; “ 22, «“ 36 in., « « 370,497. Spe Le es a Sie: : “353,835. 6. f Sate ES S20" eee ions f “ 370,355 (Punuk). 7. « «“ ‘ “ ‘ “ 2 “ 24 in., “ “ 353,203. 8. « CC Ge oe AS Shee hee ee? OnITIEe ‘s “ 354,288 (Punuk). 9g; Smallivory hooky = 95; -s2 lls, : = 354.5017 10. “ < EID 3 ne Re Sites s = 270167,6 11. Wound plug, wood,“ 19, “ 45 in., ss Sea OLO7/ 5: 12) « el ES ee WTOe | pet et ORI g = 3270'200: 13: «“ “ , “ ; “ 25, “ 72 in., “ “ 370,929. 14 “ Sy ds al 19, «“ 59 in., «“ «“ 370,192. ; NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 405 PLATE 36 Miyowagh, Old Bering Sea and Punuk cultures 1. Fishline sinker, iv Ory cut 18, depth 59 in., U.S.N.M. no. 369,889. 2 4 are avalon ee Sian: cc “ 369,865. Zz “ «“ ; «“ 5 “ TI “ 46 1 in., “ “ 353,709. 4. “ “ ; “ ; “ 24, “ 36 in., “ “ B7 OMS: Ss “ “ B “ ‘ “ ee “ 20 ine “ “ 353,945. 6 “ “ ‘ “ ; “ 4, «“ 48 in., “ “ 353,399. 7 “ oe ee O; ULSAN UM. no. 354.405; 8. e ey. | t8,depth 27 m., US. N-Msne, 360,712: 9. “ «“ : “ ; “ 19, “ 38 ile “ “ 370,059. 10. € Se eek ao Ton TOTIni: bs “ 369,679. TI. “ “ ; “ ; « 19, “ 5 in., “ «“ 369,958. 12. “ “ ; “ ; “ By “ 16 in., “ “ Boal rs: “ “ ; “ : “ Gp “ 75 in., “ “ 2535779: 14. “ “ , “ P “ 1; “ 38 in., “ “ 352,974. 15. “ “ F “ t “ 19, “ 34 in., “ “ 370,038. 16. & Seer vOneweae A i eA Saink. « “ 353,404. 17. i er IVOLY, 8) ES; w= <30uin., ce “ 369,755: 18. “ “ ; «“ ; “ : 6, “ I2 in., “ “ 353,539- 19. “ “ ; “ ; “ 235 “ 79 in., “ “ 370,693. 20. “ “ ; “ : “ it? “ 30 in., “ “ 352,961. 21. “ «“ i “ “ m5 «“ 40 in., “ “ 354,174. 22. us yi tae Gaiinished harpoon head), ivory, cut 15, depth 4o in., U.S.N.M. no. 354,173. PLATE 37 Miyowagh, Old Bering Sea culture 1. Fragmentary throwing board, wood, cut 28, depth 37 in., U.S.N.M. no. 371,192. 2. Lower end of « : SE Site Saaees Bernal G: ~ 360;877- 3. Ice creeper, bone, cut 18, depth 81 in., U.S.N.M. no. aE, 945. blo ee Dalecn as cselO wee | (Slain. é es 7Onle7. = « « ae 23 « 58 in., « “ 370,610. 6. Ivory Ghiect, cut 24, depth 80 in., U.S.N.M. no. 372,003. 7. Scraper (?), wood, cut 8, depth oe in., U.S.N.M. no. 353,861. 8. Wooden object, purchased, U.S.N.M. no. 372,038. g. Cross brace for kayak frame, wood, cut 19, depth 59 in., U.S.N.M. no. 370,193. PLATE 38 Miyowagh, Old Bering Sea and Punuk cultures 1. Knife, slate blade, wooden handle, cut 26, depth 55 in., U.S.N.M. no. 370,976. 2. “ ; “ “ ; “ “ ; “ fs, “ 29 i gloe “ “ 353,964. 2: “ «“ “ “ “ “ 22 “ 64 1 Ines “ “ 370 641. 4. Knife handle, aed: cut 24, sees 36 in., U.S.N.M. ne. 370,739. 5. £ 2 panwoynys “Se ey 32 in. 370,029. 6. «“ « 4 «“ ; «“ 18, «“ 30 in. “ « 369,731. 7, « « : « é « 7, «“ 46 in., « « 353,710. 8. « « «“ « 19 « 73 in. « « 370 236. 9. « ae bones —yiese |." 872 in., a “ 370,928. oO. “ “ ‘ «“ , «“ 18, “ 48 in., “ «“ 369,843. -_ 27 406 SMITHSONIAN MISGELLANEOUS COLLECTIONS VOL. 96 11. Knife sharpener, young walrus tusk, cut 19, depth 76 in., U.S.N.M. no. 370,246. 2 co, fe 18, 27, 13. « “ ; « « «“ é « 18, « 36 rinas «“ « 369,750. 14. “ “ ’ ¢ s : ’ w 18, "4 36 in., ih $ 369,750 a. PLATE 39 Hillside site, Old Bering Sea culture 1. Tip of harpoon blade, slate, house no. 1, U.S.N.M. no. 352,673. 2 Go a « “ : “ : “ “ ri « «“ 352,673. 3 B « en a «“ “ ih “ « 352,673. Ai ‘ ava “,among rocks, “ “ 352,888. 5. “ “ “ “ : «“« . “ “ k “ “ 352,566. 6. Harpoon or knife blade, “ , house no. 1, * "352,792: 7 « “ “ “ «“ , among rocks, “ « 352,562. 8. & tae ee eehousemon2: 7 252;A0Az 9. “ “ « “ j “ : « «“ 1 «“ “ 352,672. 10. “ “ « « : «“ ‘ “ “ iis “ “ 352,792. id sy The [7 aeamnonganockss e “ 352,882. 12) £ pore © ene ee aouse most: - 352.0725 13. Knife or lance blade, = 4 ily i 59252672" 14. “blade, slate, house no. 1, U.S.N.M. no. 352,672. 15. e “,. , first (upper) level of midden, U.S.N.M. no. 352,704. 16. . Seo shouse nose UsosNeVieemonss2:4295 17. « eae « “9 « « 352,440. 18. « “ : « i «“ « 2 « « 352,438. 19. Implement with rubbed edges, slate, between houses no. 1 and 2, U.S.N.M. no. 369,622. 20. PASSA ema rubbed enees, slate, third level of midden, U.S.N.M. no. 352,752. 21. « , * second “ «& « « “ 352,725. 22. i ie i > among rocks, UESuNe M. no. 352,888. 23. Ulu blade, Se) house no. 1, U.S.N.M. no. 352,683. 24. “ “ : ; «“ “ Te “ “ 352, 683. 25. “ “ ; “ ‘ «“ “ 2F “ “ 352,458. PLATE 40 Hillside site, Old Bering Sea culture 1. Knife Does chipped slate, house no. 2, U.S.N.M. no. 352,675. 2. “ : « « 5 “ «“ 2, “ “ 352,450. Be < Bae i “~, among rocks, « “ 352,881. rd “ “ : « ie “ « . “ «“ 352,889. 5. « ae s “ , house no. 2, “ 352,675. 6. £ Ere $ “, third level of midden, U.S.N.M. no. 352,748. "he a aa “ , house no. 1, U.S.N.M. no. 352,675. 8. “ ol. Us “, among rocks, U.S.N.M. no. 352,560. 9. . pets = “ , third level of midden, U.S.N.M. no. 352,749. 10. e ae S “, fourth (lowest) level of midden, U.S.N.M. no. 352,771. II. e “, chipped chert: house no. 2, U.S.N.M. no. 352,448. 12: ¢ “ or arrow point, chipped chert, house no. 2, U.S.N.M. no. 352,803. NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 407 13 a point, ined! jasper, house no. 2, U.S.N.M. no. 352,674. 14. si “,amonprocks,, © “ 352,566. 15. Blade, Shipped quartz, among rocks, U.S.N.M. no. 352,881. 16. “ « jasper, « « « « 352,881. 17. Knife blade, ehipped:j jasper, below ee stones, house no. 1, U.S.N.M. no. 352,656. 18. : ; “ , house no. 2, U.S.N.M. no. 352,449. 19. “ “ s “ “ : “ “ Ds “ “ 352,447. 20. Unfinished blade, chipped slate, house no. 1, “ = 3352075. 21. e PAS “ “, among rocks, “ “ 352,884. PLATE 41 Hillside site, Old Bering Sea culture 1. Drill point, slate, among rocks, U.S.N.M. no. 352,890. a “ (2?) quartz, among rocks, U.S.N.M. no. 352,890. mame “or scraper, slate, fourth (lowest) level of midden, U.S.N.M. no. 352,773. Tne “ , slate, among rocks, U.S.N.M. no. 352,569. eeinianciarll, ©, © j = “ 352,890. 6. Hand drill, chert, house no. 1, a “3521070: 7. Drill point, slate, “ oer: ": ase O77 8. Graver, chert, house no. 1, U.S.N.M. no. 352,677. 9. Rete ae among rocks, “ “ 352,570. 10. “ ; “ “ “ “ “ 352, 570. Tae emer apis (upper) evel of midden, U.S.N.M. no. 352,707. 12. Se Te eOUSEMON2 Wels: IND Minor 352,450: ee “, quartz, second level of midden, U.S.N.M. no. 352,727. 14. “, chalcedony, house no. 1, U.S.N.M. no. 352,678. 15. Side scraper, jasper, among rocks, x ein 3'5 2557.0: 11a “, quartz, first (upper) level of midden, U.S.N.M. no. 352,707. ieee ane “ , Jasper, among rocks, U.S.N.M. no. 352,570. TS “, “ , between houses no. 1 and 2, U.S.N.M. no. 369,625. NOee eee Ouse Os 2, SUES Ni Ve no3352)463 20. « “ ‘ « ' « “ iis « “ 352,808. 2s “, chalcedony, house no. 1,“ “ 352,806. Biren “, jasper, house no. 2, My “ 352,455- Dana “«, “ , below floor stones, house no. 2, U.S.N.M. no. 352,525. 2 es es eiousemos ai USN eM: no) 352,800: ier aS “«, “ , third level of midden, U.S.N.M. no. 352,751. Dow “~~, chert, fourth (lowest) level of midden, U.S.N.M. no. 352,777 Bae, mS “, Jasper, house no. 2, U.S.N.M. no. 352,454. Dor a “«, “ , fourth (lowest) level of midden, U.S.N.M. no. 352,776. BO ae ee ee whousemon 1. U-S5N- Mis no: 352.502. Bon — “ , chalcedony, house no. 1, “ 3523790: Rite “ , Jasper, house no. 1, s © 9352,674: 408 — ee oe Oar aa ee _ ~ aPwnd ea PLATE 44 Hillside site, Old Bering Sea culture Sledge runner, ivory, house no. 1, U.S.N.M. no. 352,615. “ “ 5 “ : “ “ 1 “ “ 352,615 a. “ “ ; “ : “ “ Dr “ “ 352 147 ite “ “ : “ ; “ “ ify “ “ 352,615 b. “ “ : “ F “ “ 2a “ “ 352,490. SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 PLATE 42 Hillside site, Old Bering Sea culture Side scraper, silicious slate, house no. 1, U.S.N.M. no. 352,817. & “, prase, house no. 1, U.S.N.M. no. 352,682. « “« es « a a, « « 352,800. End “ “ “ “ ie “ “ 352, 682. yy ST ae peccand level of midden, U.S.N.M. no. 352,727. a “« > —“_,house.no. 2,U.5.N.M. no. 352,452: e “—, “- , between houses no. 1 and 2, U.S.N.M. no. 369,625. & “, “ , among rocks, U.S.N.M. no. 352,889. “ “ , “ , house no. it “ “ 252) 793. é cs “ , fourth (lowest) level of midden, U.S.N.M. no. 352,774. Scraper, slate, fontth (lowest) level of midden, U.S.N.M. no. 352,770. Adzlike scraper, chert, among Rocks U.S.N.M. no. 352,568. « “ , prase, « : « « 352,567. . “, chert, house no. 1, cs “ 352,681. Scraper pias house no. 1, U.S.N.M. no. 352,677. : : “ “ Dy “ “ 352,451. ts Scraper with smoothed edges, prase, house no. 1, U.S.N.M. no. 352,817. « « « EAMES ee among rocks, « « 352,568. « « « 1) house no. 1, « « 352,794. Rubbing tool, carbonaceous slate, house no. I, S “ 352,680. . Adz blade, basalt, house no. 1, U.S.N.M. no. 352,791. PLATE 43 Hillside site, Old Bering Sea culture Rubbing stone, basalt, house no. 1, U.S.N.M. no. 352,690. ec “ , vesicular besalt, house no. 1, U.S.N.M. no. 352,812. « “ , sandstone, second level of midden, “ “352,726. é “ , basalt, below floor stones, house no. 1, U.S.N.M. no. 352,657. e “ , vesicular basalt, house no. 2, U.S.N.M. no. 352,460. a “~, scoria, house no. 1, U.S.N.M. no. 352,687. Whetstone, basalt, house no. 1, U.S.N.M. no. 352,689. s , sandstone, “ Sa, : “2521002; « « ; « «“ i; « «“ 352,815. & , diorite, € el - “ 352,689. Rubbing stone, scoria, house no. 1, se = 252A461- “ “ “ : “ “ I “ “ 352,686. “ ‘~ebasali, = ies m'' 3152685. Hammer stone, diabase, “ SO * “ 352,811. NO. I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 409 PLATE 45 Miyowagh, Old Bering Sea culture I. ee aS ivory, cut 22, coe 46 in., U.S.N.M. no. 370,522. 2. eee 275 S2urha 0S “371,085. 3. « «“ B ae We zie eS 46 i Ins, * 7 353,664. 4. Sledge shoe, “ , “ 2, “ 43 in., «“ «“ ana asi. 5. te « 7 allot ae « 28 in., “ 2353, LL0: 6. “ « . «“ f « 15, “ 14 in., « “ 354,148 a. 7. “ « ; “ : “ oF. « 9 inten (fe a 370,986. 8. «“ « ; « ; TS: ss 44 in., : 7 369,807. 9. « ee Sy ees Ts Y 75 in., . ‘ 353,791. PLATE 46 Miyowagh, Old Bering Sea culture Adz ae ivory, cut 19, depth 76 in., U.S.N.M. no. 371,9 “ “ “ , 9, U.S.N.M. no. 354,443. “ head, ane purchased, “ “ 354,464. a nae Cue 1S, depth 25 in, USIN eM. no: 360;708. “ « : wood, « 25, “ 84 in., “ “ 370,940. « abate“ 15, « 33 in., « “ 354,157: « “ oes 24, « 52 in., “« « 370,768. Pick Batarile, wood, cut 27,“ 46 in., bs © 276, SO! Base of spindle buzz, ivory, cut 9, depth 48 in., U.S.N.M. no. 354,090. “ «“ “ “ : “ s «“ 27, « 22 in., « “ 371,022. Handle of “ ae soe A SOL GACT. « “ 370,801. Rod of « « ou 19, « 24 in., « « 370,001. Drill rest, ivory, ene 21, depth 42 in., U.S.N.M. no. 370,439. Part of umiak keel, wood, cut 23, depth 44 in., U.S.N.M. no. 369,827. Endvot harpoonshatt, |“. 18) “ 44 in., 2 “ 369,821. ee en | SN lipo) feo) sienna > Ce de PLATE 47 Miyowagh, Old Bering Sea culture 1. Drill bow, wood, cut 19, depth 45 in., U.S.N.M. no. 370,103. e=eeaiwnandie; bone, © 23, “ 72) in., g “ 371,980. 3 “ “ ; “ ; “ 25 “ 89 in., “ “ 370,953- AS ee WOOC = atc 2) sl AONIt., s ~ 252043. 5. Wooden handle, cut 27, depth 26 in., U.S.N.M. no. 371,048. 6. Ivory handle, SELON OOMINE: s SGA DE. 7 atts ow UesaNe Vie no7 354,300: 8. Ivory wedge, Seo: a “254°205. 9. Wedge, used secondarily as drill rest, cut 13, depth 60 in., U.S.N.M. no. 353,959. Scraper (?), wood, cut 27, depth 46 in., U.S.N.M. no. 371,154. Seraper handle (7), “ 27, “42 in, “ “ cy Ay p pAeaBe: ivory, cut 25, depth 48 in., U. S. N.M. no. 370,886. “ “ ; 19, 76 in., Hook, Son ee2 eee sais ATOATc: “ “ “ 370,250. 379,517- . Carving of human head, wood, cut 10, depth 54 in., U.S.N.M. no. 353,963. Ladle, bone, cut 23, depth 81 in., U.S.N.M. no. 370,697. Pottery paddle, cut 9, U.S.N.M. no. 354,414. 410 CO Oo ON OC os ae —~ _ ~ Ny Oe | Oe oR ON ON eon NN No oS a) tae: ORE es Se DAKE No SMITHSONIAN MISCELLANEOUS COLLECTIONS Drill point, bone, cut 22, depth “ “ “ “ “ ’ ’ ’ Handtdrill es “ «“ “ « } ’ “ OE tts PLATE 48 Miyowagh, Old Bering Sea culture “ 27, 44 in., Ts, 8 Bepsie. « Loan) So Ganinee «“ 200 eeeOyine s 24 OFine e , Ivory, Oe OS RLites ss Reamer, bone, cut 29, depth 17 in., U.S.N.M. Ce AIVOLY abe 23) ke eee nee é Rubbing tool, bone, cut 23, depth 58 in., U.S. “ “ ; “ : “ 2B “ We in., “ “ F «“ . “ 22. “ 36 ifl., Reamer, bone, cut 18, depth 59 in., U.S.N.M. “ ; “ : “ 24, “ 18 in., “ Awl, bone, ueoohe cam SSmnih. ss “ . “ : “ 19, “ 51 in., “ “ “ “ 18, “ 25 in., “ Flaking tool, bone, cut “ bi “ ’ ’ “ “ “ ’ ’ Awl, ivory, cut “ “ “ ’ ’ 24, “ “ “ ’ ’ 25; 30 in., U.S.N.M. no. 370,487. 2 Belg Bile 369,720. 369,906. ee Yd ea the “370,734. 370,207. no. 371,205. “370,558. N.M. no. 370,629. ‘ “ 370,666. 370,496. no. 369,890. Sa 7O.7aloe 370,621. oe 370M Se “ 369,700. “ “ I, depth 12 in., U.S.N.M. no. 353,915. ie PTA TG ef “ 369,683. iy GD ole : “ 369,932. 7, depth 62 in., U.S.N.M. no. 353,781. “54 in., : “370,784. CGI iis, f “ 370,906. Implement, ivory, cut 18, depth 27 in., U.S.N.M. no. 369,719. Ivory pick, house no. PLATE 49 Hillside site, Old Bering Sea culture 2, U.S.N.M. no. 352,466. Te ‘ DBE feliitey, 2, “ 352,467. Ie s “ 352,605. Ile s “> 2152;612: il, s “ 352,604. Ty, i. = 352,010: PLATE 50 Miyowagh, Old Bering Sea culture Wooden paddle, cut 23, depth 81 in., U.S.N.M. no. 370,699. Mattock blade, bone, cut 24, depth 25 in., U.S.N.M. no. 370,729. Sledge cross bar, cut ga, depth 36 in., U.S.N.M. no. 354,349. Knife, bone, cut 19, depth 67 in., U.S.N.M. no. 370,230. Ajagaq, bone, cut 7, “ “ “ 37 in., 353,690. Snow shovel, bone, cut 18, depth 49 in., U.S.N.M. no. 369,860. VOL. 96 NO. SO SIWAKERY Do om ~ = — te ee I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS Ail PLATE 51 Miyowagh, Old Bering Sea culture Ulu, wooden handle, stone blade, cut 22, depth 86 in., U.S.N.M. no. 370,286. “ “ “ “ “ 2 “ 18, “ 59 in. , “ “ 369,887. Ulu handle, wood, en 9, ocecl 20 in., U.S.N.M. no. 354,417. Sie Wi doa Meh era Lo 86 in., « “ 370,289. « « : « : “ 25, “ 33 in., «“ «“ 370,860. « « ; “ ? “ 19, “ 45 in., “ “ 370,094. «“ «“ 2 « : “ 27, «“ 36 in., «“ “ 371,100. Fat scraper, ivory, cut 12, depth 25 in., “ m3 53,982: . fee ogee) WES INL ML. ino! 35454:76- Scraper, bone, cut 10, depth 22 in., U.S.N.M. no. 353,978. Fat scraper, ivory, cut 10, depth 22 in., U.S.N.M. no. 353,979. Ivory vessel, cut 16, oo 23 in., U.S.N.M. no. 354,222. Ladle, wood, “ 23, 72 in. id =9370}070: SeETIVOLY, fa TOW 1 9 3O)Ine i 37.0;016: Spoomyantler, — 19). < (67 ins, ‘ = 270:224° adem eco Toh, qouin.. « “ 369,845. PLATE 52 Hillside site, Old Bering Sea culture Potsherd, first (upper) level of midden, U.S.N.M. no. 352,702. “~~ , among rocks, U.S.N.M. no. 352,880. “ : “ “ : “ “ 352,559. “ . “ “ : “ “ 352,880. “ P “ “ . “ “ 352 ,880. “ . “ “ . “ “ 352,880. we , s ae « “ 352,880. E , house no. 1, < “ 352,789. “ : “ “ I ; “ “ 352,789. “ : “ “ ng “ “ 352,671. “ : “ “ ie “ “ 352,671. “« , among rocks, by » 352,559- < , first (upper) level of midden, U.S.N.M. no. 352,702. « , below floor stones, house no. 1, “ ©. 352,055. PLATE 53 Baleen vessel, below floor stones of house no. 2, U.S.N.M. no. 352,515. Pottery lamp, from old grave at Kowieruk, east of Teller, Seward Peninsula, U.S.N.M. no. 347,877. PLATE 54 Hillside site, Old Bering Sea culture Bottom of baleen vessel, wood, beneath floor stones, house no. 2, U.S.N.M. no 352,536. Bottom of baleen vessel, wood, beneath floor stones, house no. 2, U.S.N.M. no 352,516. 412 SI AK SOIT EN ee CSS S Ouice se i SMITHSONIAN MISCELLANEOUS COLLECTIONS Bottom of baleen vessel, wood, house no. 1, U.S.N.M. no. 352,627. VOL. 96 Fragment of wooden vessel, beneath floor stones, house no. 2, U.S.N.M. no. 352,538. Fragment of wooden vessel, house no. 1, U.S.N.M. no. 352,638. Wooden shovel (2), house no. 1, U.S.N.M. no. 352,627. “ “ “ “ “ “ 352, 628. 352,627. “ “ “ “ “ “ “ PLATE 55 Miyowagh, Old Bering Sea culture Toy bow, baleen, cut 19, depth 51 in., U.S.N.M. no. 370,133. Part of a baleen snare, cut 23, depth 39 in., U.S.N.M. no. 370,592. 370,863. 578. Piece of knotted baleen, cut 25, “ 33 in., t: & Ventilator rim (?), cut 23, depth 32 in., U.S.N.M. no. 370 Drum handle and part of rim, cut 16, depth 23 in., U.S.N.M. no. 354,224. Wooden shaft with baleen wrapping, cut 9, U.S.N.M. no. 354,420. Ice scoop, baleen, cut 18, depth 44 in., = = 260) S rds Ring of baleen, cut 9, depth 48 in., U.S.N.M. no. 354,088. Bishline:! 0°= 59" S268) eo taghne, - Se s7Ooyile PLATE 56 Miyowagh, Old Bering Sea culture Engraving tool, wooden handle, jasper point, cut 19, depth 38 in., 370,059. U.S.N.M. no. Engraving tool, wooden handle, jasper point, cut 19, depth 67 in., U.S.N.M. no. 370,223. Fishhook (?), cut 9, U.S.N.M. no. 354,474. Toy knife, baleen, cut 19, depth 45 in., U.S.N.M. no. 370,099. Toy, baleen, cut 19, depth 38 in., U.S.N.M. no. 370,069. Toy paddle, baleen, cut 27, depth 26 in., U.S.N.M. no. 371,052. Toy, baleen, cut 19, depth 42 in., U.S.N.M. no. 370,081. Plaited object, baleen, cut 18, depth 27 in., U.S.N.M. no. 369,725. “ “ “ “ , 23; “ 22)ine, . “ 370,560. Pieces of ord lashed with baleen, cut 18, depth 40 in., U.S.N.M. no. 369,783 Roll of baleen, cut 18, depth 32 in., U.S.N.M. no. Oo Tae Part of snare (?), cut 18, depth 32 a a $9369;746: . Sliver of bone attached to wooden shaft, cut 27, depth 16 in., U.S.N.M. no. 371,000. Bundle of grass, wrapped with baleen, cut 27, depth 26 in., U.S.N.M. no. 371,053. Toy bow, baleen, cut 18, depth 32 in., U.S.N.M. no. 369,742. Human hair tied to piece of wood, cut 21, depth 24 in., U.S.N.M. no. 370,403. PLATE 57 Miyowagh, Old Bering Sea culture Lance foreshaft, wood, cut 23, depth 81 in., U.S.N.M. no. Part of drill shaft, mood cut 23, depth 81 in., “ s “ “ “ “ : 4 “ 18, “ 36 Bly, “ “ End of arrow i 4c) anion 2 ese ian : S “ “ “ “ ' “ ‘ “ 18, “ 66 in., “ “ 370,701. 379,795. 369,767. 369,916. 379,034. NO. 6. re 8. 9; lie 52: 13. 14. 15. 16. 7: 18. 19. 20. DT wa. _ SON ONS CEO CO ON ONE one Os I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 413 Handle of engraving tool (?), cut 19, depth 79 in., U.S.N.M. no. 370,263. Part of snare (?), wood, cut 9, U.S.N.M. no. 354,422. Section of drum rim, wood, cut 16, depth 46 in., U.S.N.M. no. 354,255. 10. Fragmentary wooden rods, painted red, cut 23, depth 81 in., U.S.N.M. no. 370,700. Wooden object, cut 24, depth 68 in., U.S.N.M. no. 370,816. « are 18, Ko A36 in., “« « 369,759. «“ « ; « 18, « 36 in., « «“ 369,767. « «“ ; “ 18, “ 49 in., «“ “ 369,861. s “, painted red, cut 18, depth 59 in., U.S.N.M. no. 369,893. s ee cut 28) depth 2imine. U:S.Ne Me no: 371.178: « oO Shey « 81 in., “ “ 370,703. Wooden cylinder, cut 23, depth 32 in., “ a 7ON5 70: ss “ , painted red, cut 23, depth 45 in., U.S.N.M. no. 370,596. Piece of drying rack slat, wood, cut 23, depth 81 in., “ BON OS Wooden object, cut 18, depth 4o in., U.S.N.M. no. 369,782. Part of drying rack frame, cut 19, depth 45 in., U.S.N.M. no. 370,108. PLATE 58 Miyowagh, Old Bering Sea culture Snow goggles, wood, cut 19, depth 54 in., U.S.N.M. no. 370,150. “ «“ A « 24, «“ 39 in., «“ « 370,753, Bone button, leu 24, depth 64 in., U.S. N. M. no. 370,794. ivonvaouttons 19; gOun, 4 “ 370,017. «“ “ : « 24, « 25 in., « « 371,990. Piece of brow band, ivory, cut 18, depth 81 in., U.S.N.M. no. 369,947. Brow band, ivory, cut 7, depth 75 in., U.S.N.M. no. 353,780. Piece of brow band, ivory, cut 23, depth 72 in., U.S.N.M. no. 370,663. Brow band, ivory, cut 27, depth 52 in., U.S.N.M. no. 371,159. Comb, ivory, cut 7, depth 67 in., U.S.N.M. no. 353,766. Shuttle, ivory, cut 7, depth 67 in., SES SO aioe Piece of brow band, ivory, cut 3, depth 60 in., U.S.N.M. no. 353,291. Ivory object, cut 23, depth 58 in., U.S.N.M. no. 370,625. Potsherd with human figure decoration, cut 9 a, depth 60 in., U.S.N.M. no. 354,369. Small bone plaque, cut 5, depth 32 in., U.S.N.M. no. 353,503. . Fused walrus teeth, used as drill rest, cut 16, depth 36 in., U.S.N.M. no. 354,243. lyory object, cut 7, depth 67 in., U.S.N.M. no. 353,772. “ : “ 8, “ 20 in., “ “ 353, 843. PLATE 59 Miyowagh, Old Bering Sea culture Toy kayak, wood, cut 18, depth 44 in., U.S.N.M. no. 369,828. Cross piece of toy umiak, wood, cut 13, depth 54 in., U.S.N.M. no. 353,960. Keel piece “« « « ee «“ 18, “ 32 in. be «“ « 369, 746. Figure of man in toy eaale (?), wood, cut 18, depth 44 in., U.S.N.M. no. 369,829. - Figure of man in toy kayak (?), wood, cut 6, depth 42 in., U.S.N.M. no. 369,596. Toy umiak, bark, cut 19, depth 54 in., U.S.N.M. no. Eon End seat for toy umiak, bark, cut 18, depth 32 in., U.S.N.M. no. 369,741. ATA SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 8, 9. Toy sledge runner, wood, cut 25, depth 43 in., U.S.N.M. no. 370,878. 10. Cross bars of toy sledge, wood, cut 25, depth 43 in., “ = 3 7O STS. 11. Toy sledge runner, ivory, cut 18, depth 12 in., U.S.N.M. no. 369,681. 12. Ivory top (?), cut 18, depth 12 in., U.S.N.M. no. 369,683. 13. Toy whales, bark, U.S.N.M. no. 354,481. 14. Toy whale, “ , cut 7, depth 75 in., U.S.N.M. no. 353,785. 15. Wooden object with spiral groove, cut 10, depth 54 in., U.S.N.M. no. 354,023. 16. Doll, wood, cut 19, depth 34 in., U.S.N.M. no. 370,043. “ “ 17.- “ , bark, = Gone ong suin:, 354,092. 18. Toy harpoon head, ivory, cut 24, depth 54 in., U.S.N.M. no. 371,995. i@), 2 70 eC hae oat agen OMe cava SOIT Ss ~ 237 SAs: 20. “ “ “ F “ : “ 19, “ 56 in., “ “ 271 Ole 21. “ “ “ “ ; “ 7h “ 75 in., “ “ 353,784. 22. Toy winged abiest, eee Sat meen LOOMS - = 37h S50: 22. “ “ , “ ’ “ 2h “ 46 in., “ “ 371,144. 24.5% é oh a ote Seen, L2G uiiice : “ 353,486. 25. Toy polar bear, . SP 2A, Pee eSauine 4 = 3705770: 26. Toy bow, wood, cut 19, deneh gain, URS NE M. no. 370,047. 2g “ “ ; “ ‘ “ 24, “ 54 in., “ “ 370, 789. PLATE 60 Miyowagh, early Punuk Circular dish, whale bone, cut 17, depth 34 in., U.S.N.M. no. 369,655. Adz head, ivory, cut 19, depth 42 in., U.S.N.M. no. 371,903. Ivory object, cut 9, depth 36 in., U.S.N.M. no. 354,083. Cord hasdle (?), ivory, cut 19, depth 17 in., U.S.N.M. no. 369,977. “ « ; « : «“ 23, «“ 24 in., «“ « 370,548. Broken snow goggles, ivory, surface of ridden: y “ 354,478. Ornamental plug, ivory, cut 2, depth 28 in., & = a53nlmos Toy bow, reinforced with baleen, cut 19, depth 42 in., U.S.N.M. no. 370, O81. Drill shaft, wood and ivory, purchased, U.S.N.M. no. 354,495. Engraving tool, ivory, cut 27, depth 22 in., “ Ses IRO2Gs “ “ “ ; “ ae “ 20 IMs, “ “ 353, 209. Bone shovel, Git 4, depth 15 in., U.S.N.M. no. 353,310. ty So Cassie en. CaeS co) Se i _ iS) PLATE 61 1. View of levoghiyogq, on the gravel plain, with Miyowagh in foreground. 2. Excavations at house no. 7, levoghiyoq; stone entrance passage in foreground. 3. House no. 7 at a later stage of excavation, showing the mass of fallen wall and roof timbers, and stone walled entrance in foreground. 4. View of the same looking in opposite direction, toward the south. PLATE 62 2. Views of the fallen wall and roof timbers of house no. 7, Ievoghiyoq. 3. Ruins of house no. 6, a stone-walled house of later type at levoghiyoq. 4. Round antechamber at end of entrance passage to house no. 6. NO. Es zi 4. 5 to wn — ASD ON ANA! DY oO ed ONE i se ea) I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 415 PLATE 63 Piece of blubber-soaked walrus skin in one of the cuts at Seklowaghyaget. 3. House no. 8, Seklowaghyaget. ° Oval enlargement of the entrance passage to house no. 8. View of same before the fallen roof material—whale ribs—had been removed. PLATE 64 Oval annex to passage of house no. 9, with stone walls and floor and roof of whale Tibss Uncovering the wooden floor of house no. 10, the latest house excavated at Gambell; abandoned about 50 years ago. Underground cache made of whale bones, Gambell. Small underground house of whale bones and stones at Naskok, east of Gambell: abandoned about 50 years ago. Ruins of square stone house at Mirrukta, east of Gambell. PLATE 65 levoghiyoq, Punuk culture Broken needle case, ivory, cut 5, depth 43 in., U.S.N.M. no. 355,461. Needle case, ivory, cut 9, depth 12 in., U.S.N.M. no. 372,080. hoitehandies = = t = 28ane, a “ 354,610. Fragmentary dart socket piece, ivory, cut I, depth 31 in., U.S.N.M. no. 354,649. Fragmentary dart socket piece, ivory, cut 5, depth 24 in., U.S.N.M. no. 355,259. Link ornament, ivory, cut 8, depth 8 in., U.S.N.M. no. 372,052. Ivory object, cut 1, depth 42 in., U.S.N. M. no. 354,744- “ “ : Per wre; Cok I. 355,339: Wrist SEE: ivory, cut 1, depth 36 in., U.S.N.M. no. 354,702. “ ; “ F “ ins “ 44 in. “ “ 354,760. « UE anes a «“ 36 in., « « 354,703. PLATE 66 levoghiyoq, Punuk culture Fragment of wrist guard, ivory, purchased, U.S.N.M. no. 372,100. Wrist guard, ivory, purchased, U.S.N.M. no. 355,767. Harpoon socket piece, ivory, purchased, U.S.N.M. no. 354,496. Ivory object, cut 1, depth 42 in., U.S.N.M. no. 354,743. “ “ s “ 5, “ 14 i lop “ “ 355,181. “ “ : “ 8, “ 44 in., “ “ 372,072. Needle case, purchased, U.S.N.M. no. 372,144. Ivory object, a ' 7 ass bey Tie PLATE 67 Seklowaghyaget and old section of Gambell, Punuk culture Ulu handle, ivory, Seklowaghyaget, purchased, U.S.N.M. no. 372,743. Base of spindle buzz, ivory, Seklowaghyaget, ee ey U.S.N.M. no. 372,138. Broken dart socket piece, ivory, “ ; ; . eS aA Tea: 416 ON boo SION Ao Se 1-8. WN oS NCA Gs SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 Bodkin, ivory, Seklowaghyaget, purchased, U.S.N.M. no. 356,125. Wrist guard, ivory, old section of Gambell, cut 1, depth 8 in., U.S.N.M. no. 356,192. Ivory object, Seklowaghyaget, purchased, U.S.N.M. no. 364,180. Wrist guard, Seklowaghyaget, cut 3, depth 20 in., U.S.N.M. no. 355,952. Ivory object, w , purchased, U.S.N.M. no. 356,131. PLATE 68 levoghiyoq and Seklowaghyaget, Punuk culture Ivory trident, levoghiyoq, cut 4, depth 31 in., U.S.N.M. no. 355,064. “ “ «“ , “ A, “ 31 in. «“ “ ana 065. # - | Seklowaghyaget, purchased, U. S N.M. no. 356,126. “ “ ‘ “ “ 355) 202. . “, levoghiyoq, cut & depth om in., U.S.N.M. no. 356,127. Ivory object, Seklowaghyaget, purchased, U.S.N.M. no. 372,140 a. % ae ; - ; vi “ 356,128. “ “ ; “ ; “ : “ “ 356,129. PLATE 69 Reverse of objects illustrated on pl. 68. PLATE 70 levoghiyoq, Punuk culture Harpoon head, ivory, cut 1, depth 12 in., U.S.N.M. no. 354,560. s (DONC es 5, tps welLOhimn : “ 355,179. “ “ ; “ ; “ if, «“« 22 1A “ “ 354,970. s IS NOE) © inh GS ap) Tha, ce “ 354,593. “ “ ; “ ; “ 5, “ 2 in., “ “ 355,252. . o DOney ato. ae Suinier 4 “ 372,060. oh SAG AIVOLYs) amily Reractes eA ACnae, $ 3545702: = A es Me vse eee Pane EATING e 2355) als “ “ F “ F ‘ 6, “ 36 in., “ a“ 355,045. ' Sse he ox, 0 WO eae? eS onulles : “3551356. “ “ , “ , “ 6, “ 30 in., “ “ 355,596. “ “ ; “ ; “ 6, “ 22 in., “ “ 355,557: “ “ : “ , “ 6, “ 5 anes “ “ 355,471. “ “ ; “ 5 “ 6, “ 5 in., “ “ 355,139. . Fe cake tet tate Oven Eh aaa IASI « “372,008: “ “ ; “ i “ 9, “ I2 in., “ “ 372i O078e “ “ ‘ “ : “ 9, “ 31 in., “ “ 372,093. S ees cere (ae M2) Si gle. S 4 355,073: “ “ , “ : “ I, “ 50 in., “ “ 354,779. . oN EN elie eee ie Oui % es aye bay tii = sis 00) AS. cmd As ome IT : “ 355,060. “ “ : “ ; “ 6, “ 9 in., “ “ 355,530. “ “ ‘ “ ; “ Be “ 22 ines “ «“ 354,859. NO. 28. 29. I ARCHEOLOGY PLATE 71 * ST. LAWRENCE ISLAND—-COLLINS 417 Seklowaghyaget and old section of Gambell, Punuk and modern Harpoon ae “ 356,180. ’ , , bone, house no. ’ ’ ’ bone, ivory, ’ ’ ’ ’ ivory, | - 8 8 8, 8 8 8, ivory, Seklowaghyaget, “ “ 2; SN NM NN HH KN YN cut 2, depth 22 in., U.S.N.M. «“ 49 ime, «“ PeTS Gin & «“ 18 in ‘ “ cits ats oY Sy 112) Stripe ‘ yeah ines us So wa T2Y I. . Si Be Tia . Tits} Sad s “522i. i Se its) Tins Se) Sa tie . ME ins < - 356,337. 356,335. 356,322. 356,333. 356,336. 356,323. no. 355,863. “355,903. “ 355,942. “356,039. my 55: 042" “ . 355,822. * 355;879- “ 355,821. “355,863. “ 355,841. “ 355,868. “ 355,843. te 372013; “356,029. , bone, exact provenience unknown, purchased, U.S.N.M. no. 356,713. ivory, old section of Gambell, ’ cut 1, depth 8 in., Harpoon head, bone, old section of Gambell, cut 1, depth 16 in., 356,203. Harpoon head, ivory, old section of Gambell, cut 2, depth 20 in., 356,311. Harpoon head, ivory, old section of Gambell, cut 2, depth fo in., 356,298. Harpoon peas, ivory, house no. 10, U.S.N.M. no.356,372. “ , St. Lawrence Island, collected by E. W. Nelson, U.S.N.M. no. ’ 126,915. Harpoon head, ivory, old section of Gambell, cut 1, depth 8 in., 350,178. U.S.N.M. no. U.S.N.M. no. U.S.N.M. no. US NEM not U.S.N.M. no. Harpoon head, ivory, St. Lawrence Island, collected by Riley D. Moore. PLATE 72 Various localities, Punuk culture Whaling harpoon head, 371,846. Whaling harpoon head, 353,928. Whaling harpoon head, ivory, Kitneapalok, no. 371,705. ivory, Miyowagh, ivory, Miyowagh, cut 18, depth 60 in., cut 12, depth 27 in., U.S.N.M. no. U.S.N.M. no 20 miles south of Gambell, U.S.N.M. oe ty SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 Whaling harpoon head, ivory, Seklowaghyaget, purchased, U.S.N.M. no. 356,121. ie sc “, “ , old section of Gambell, cut 1, depth 16 in., U.S.N.M. no. 356,201. Whaling harpoon head, ivory, old section of Gambell, cut 1, depth 8 in., U.S.N.M. no. 356,175. Whaling harpoon head, ivory, Kukuliak, purchased, U.S.N.M. no. 356,576. Lance head, ivory, Seklowaghyaget, purchased, ¢ * 27 SOs i “ , “ , Kukuliak, purchased, U.S.N.M. no. 356,715. PLATE 73 Various localities, Punuk and modern Harpoon foreshaft, ivory, Seklowaghyaget, cut 2, depth 28 in., U.S.N.M. no: 355,878. Harpoon foreshaft, ivory, Seklowaghyaget, surface, U.S.N.M. no. 371,639. a i. , “ , modern, Punuk Island, § *)3242701r Harpoon ice pick, bone, Ievoghiyogq, cut 5, depth 24 in., U.S.N.M. no. 355,282. Toy harpoon socket piece, ivory, Ievoghiyoq, cut 6, depth 22 in., U.S.N.M. no. 3551572: Harpoon socket piece, ivory, Seklowaghyaget, surface, U.S.N.M. no. 371,631. - . “ , house no. 9, U.S.N.M. no. 356,356. Harpoon ice pick, house no. 8, U.S.N.M. no. 356,348. ay “ — “ , bone, modern, collected by E. W. Nelson, U.S.N.M. no. 126,912. Mouthpiece for bladder float, ivory, levoghiyoq, cut 6, depth 30 in., U.S.N.M. no. 355,592. Mouthpiece for bladder float, ivory, Miyowagh, cut 4, depth 20 in., U.S.N.M. no. 3551322. Mouthpiece for bladder float, ivory, levoghiyoq, cut 6, depth 4o in., U.S.N.M. no. 355,656. Finger rest for harpoon shaft, ivory, levoghiyoq, cut 5, depth 18 in., U.S.N.M. no. 355,159. Finger rest for harpoon shaft, ivory, old section of Gambell, cut 1, depth 16 in., ~ U.S.N.M. no. 356,223. Mouthpiece for water bag, ivory, Seklowaghyaget, surface, U.S.N.M. no. 371,641. Mouthpiece for seal skin float, ivory, levoghiyoq, cut 9, depth 19 in., U.S.N.M. no. 371,381. Spur for end of throwing board, ivory, Ievoghiyoq, cut 5, depth 12 in., U.S.N.M. no. 355,128. Peg for end of dart shaft, ivory, Ievoghiyoq, cut 1, depth 50 in., U.S.N.M. no. 354,778. Wooden plug for seal skin float, Ievoghiyoq, cut 5, depth 31 in., U.S.N.M. no. 355;324- Float bar, wood, Ievoghiyogq, cut 4, depth 53 in., U.S.N.M. no. 355,098. ‘ PLATE 74 Various localities, Punuk and modern Side prong for bird dart, ivory, Ievoghiyoq, cut 4, depth 45 in., U.S.N.M. no. 355,076. Side prong for bird dart, ivory, Ievoghiyoq, cut 1, depth 28 in., U.S.N.M. no. 354,623. | NO. Bown I ARCHEOLOGY OF ST. LAWRENCE ISLAND—COLLINS 419 End prong for bird dart, ivory, levoghiyoq, cut 5, depth 12 in., U.S.N.M. no 355,194. End prong for bird dart, ivory, levoghiyoq, cut 8, depth 29 in., U.S.N.M. no. 371,277. End prong for bird dart, ivory, levoghiyoq, cut 1, depth 44 in., U.S.N.M. no 354,764. Arrowhead, WORy, Ievoghiyoq, cut 5, depth 24 in., U.S.N.M. no. 355,277. i s : s , U.S.N.M. no. 355, 738. . , bone, s , cut 1, depth 12 in., U.S.N.M. no. 354,537. “ ; “ ; “ : “ 9, “ 12 in., “ “ B75 0: “ ; “ “ : “ 6, “ 22 in., “ “ 355, 567. Blunt arrow ane ivory, Seklowaghyaget, cut 2, depth 12 in., U.S.N.M. no. Blunt arrow tip, ivory, Seklowaghyaget, cut 8, depth 60 in., U.S.N.M. no. Blunt arrow tip, ivory, Seklowaghyaget, cut 8, depth 67 in., U.S.N.M. no. Arrowhead, Hone; tv cemivod cut 5, depth 39 in., U.S.N.M. no. 355,379. “ ' ‘ ; “ it “ 28 in., “ ; “ 354, 616. . Bow brace, bone, ee Oe! les 5 in., “ 355,485. Y ae old) section of Gambell, cut 1, depth 8 in., U.S.N.M. no. Bow brace, baie levoghiyoq, cut 8, depth 44 in., U.S.N.M. no. 371,341. Sinew twister, ivory, “ Tee & (ae J2reine, . Wait atetay “ “ “ «“ : “ 5, “ 18i in. “ “ 355,164. «“ “ Wrist eed ivory, aid section of Gambell, surface, SUP AWalit . «, “ , levoghiyoq, purchased, U.S.N.M. no. 355,768. “ “ : “ ; “ : “ f “ “ 357 ; 769. a « “, modern St. Lawrence form, purchased, U.S.N.M. no. PLATE 75 Various localities, Punuk and modern Prong for a spear, lev PEniyod: cut 1, depth g in., U.S.N.M. no. 355,150. “ “ : ; “ I “ 9 in., “ “ 355, 194. Ty sold eeeiion of Gambell, cut 1, depth ai in., U.S.N.M. no. 356,213. on poe figure, purchased, U.S.N.M. no. 355,778. _ pt loan act, cut 7, depth 12 in., U.S.N.M. no. 371,488. « «“ « : “ Fe « 12 in., “ & 371,488. . - 3 panectaned U.S.N.M. no. 356,141. C : “, Seklowaghyaget, cut 7, depth 12 in., U.S.N.M. no. 371,488. Ivory doll, Cape Eaten. U.S.N.M. no. 343,004. 424 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 13. Doll head, ivory, Seklowaghyaget, purchased, U.S.N.M. no. 356,726. i. “ , “ , Punuk Island, U.S.N.M. no. 356,701. 15. Ivory top, Seklowaghyaget, cut 7, depth 36 in., U.S.N.M. no. 371,520. 16, Whale, ivory, Ievoghiyoq, “ 7, “ 8 in., a “ 355,676. 17e « ,° * |, house’no.-9, U-SIN- Me no: 356,357; TS “ : “ F “ «“ 8, “ “ 356,325. PLATE 84 Various localities, Punuk and modern Wooden fire hearth, house no. 7, Ievoghiyoq, U.S.N.M. no. 355,722. Boat paddle, house no. 7, levoghiyoq, U.S.N.M. no. 355,720. Half of pottery lamp, levoghiyoq, cut 6, depth 5 in., U.S.N.M. no. 355,464. Lamp, St. Lawrence Island, modern type, U.S.N.M. no. 342,736. Cooking pot, modern type, house no. 9, U.S.N.M. no. 356,349. Sh 9 oS) ae ((S6£ a3vd 99s ‘uor}eue[dxo 10,7) GNV1S| 3ADONSYMV 7 LS 43AO SLSVOD HLYON GNV LSAM NO SSALIS OWINSS G10 LV ‘AHOVD LVAW GNV “Td ‘tL ON ‘96 “10A *“SANOLS ONIdWNL,, “SNSQGIW AO SM3IA SNOILDA1109 SNOANVTISDSIN NVINOSHLIWS 28 ((S6€ a5ed 99s ‘uorjeuridxea 10,7) ONV1S| ADNAYMV7 LS AO ONF NYSALSVAHLNOS AAO ‘ONV1IS| MHNNNd ‘NINY ASNOH GNV SNSGCIW AO SM3ZIA @ 1d “1 ‘ON 96 “10A SNOILD41109 SNOANVTISDSIN NVINOSHLIWS (‘S68 ased 9as ‘uoleuri[dxe 10,7) GQNV1SI ADNAYMV 7 LS dO ONA N&SLSVAHLNOS “MVOATVIM AdVD 'NINY ASNOH GNv SN3GGCIW 40 SM3FIA € “1d ‘1 ON ‘96 “10A SNOILD31109 SNOANVTISOSIW NVINOSHLIWS SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO. 1, PL. 4 VIEWS AT GAMBELL, NORTHWESTERN END OF ST. LAWRENCE ISLAND (For explanation, see page 395.) (£6 ased 99s ‘uoneur[dxea 10,7) TWIFEWVS ‘ALIS 30GISTTIH LY SNOILVAVOX A lite oc as ei G “1d ‘1 ON ‘96 “1OA SNOILD31109 SNOANVT1SZ9SINW NVINOSHLIWS 9 1d ‘lt ‘ON ‘96 10A ($68 osed 99s ‘uoljeur[dxa 10,7) T1ISEWNVS “HOVMOAIW LY SNOILVAVOX9A SNOIL937109 SNOANVITS9DSIW NVINOSHLIWS SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO. 1, PL; 7 EXCAVATIONS AT MIYOWAGH, GAMBELL (lor explanation, see page 396.) ((96€ ased 9as ‘uoljeur[dxea 10,7) W1FaWNvdD “HOVMOAIW Lv SNOILVAVDXO 81d ‘tL ON ‘96 10A SNOILD3S1109 SNOANVTISDSIN NVINOSHLIWS ((968 ased 9as ‘uoneuridxa 10,;) TIAENVS “HOVYMOAIW LY SNOILVAVOX A SNOIL931109 SNOANV1T1S9SIW NVINOSHLIWS Ol “1d ‘lt ‘ON ‘96 “10A ((96® a8ed 99s ‘uorjeurldxa 10,7) TIFAEWNVD ‘HOVYMOAIW LY SNOILVAVOXy SNOILD31109 SNOANVI11S9SIW NVINOSHLIWS ((96£ ased 9as ‘uoneur[dxe 10,7) (GZ OGNV ‘v2 “€2Z ‘81 SLND) WAEWVD ‘HOVMOAIW LV SNOILVAVOX9A LL “Id ‘L ‘ON “96 “10A SNOILD371109 SNOANVT1IS9OSIW NVINOSHLIWS SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO. 1, PL. 12 IVORY OBJECTS FROM THE HILLSIDE SITE, OLD BERING SEA CULTURE. VY NATURAL SIZE. (For explanation, see pages 396-397.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO. 1, PL. 13 DECORATED IVORY OBJECTS FROM THE HILLSIDE SITE, OLD BERING SEA CULTURE. APPROX. 23 NATURAL SIZE. (For explanation, see page 397.) ((46© ased 99s ‘uoljeuRldxa 10,7) “AZIS IVWYNLYN G/p “XOUddY ‘9 ‘AZIS IVHNLVYN G/€ “XOUddYy ‘¢G ‘AZIS IWYUNLVN ‘py “€ ‘GNVIS| 3JGANOIC AILLIT WOUSs ‘9-€ “AZIS IVUNLYN % “XOUddY -ALIS ACISTTIH WO ‘2 *l “SYNLINO VAS ONIYNSG GIO ‘sLoargao ANOA| GQaLvyoos0d Vi “Id “1 ‘ON ‘96 “10A SNOILDAT110O9 SNOANVTISOSIW NVINOSHLIWS SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO. 1, PL. 15 Le DECORATED IVORY OBJECTS FROM MIYOWAGH, OLD BERING SEA CULTURE. APPROX. 3/5 NATURAL SIZE. (For explanation, see page 397.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO. 1, PL. 16 gore BS? sel Se. j \\ te \ DECORATED IVORY OBJECT FROM MIYOWAGH, OLD BERING SEA CULTURE. 5/6 NATURAL SIZE. (For explanation, see page 398.) SMITHSONIAN MISCELLANEOUS COLLECTIONS Zs VOL. 96, NO. 1, PL. 17 DECORATED IVORY SCRAPERS, ADZ HANDLE, AND NEEDLE CASES FROM MIYOWAGH, OLD BERING SEA CULTURE. APPROX. 3/5 NATURAL SIZE. (For explanation, see page 398.) 29 Cg6e ased 99s ‘uoljeur[dxe 10,7) ‘G/| GADYVIN”A S3YNLINO VAS ONIYNSG A1O ‘HOVYMOAICW WO 81 Id ‘Ll “ON ‘96 “10A SASVD AIGSAN ANYOA! GaLlvyHoosaGd SNOILD4A1109 SNOANVTISDSIN NVINOSHLIWS SMITHSONIAN MISCELLANEOUS COLLECTIONS IVORY OBJECTS FROM MIYOWAGH, OLD BERING SEA CULTURE. APPROX. 3/5 NATURAL SIZE. (For explanation, see page 398.) VOL. 96, NO. 1, PL. ((g6£ ased 99s ‘uoljeuR[dxe 10,7) "AZIS IVYNLVYN % “XOUddyY SAYNLINO VAS ONINAG G10 ‘HDSVMOAIW Woes SLOAFSO GAONIM AXOAI OZ “Id ‘1 (ON ‘96 “10A SNOILO31109 SNOANVT1S9SIW NVINOSHLIWS ((g6€ ased 3as ‘uoljeur[dxa 10,7) ‘SZIS IWUYNLYN % “XOUddY ‘07 3ALYV1I1d NO NMOHS SLOArSEO AO ASHAARaY le 1d ‘| ON ‘96 10A SNOIL941109 SNOANV11S9SIW NVINOSHLIWS SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL-96, NO: 1, PIES22 EXAMPLES OF EARLY PUNUK ART FROM MIYOWAGH. APPROX. 24 NATURAL SIZE. (For explanation, see pages 398-399.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO. 1, PL. 23 IVORY HARPOON HEADS FROM HILLSIDE SITE, OLD BERING SEA CULTURE. y APPROX. 14 NATURAL SIZE. (For explanation, see page 399.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO. 1, PL. 24 14 : eat 23 HARPOON HEADS FROM MIYOWAGH, OLD BERING SEA AND EARLY PUNUK CULTURES. APPROX. 14 NATURAL SIZE. (For explanation, see page 399.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO. 1, PL. 25 REVERSE OF HARPOON HEADS SHOWN ON PLATE 24. APPROX. 14 NATURAL SIZE. (For explanation, see page 400.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO. 1, PL. 26 16 19 20 IVORY HARPOON HEADS FROM MIYOWAGH, OLD BERING SEA AND EARLY PUNUK CULTURES. APPROX. 1, NATURAL SIZE. (For explanation, see page 400.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO. 1, PL. 27 HARPOON HEADS AND SOCKET PIECE, ST. LAWRENCE AND LITTLE DIOMEDE ISLANDS. APPROX. 14 NATURAL SIZE. (Por explanatien, see page 400.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO. 1, PL. 28 HARPOON HEADS OF EARLY PUNUK TYPE FROM MIYOWAGH. APPROX. 14 NATURAL SIZE. (For explanation, see page 401.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO. 1, PL. 29 aul 12 mm IMPLEMENT TYPES OF OLD BERING SEA CULTURE, FROM THE HILLSIDE SITE. APPROX. 12 NATURAL SIZE. (For explanation, see page 401.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO. 1, PL. 30 24 28 29 IMPLEMENT TYPES OF OLD BERING SEA CULTURE, FROM THE HILLSIDE SITE. APPROX. 2/5 NATURAL SIZE. (For explanation, see page 402.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO. 1, PL. 31 16 +7 HARPOON PARTS, OLD BERING SEA CULTURE, FROM MIYOWAGH. APPROX. 14 NATURAL SIZE. (For explanation, see pages 402-403.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO. 1, PL. 32 HARPOON PARTS, OLD BERING SEA CULTURE, FROM MIYOWAGH. APPROX. 14 NATURAL SIZE. (For explanation, see page 403.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO. 1, PL. 33 43 20 DART POINTS, ETC., OLD BERING SEA AND PUNUK CULTURES, FROM MIYOWAGH. APPROX. lg NATURAL SIZE. (For explanation, see pages 403-404.) 30 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO. 1, PL. 34 BONE AND IVORY ARROWHEADS, OLD BERING SEA AND PUNUK CULTURES, FROM MIYOWAGH, APPROX. 12 NATURAL SIZE. (For explanation, see page 404.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO. 1, PL. 35 13 14 MEAT HOOKS AND WOUND PLUGS. OLD BERING SEA AND PUNUK CULTURES, FROM MIYOWAGH, APPROX. ly NATURAL SIZE. (For explanation, see page 404.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO. 1, PL. 36 ceil hee 22: Pe eae rr FS? FISHLINE SINKERS, OLD BERING SEA AND PUNUK CULTURES, FROM MIYOWAGH. APPROX. 12 NATURAL SIZE. (For explanation, see page 405.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO. 1, PL. 37 THROWING BOARDS, ICE CREEPERS, ETC., OLD BERING SEA CULTURE, FROM MIYOWAGH. APPROX. 14 NATURAL SIZE. (For explanation, see page 405.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO. 1, PL. 38 8 9 10 4 12 13 14 KNIVES AND KNIFE SHARPENERS, OLD BERING SEA AND PUNUK CULTURES, MIYOWAGH (For explanation, see pages 405-406.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO. 1, PL. 39 e4 IMPLEMENTS OF RUBBED SLATE FROM HILLSIDE SITE, OLD BERING SEA CULTURE. APPROX. 14 NATURAL SIZE. (For explanation, see page 406.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO. 1, PL. 40 IMPLEMENTS OF CHIPPED STONE FROM THE HILLSIDE SITE, OLD BERING SEA CULTURE. 4%, NATURAL SIZE. (For explanation, see pages 406-407.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO. 1, PL. 41 31 DRILL POINTS, GRAVERS, AND SIDE SCRAPERS FROM HILLSIDE SITE, OLD BERING SEA CULTURE. APPROX. 14 NATURAL SIZE. (For explanation, see page 407.) SMITHSONIAN MISCELLANEOUS [COLLECTIONS VOL. 96, NO. 1, PL. 42 SIDE AND END SCRAPERS, RUBBING TOOLS, ETC., FROM HILLSIDE SITE, OLD BERING SEA CULTURE. 14 NATURAL SIZE. (For explanation, see page 408.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO. 1, PL RUBBING STONES, WHETSTONES, ETC., FROM HILLSIDE SITE, OLD BERING SEA CULTURE. APPROX. 13 NATURAL SIZE. (For explanation, see page 408.) 43 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO. 1, PL. 44 IVORY SLEDGE RUNNERS FROM HILLSIDE SITE, OLD BERING SEA CULTURE. APPROX. 4 NATURAL SIZE. (For explanation, see page 408.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO. 1, PL IVORY SLEDGE RUNNERS AND SHOES, MIYOWAGH, OLD BERING SEA CULTURE (For explanation, see page 409.) 45 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO. 1, PL. 46 IMPLEMENT TYPES OF THE OLD BERING SEA CULTURE, MIYOWAGH (For explanation, see page 409.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO. 1, PL. 47 IMPLEMENT TYPES OF OLD BERING SEA CULTURE, MIYOWAGH, APPROX. 3 NATURAL SIZE. (lor explanation, see page 409.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO. 1, PL. 48 IMPLEMENT TYPES OF OLD BERING SEA CULTURE, MIYOWAGH (For explanation, see page 410.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO. 1, PL. 49 IVORY PICKS, OLD BERING SEA CULTURE. HILLSIDE SITE APPROX. 13 NATURAL SIZE. (For explanaticn, see page 410.) By SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96. NO. 1, PL. 50 IMPLEMENT TYPES, OLD BERING SEA CULTURE, MIYOWAGH (For explanation, see page 410.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO. 1, PL. 51 IMPLEMENT TYPES, OLD BERING SEA CULTURE, MIYOWAGH For explanation, see page 411.) I pas | SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO. 1, PL. 52 POTSHERDS, OLD BERING SEA CULTURE, HILLSIDE SITE. APPROX. 43 NATURAL SIZE. (For explanation, see page 411.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO. 1, PL. 53 1. BALEEN VESSEL, OLD BERING SEA CULTURE, HILLSIDE SITE. 3, NATURAL SIZE. 2. POTTERY LAMP, KOWIERUK, SEWARD PENINSULA. (For explanation, see page 411.) 3 NATURAL SIZE. SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO. 1, PL. 54 WOODEN BASES FOR BALEEN VESSELS, ETC., OLD BERING SEA CULTURE, HILLSIDE SITE. APPROX. 43 NATURAL SIZE. (For explanation, see pages 411-412) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO. 1, PL. 55 IMPLEMENT TYPES OF OLD BERING SEA CULTURE, MIYOWAGH (Por explanation, see page 412.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO. 1, PL. 56 tin. 3 ¥ "37009 BALEEN OBJECTS, ETC., OLD BERING SEA CULTURE, MIYOWAGH (For explanation, see page 412.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO. 1, PL. 57 WOODEN OBJECTS AND FRAGMENTS, OLD BERING SEA CULTURE, MIYOWAGH (For explanation, see pages 412-413.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO. 1, PL. 58 SNOW GOGGLES, BUTTONS, BROW BANDS, ETC., OLD BERING SEA CULTURE, MIYOWAGH (For explanation, see page 413.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO. 1, PL. 59 418 i9 20 TOYS, OLD BERING SEA CULTURE, MIYOWAGH. APPROX. 12 NATURAL SIZE. (For explanation, see pages 413-414.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO. 1, PL. 60 IMPLEMENT TYPES OF EARLY PUNUK CULTURE, MIYOWAGH (For explanation, see page 414.) (‘Fit ased 99s ‘uolnvuridxsa 10,7) TISENVD COAIHDSOAR! LY SNOILVAVOXS 19 “1d “| ON ‘96 “10A SNOILD31109 SNOANVT1IS90SIW NVINOSHLINS eee c9 “1d ‘1 ‘ON ‘96 “10A SNOILD31109 SNOANW1139SIW NVINOSHLIWS (Si a8ed 99s ‘uoljeurjdxo 10,7) TW1ISEWNVD “LSE9DVAHSDVYMOMWMAS LV SNOILVAVDXA €9 1d ‘1 ‘ON ‘96 110A SNOILD31109 SNOANVT1S9SIW NVINOSHLIWNS SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO. 1, PL. 64 HOUSE RUINS AT OLD SECTION OF GAMBELL, AND AT NASKOK AND MIRRUKTA, EAST OF GAMBELL (For explanation, see page 415.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO. 1, PL DECORATED OBJECTS FROM IEVOGHIYOQ, PUNUK CULTURE. APPROX. 3/5 NATURAL SIZE. (For explanation, see page 415.) a5h5}3) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO. 1, PL. 66 DECORATED OBJECTS FROM IEVOGHIYOQ, PUNUK CULTURE. APPROX. 3/5 NATURAL. SIZE. (For explanation, see page 415.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOESS6y NON IPE DECORATED OBJECTS FROM SEPLOWAGHYAGET AND OLD SECTION OF GAMBELL, PUNUK CULTURE. APPROX. 3/5 NATURAL SIZE. (For explanation, see pages 415-416.) 67 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO. 1, PL. 68 IVORY OBJECTS FROM IEVOGHIYOQ AND SEKLOWAGHYAGET, PUNUK CULTURE. APPROX. 12 NATURAL SIZE. (For explanation, see page 416.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO. 1, PL. 69 ey Wot ae SS REVERSE OF THE OBJECTS SHOWN ON PLATE 68 (For explanation, see page 416.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO. 1, PL. 70 19 HARPOON HEADS FROM IEVOGHIYOQ, PUNUK CULTURE. APPROX. 14 NATURAL SIZE. (For explanation, see page 416.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO. 1, PL. 71 HARPOON HEADS FROM SEKLOWAGHYAGET AND THE OLD SECTION OF GAMBELL, PUNUK CULTURE AND MODERN. APPROX. 1, NATURAL SIZE. (For explanation, see page 417.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO. 1, PL. 72 v4 8 WHALING HARPOON HEADS FROM THE PUNUK SITES. APPROX. 12 NATURAL SIZE. (For explanation, see pages 417-418.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO. 1, PL. 73 IMPLEMENT TYPES, PUNUK AND MODERN. APPROX. 14 NATURAL SIZE. (For explanation, see page 418.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO. 1, PL. 74 10 IMPLEMENT TYPES, PUNUK AND MODERN. APPROX. 14 NATURAL SIZE. (For explanation, see pages 418-419.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO. 1, PL. 75 IMPLEMENT TYPES, PUNUK AND MODERN (For explanation, see pages 419-420.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO. 1, PL. 76 BOLA WEIGHTS AND PIECES OF PLATE ARMOR, PUNUK CULTURE. APPROX. 2/5 NATURAL SIZE. (For explanation, see page 420.) Pods Lr VOL. 96, NO. SMITHSONIAN MISCELLANEOUS COLLECTIONS SLEDGE RUNNERS OF THE PUNUK CULTURE 20.) 0-4 > (For explanation, see pages 4 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO. 1, PL. 78 IMPLEMENT TYPES OF THE PUNUK CULTURE (For explanation, see page 421.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO: 1, PL. 79 IMPLEMENT TYPES OF THE PUNUK CULTURE (For explanation, see page 421.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO. 1, PL IMPLEMENT TYPES OF THE PUNUK CULTURE (For explanation, see page 422. . 80 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO. 1, PL. 81 16 17 18 19 IMPLEMENT TYPES OF THE PUNUK CULTURE (For explanation, see page 422.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO. 1, PL. 82 ORNAMENTS, ETC., OF THE PUNUK CULTURE. APPROX. ly NATURAL SIZE. (For explanation, see pages 422-423.) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO. 1, PL. 83 TOYS, PUNUK AND MODERN (For explanation, see pages 423-424.) (ber abud aos ‘uotjeurjdxo 1G.) ) “AZIS IVYUNLVYN 9/1 ‘2 (3ZIS IWYUNLYN “% *KOUddY ‘G-E ‘1 "NYSGOW GNV MANNd ‘HLYVAH BYIA GNV “3100Vd LVOG STISSS3SA AvX3aL10d v8 1d °t ON ‘96 10A SNOIL931109 SNOANVTIS9SIW NVINOSHLIWS INDEX A Adz blades, 154 Adzes, 78, 159-160, 179, 186, 233-234, 333-334 Ainu, 275-270, 206, 318, 327, 354 Ajagaq, 178 Aleutian Islands, 276-277, 289, 290, 316, 335, 343, 344, 345, 348, 372, 373, 374, 375, 376-378 Amur, 296, 298, 299, 300 Andersson, eG. 354 Animal remains seal noses, 66, 248 seal skull, 60, 248 See also Barnacles; Birds; Fish; Mammals; Mollusks Armor American Indian, 326-327, 332-333 Asiatic, 327 ff at Gambell, 189, 224-225 Eskimo, relationships of, 325 ff. Arrowheads, 135-137, 221-223, 323-324, 331, 305-306 blunt pointed, 324-325, 382 Arrow shafts, 137, 240 Art Aleutian Islands, 289, 2090 Asia, northeastern, 293, 296, 208 299, 300, 303 British Columbia and Washington, 201 China, 297, 298-301 Columbia River, 291 Cook Inlet, 290, 291 Dorset culture, 289, 293, 206 Eurasia, northern, 293, 294, 302- 303 Iron age, 294 Kodiak Island, 290 Maglemose, 294, 205 Megalithic, 294 Melanesian, 300, 301 modern Eskimo, 287-288 Neolithic, 294, 296 Northwest Coast, 292, 298-299, 300 Paleolithic, 294, 295, 206 pictographic, 294, 366 Thule culture, 289 Art, Old Bering Sea affinities of, 287 ff. at Hillside Site, 40-52 at Little Diomede Island, at Miyowagh, 76-92 design elements of, 46-47, 80-81, 82, 91-92 33 SRS) Art, Old Bering Sea—Cont’d distribution of, around Bering Strait, 286 distribution of, at Gambell, 202 Art, Punuk affinities of, 287-289, 291, 301-303 at levoghiyoq, Seklowaghyaget, and old section of Gambell, 192-202 at Miyowagh, 92-97 distribution of, around Bering Strait, 287 distribution of, at Gambell, 202 Awls, 164, 233 B Baleen objects fishing line, 172 ice scoop, 171, 240 knots, 171, 240 netting, 175 plaited, 172, 240 snare, 172 toy bows, 172 toys, etc., 179 ventilator rim, 171 vessels, 169-170, 239, 350 Barnacles, 250 Beechey, F. W., 283 Beothuk, 373 Bering, Vitus, 17, 18, 377 Billings, Joseph, 19 Birch bark, rolls of, 175 Bird bones excavated, 249-250, 251 bones, one inserted in another, 240 darts, 323, 382 end prongs, 132-133 side prongs, 131-132, 221 figures, ivory, 50, IQI, 212, 244, 306, 366 snares, 186 Birds, species of, on St. Lawrence, 15 Birket-Smith, Kaj, 1, 8-9, 265, 268, 277, 279, 281, 283, 284, 338, 340, 343-345, 347, 379-380 Birnirk culture, 103, 118, 307-308, 311- 312, 310, 322, 323,.324,, 325, 365-366, 372, 379 Bishop, C. W., 280 Boas, Franz, 1, 3, 4, 202, 352, 353, 367, 370-371 Boat hooks, 238 Bodkin, 195 Bogoras, W., 3, 270, 271, 285, 370 425 426 Bolas, 227-228, 325, 365, 372, 382 Bone dish, 179 Bone knife, 162, 235-236, 333 Bow braces, 223, 331 Bows, 134, 323 toy, 134, 172 Brandt, G. Herman, 25 British Columbia, 280, 286, 290, 201, 292, 338 Brow bands, ivory, 79, 177, 243 Brower, Charles, 264 Burials, human, 64-65, 76, 186, 188, 246-247 Buttons, ivory, 177 C Caribou Eskimos, 373, 379-380 Cassedy, E. G., iv Central regions (Canada), 1 ff., 305, 309, 314-315, 353 _Cernecov, V., 320, 343, 383 Chain, ivory, 43 Chambers, Moreau B., iii, 25, 60, 61, 66, 67, 68, 216 Chaney, Ralph W., 14 Check-stamped pottery, 160, 349-350 China, 7. 2098-301, 322, 327 ff., S37: 33 Chukchee, 270, 293, 318, 322, 323, 325, _ 326, 370-371 Clothing gut parka, 177 seal and bird skin, 64, 177 Coast eS: 6-7, 338, 344, 348, 381- 382 Collier, A. J., 14 Collins, Henry B., Jr., 267, 289, 301, 306, 309, 311, 320, 352, 365, 306, 368 Combs, 78, 175, 243-245 Commander Islands, 377 Cook, oe James, 18, 274, 276-277, 284 Cooking pots pottery, 341-342, 346-348 pottery, fragments of, 188, 190, 238 St. Lawrence Island, 341-342, 346- 167-168, 347 Thule culture, 347 GranzD eer Creel, Herlee G., 299 Culture elements distribution as age criterion, 370 modern Alaskan Eskimo, lacking in Old Bering Sea, 360 Old Bering Sea, listed, 356-358 Punuk, listed, 357-359 Thule, at Point Barrow, 364 Thule, on St. Lawrence Island, 362-364 Cylinders, wood, 174, 240 INDEX D IDRWIIS WW, 1H, S377 Dart socket pieces, 126, 195 Deshnev, Simeon, 16 Diomede Island, 11, 53-56, 286, 287, , 313, 346, 379 Disko Bay, 335-336, 355 Dog skulls, types of, 248 Dog traction, 242, 338 ff., 366 Dolls, 178, 244, 245 Dorset culture, 289, 293, 205, 206, 315- 316, 335, 336, 338, 348, 370, 372-373, 379 Doty, W. F., 24 Drill bows, 162, 237 mouthpieces, 162, 236 points, bone, 163, 237 points, stone, 149-150 rests, 162, 236 shafts, 163, 180, 237 Drills, hand, 163, 237 Drum handles, 54, 174, 239 rims, 66, 174 Drying rack, 174 EK Ear ornaments, 243, 305, 301 Egede, Hans, 1 Engraving tools, 173, 180, 237, 303- 304 Eskimo culture American origin of, 379-380 changes in, 380 Old World origin of, 381 theories regarding origin of, I-12 uniformity of, 370 Eskimo language, 4, 367 F Fat scrapers, ivory, 46, 80, 83-84, 164- 166, 233, 351 Finland, 322, 325 Finmark, 303, 320 inn Ds Jk 340 Fire hearth, wood, 186, 237 Fischer-Mgller, K., 368-360 Fish bones, excavated, 250 hook, 172, 225 lies 172 spears, points for, 133, 140, 225 Fishline sinkers, 140-143, 186, 226 Float bars, 130, 221 mouthpieces, 130, 220-221 plugs, 130 INDEX Folklore, 3, 367, 370-371, 377-378 Ford, James A., 111, 25, 64, 69-70, 183, 3II -312 Formosa, 337 Fraser River, 337 Friedmann, Herbert, iii, 240 G Gambell (Sevuokok) abandoned sites, 26-27, 31 description of, 31-33 excavations at, 35-36 old section of excavations at, 180-1091 harpoon heads, 212-215 houses, 189-191 Gambell, V. C., 2 Gest OtomWe 25,27, 30-31, 55, 243 Gilyak, 268, 275, 293, 206, 318, 327, 354 Glass bead, 188 Gordon, G. B., 257 Grass in house walls, 272, 273, 274 toy (?) wrapped with baleen, 172 Gravers, stone, 150 Greenland, 1 ff., 265-266, 268, 282, 305, 309, 314-315, 316, 322, 335-330, 352, 354, 355, 372, 380 H Hallowell, A..I., 381-382 Harpoon blades, 147-148 finger rests, 129, 219 foreshafts, 43-44, 124-126, 218 ice picks, 128-129, 218 rests, 238 shafts, 126, 219 socket pieces, 194, 218 Harpoon heads Ainu, 318 Bering Strait, 307, 308 Birnirk type, 307-308, 311-312, 310, 365 chart showing chronological de- velopment of, 216 coast culture, 382 Cook Inlet, 315-317 distribution of, according to site and depth, 218 Dorset culture, 315-316 drilled lashing holes on, 308-309 drilled rivet holes, 309-310 Europe, central and northern, 320 Japan, 318-319 Kolyma River, 307, 319 Kurile Islands, 318 Negritos, 319 48, 86-87, 126-128, 427 Harpoon heads—Cont’d Northwest Coast, 317 Nunivak Island, 316 Polar Eskimos, 314 significance of, in chronology, 97- 08 Thule culture, 310-315 toy, 178 whaling, 215-217 Ya-mal Peninsula, 320 Harpoon heads, Old Bering Sea cul- ture decoration of, at Hillside site, 51- 52 decoration of, at Miyowagh, 87 types described from Hillside site and Miyowagh, 97-115 Harpoon heads, Punuk culture decoration on, at Ievoghiyoq, 194- 195 decoration on, at Miyowagh, 93-94, 95 decoration on, at Seklowaghyaget and old section of Gambell, 197 types described, from Ievoghiyogq, 203-208 types described, from Miyowagh, II5-124 types described, from old section of Gambell, 212-215 types described, from Seklowagh- yaget, 209-212 Hatt, Gudmund, 6-7, 9, 285, 338, 340, 344, 352, 381-382 Henderson, E. Pa is Hightower, (Gs de Hillside site decorated objects from, 40-52 discovery of, 34-35, 36 excavations at, 37-40 house ruins at, 37-38, 39-40 Hitchcock, R., 276 Hoffman, W. Ae 204 Holm, e I Holmberg, H. J., 273, 277-278 Hooks dog humerus, 237-238 fish, 172 wooden, 173 Hooper, Capt. C. L., 22-23 Hough, Walter, 326-327, 328, 341, 346 Houses Ainu, 275-276 Aleutian Islands, 276-277 Bering Strait, 261- -204 Chukchee, 270 circular, 282, 283-284 coast culture, 382 discussion of, 256-286 entrance passage, 280-281 gabled roof, 281 Gilyak, 268, 275 428 Houses—Cont’d Greenland, 265, 266, 268, 282 Kamchadal, 274-275 kashims, 257, 272-273 Kodiak Island, 277-278 Koryak, 271-274 Kotzebue Sound, 269, 283 Mackenzie, 264, 265, 268-269 Metlatavik, 262-264 Norton Sound, 260 Northwest Coast, 278-280, 281 Nunivak Island, 257, 258-260 octagonal, 284 Point Barrow, 264, 265, 266-268, 269, 279, 281, 282, 366 Point Hope, 264 Prince William Sound, 278 roof entrance, 280-281 ruins of Hillside site, 37-38, 39-40 Ievoghiyog, 183-186 Mirrukta, 191-192 Miyowagh, 68, 69-76 old section of Gambell, 189-191 Punuk Island, 28 Seklowaghyaget, 187-188 Siberia, northeastern, 1-2 St. Lawrence Island, 260-261, 270, 273, 284, 285-286 Salish, 286 shape of, 282-285 Shishmareff, 264 Siberian Eskimo, 270, 283 Southwest Alaska, 257, 266, 260, 272-273, 274, 276, 284, 285 Thule, 266, 267, 268, 282-283, 286 Yukaghir, 271 Hrdlicka, A., iii, 12, 25, 26, 28, 54, 197, 240, 246, 330, 335, 344-345, 367-368 Human face, carving of, 43, 54 Human figures carvings, 43, 245 incised, 89 I Ice creepers, 134, 227, 322 Ice scoop, 171, 240 Ievoghiyoq, 33 description of, 181-182 excavations at, 181-186 houses at, 183-186 Imrey, Frank, iv Indo-China, 301, 337, 338 Inland culture, 6-7, 9, 325, 381-382 Iron pyrites, 243 i) Japan, 276, 318-319, 321, 328-320, 337, 348, 349, 375 INDEX Jenness, Diamond, I, I1-12, 25, 28, 20, 215, 267, 290, 297, 301, 307, 308, 313, 315-316, 335, 342, 348, 367, 372-373, 379 Jesup Expedition, 3, 13, 370, 377 Jochelson, W., 3, 271, 272, 275, 280, 289, 330, 348, 370, 373, 376, 377 Jones, Capt. E. D., iti, 96 Jumping stones, 26, 354-355 K Kachemak Bay culture, 290, 291, 305, 316-317, 335, 337, 344, 348, S 5 Kamchadal, 274-275, 277, 322, 370-371, 8 37 Kamchatka, 296, 318, 322, 345, 348, 375-378 Kayak, 253, 382 paddle, 186 toys, 159, 244, 245 Kellogg, Remington, iii, 64, 247, 248 Kialegak, Cape, 22, 25, 29-31, 240, 250 Kidder, A. V., 96-97 Kirk, E. W., 349 Kitneapalok, 26, 216 Knife handles, wood, 143-145, 232 handles, bone and ivory, 145-146, 231-232 sharpeners, 174, 232, 333 Knives, 140, 333 crooked, 333, 382 ivory, 192, 235 Kodiak Island, 240, 277, 290, 331, 335, 344-345, 348, 376 Kola Peninsula, 303, 320, 350, 354 Koryak, 271-274, 293, 318, 322, 325, 326, 370-371, 378 Kotzebue, Otto von, 19-22, 264 Kowieruk pottery lamp from, 238, 342 Krasheninnikoff, S., 274 Kroeber, A. L., 7-8 Kukuliak, 25, 27, 243 Kurile Islands, 78, 276, 206, 318, 337, 345, 348, 354, 376 i Labrets, 281, 305, 376 Ladles, 166 de Laguna, Frederica, 278, 292, 2093, 204-295, 302-303, 316, 330, 334, 335, 344, 345, 373-375, 376 Lamps, pottery distribution and relationships of, 343 ff. Ertebglle culture, 343 INDEX Lamps, pottery—Cont’d fragments of, 166-168, 188, 190, 238 Kowieruk, 342 St. Lawrence Island, 341-342, 346 Wales, 342 Yukon-Kuskokwim, 342 Lamps, stone Aleutian Islands, 343, 344, 345, 348 ; Caribou Eskimo, 343 distribution and relationships of, 343 I. Kamchadal, 344, 345 Kurile Islands, 345 North Alaska, 267, 366 South Alaska, 345 Thule culture, 346 Lance foreshafts, 126 heads, 217 Langsdorff, G. H. von, 276, 278 Lapps, 203, 354 Lashing holes on harpoon heads, 215, 308-300, 366 Laufer, Berthold, 3, 30, 304, 326, 327, z 328-330, 331 Link ornaments, ivory, 192, 241, 305 Lisiansky, U., 277, 278 M Mackenzie house type, 264, 265, 268- 269 Mammals bones of, excavated, 247-240 species on St. Lawrence Island, 15 Manca, Harry E., 25 Manchuria, 331, 337, 338, 354 Markham, C. H., 1-2 Maskin, Phillip, iii Masks, absence of, 281, 305 Mason, J. Alden, 338, 342, 365, 368, 372 Mathiassen, Therkel, 1, 8, 9-10, 99, 256, 265-266, 267, 268, 270, 277, 282-283, 284, 290, 308, 311, 314, 315, 336, 338, 340, 341, 343, 346, 347, 348, 351, 352, 361 ff., 373, 381 Mattock, 161, 235 Meat hooks, 138-140, 227 Metal from House no. 10, 191, 212 knife blades of, 146 points on engraving tools, 180, 237 possible early source of, 304-305 prehistoric use of, on St. Lawrence Island, 29-30, 303-304 Metlatavik house type, 262-264, 265 Miller, Gerrit S., Jr., 248 Mirrukta, 27, 191-192 Missugameet, 27 429 Miyowagh, 31 description of, 56-58 excavations at, 58-76 houses at, 69-77 Mollusks, 250 Mongolia, 331 Moore, Riley D., 24, 108, 214, 246 Miller, G. F., 16, 17-18 Murdoch, John, 1, 2, 138, 264, 265 Mussel shells, 59 N Needle cases, 45, 90, 192, 194, 351-354, 305, 382 Needles, 164, 233 Nelson, E. W., 1, 24, 126, 238, 257, 200, 282, 283, 341 Neo-Eskimo culture, 5-6, 379 Net sinkers, 226-227 Netting, baleen, 175 New England, 337, 338 Nordenskidld, A. E., 1, 270-271 Northwest Coast, 12-13, 278-281, 201- 293, 298-299, 300, 317, 332, 348, 360, 370, 378, 381 Nunivak Island, 243, 257, 258-260, 349 O Old Bering Sea culture art of, at Hillside site, 40-52 art of, at Little Diomede Island, oS aos ee art of, at Miyowagh, 76-92 clothing and ornaments, 177-178 design elements, 46-47, 80-81, 82, QI-92 discovery, II-13 elements, 356-358 evidences of, at Punuk Island and Cape Kialegak, 28, 29, 30 general character of, 253-254, 350, 360 harpoon heads, 97-115 houses, 37-38, 39-40, 73-76, 260 implement types, 124-176 relationships of, 361 ff., toys, 178 Ornaments, 177, 192, 242-243, 305-306 Ostiak, 303, 325 lee Pail handles, 78, 86 Palae-Eskimo culture, 5-6, 379 Pendants bear teeth, 242 dog teeth, 242 ivory, 242, 305, 365, 372 Petitot, E., 265 430 Petroff, Ivan, 257, 278 Physical type, 246-247, 367-369 Picks, 160-161, 234-235 Pilsbry, H. A., 1v, 250 Point Barrow, 264, 265, 266-268, 4 ff. Polar bear, ivory, 49, 178 Porsild, Morten P., 355 Porter, Robert P., 276 Pottery, 166-160, 188, 190, 238-239, 320, 341 ff., 376, 382 Pre-Eskimo remains, absence of at Bering Strait, 378 Punuk culture art of, at Ievoghiyog, Seklowagh- yaget, and old section of Gambell, 192-202 art of, at Miyowagh, 92-97 discovery of, 28-29 early use of metal, 29-30, 303-305 elements, 357-359, 360 general character of, 254-255 harpoon heads, 115-124, 203-208, 209-215 houses, 70-73, 76, 183-186, 260-261 implement types, 179-181, 217-246 relationships of, 361 ff. Punuk Island, 18, 22, 25, 27-31, 55-56 Pyrites, 243, 361 R Rasmussen, Knud, 1, 8, 308, 317 Rainey, Froelich G., 27, 31 Ratzel, F., 328 Reamers, 163, 237 Red pigment in incisions On ivory, 196, 200, 206, 207, 200, 212, 287 on boat paddle, 158 on wooden objects, 174, 237, 240 Rehder, Harald A., iv, 250 Reid, E. D., iv, 250 Rink Elie Rivet holes on harpoon heads, 215, 309-310, 366 Rubbing stones, 154-155 Rubbing tools, bone, 163, 237 S Sagoskin, L., 257-260 St. Lawrence Island birds, 15 description of, 13, 19-24 discovery of, 17 mammals, I5 vegetation, 15 Salmon spears, 225 Samoyed, 325, 354 Sarfert, E., 284 INDEX Sauer, Martin, 276, 278 Savunga, 27 Scandinavia, 294, 302, 303, 317, 320, _ 322, 328, 337 Scapulimancy, absence of, 248 Schmidt, A. V., 303 Schmitt, Waldo L., iv Schrenck, L. von, 275, 276 Scrapers bone, 166, 382 stone, 150-154 Seal importance of, 248 noses, 66, 248 skull, 60, 248 Sealing scratcher, 219, 322, 365, 372 Seklowaghyaget, 31 description of, 186-187 excavations at, 186-189 house at, 187-188 Sequoia, fossils of, 14 Sevuokok, see Gambell Shishmareff, Lt. G. S., 22 Shovels, wooden, 173 Shuttles, 175 Silook, Paul, iii, 64, 67 Simpson, John, 264, 268, 269 Sinew-backed bow, 323, 382 Sinew twisters, 223, 331 Sledges, 155, 228-230, 338-341, 382 cross bars, 15 runners, 155-157, 228-231, 339-340 shoes, 157-158, 340 toy, 158, 244 Smith, Harlan I., 3, 201 Snow goggles, 178, 180, 244 Snow shovels, 161-162, 180, 235 Solberg, O., 303, 336 Sollas, W. J., 204 Soonogoruk, Moses, ili Spindle buzzes, 176, 196, 245 Spoons, 233 Steensby, H. P., 1, 4-6, 265, 266, 268, 269, 282, 284 Stefansson, V., I, 311, 323, 342 Steller, G. W., 274, 278 Sternberg, L., 275 Stone flakers, 163, 237 Stone implements, 147-155, 232, 334- 338, 382 Su-shén, 304, 320, 331 Sverdrup, Hy Us, 2737s to Swanton, John R., 3, 327 Synd, Lieutenant, 18 Ab Tattooing on ivory doll, 245 Teeth dog, perforated, 243 polar bear, perforated, 240, 243 Tet y).3 INDEX 431 Textile, 60, 245-246 V Thalbitzer, William, I, 4, 257, 265, 269, 282, 352, 354, 355 V li Rae, 64, 365 Thimble guard, 241 ae ae ee SET BNF 1 30300) Thordeman, Bengt, 328 Vogul, 303, 325 Throwing boards, 133, 219-220, 322- eg 323, 382 spur for end of, 133-134, 220 W Thule culture, 9-11, 140, 148, 162, 204- A, Abies, Ais Phiy Arie, ey B25ue23 230), 2405 (241) 2600; 267, 268, 282-283, 286, 280, 306, 309, 310-316, 321, 322, 324, 325, 333, 334, 336, 341, 346, 347, 348, 351, 356 elements, at Point Barrow, 364 elements, on St. Lawrence Island, 362-364 origin of, 372, 379 physical type associated with, 367- 309 relationships of, 315-317, 364-366 return movement of, to Alaska, 267, 306, 324, 334, 345, 366, 369-371, 377 Toboggans, 509, 158, 240, 338 ff., 382 Tops, 178, 245 Torii, R., 276 Toys, 134, 158, 159, 172, 178-179, 180, IQI, 244-245, 305-306 Trace buckles, for dog harness, 242 Tube, bone, 240 Tungus, 293, 325, 354 1G} Ulu blades, 149 Ulus, 44, 94, 164, 195, 232-233, 350- a5 Umiaks, 253, 382 paddle and keel, 158 toy, 159 Wales, Cape Prince of, 11, 261-262, 286, 287, 308, 313, 342, 375, 379 Walrus importance of, 248 ivory carving of, 86, 91 Waterman, T. T., 278, 280, 283-284 Webber, John, 277, 284 Wedges, I61, 235 Weltfish, Gene, 245 Whales, figures of, 178, 244 Whetstones, 154-155 Wilson, J. W., iv Winged objects developmental stages of, 200-201 from Hillside site, 41-43 from levoghiyoq and Seklowagh- yaget, 199-200 from Miyowagh, 87-89 possible use of, 200-201 toy, 178 Wissler, Clark, 300, 324 Wooden vessels, 170, 239 Wound plugs, 138, 227 Wrangell, F. von, 1, 270 Wrist guards, 95-96, 193, 195, 224, 325, 331, 372, 383 AY Yakut, 203 Ya-mal Peninsula, 320, 343, 382-383 Yukaghir, 271, 293 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOLUME 96 NUMBER 2 WAVE LENGTHS OF RADIATION IN THE VISIBLE SPECTRUM PROMOTING THE GERMINATION OF LIGHT-SENSITIVE LETTUCE SEED (WitTH ONE PLATE) BY LEWIS H. FLINT Boyce Thompson Institute for Plant Research, Inc. AND E. D. McALISTER Division of Radiation and Organisms, Smithsonian Institution (PUBLICATION 3414) . GITY OF WASHINGTON PUBLISHED BY THE SMITHSONIAN INSTITUTION JUNE 16, 1937 j ‘ ; NGA, ‘ in ‘ . a j 4 var, i * fy Te ee es pe os 1 K } ( | THe Lord Baltimore Press Z BALTIMORE, MD., U. 8. A. WAVE LENGTHS OF RADIATION IN THE VISIBLE SPECTRUM PROMOTING THE GERMINATION OF LIGHT=SENSITIVE LEPTUCE SEED By LEWIS H. FLINT,’ Boyce Thompson Institute for Plant Research, Inc. AND E. D. McALISTER, Division of Radiation and Organisms, Smithsonian Institution (WitH ONE PLATE) INTRODUCTION The purpose of this investigation is to fix definitely the wave lengths of radiation which maximally promote the germination of light-sensi- tive lettuce seed * and to study their relationships to the wave lengths of radiation most effective in other fundamental plant phenomena. In a previous publication (Flint and McAlister, 1935) the wave lengths of radiation in the visible spectrum inhibiting the germination of light-sensitive lettuce seed were reported as occurring in two re- gions: one in the violet-blue-green region (with principal maxima at about 4400 and 4800 angstroms), and the other in the red-to-near- infrared region (with one maximum, at about 7600 angstroms). The earlier work of Flint (1934, a, b) had indicated the general range of the radiation promoting germination as 5200 to 7000 A, but the curve of relative effectiveness of the radiation within this range was unknown. The accompanying graph (fig. 1), reproduced from a re- cent paper by Flint and McAlister (1936), illustrates the state of the study of radiation in relation to the germination of this particular seed at the beginning of the researches here reported for the same material. It may be noted from figure 1 that at the intensities of radiation applied in the region 5200-7000 A a germination of 100 percent was secured at the longer wave lengths. It follows that differences in the * Formerly of the Division of Seed Investigations, Bureau of Plant Industry, U. S. Department of Agriculture. * Light-sensitive Arlington Fancy lettuce seed was used throughout the co- operative researches. SMITHSONIAN MISCELLANEOUS COLLECTIONS, VOL. 96, No. 2 2 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 effectiveness of one part of this region over another were masked by the completeness of the germination. In the study of the effectiveness of radiation in the violet-blue-green region it had been found that the principal critical wave lengths in the regions 4400 A and 4800 A were identical (within the range of experimental error) with those reported by Johnston (1934) and others for the phototropic response of oat coleoptiles. This analogy suggested that similar pigments were involved in the absorption in both instances, and that the results thus might well be of significance beyond their immediate application to the problem of dormancy in a particular kind of seed. This possibility has been made more prob- 100 wu (@) TInhibition<—>Promotion ce) 4400 4800 5200 7000 8000 Fic. 1—The action of radiation of specific wave-lengths in relation to the germination of light-sensitive lettuce seed. The percentages of germination in the spectrum (immediately following an exposure to red light sufficient to effect 50% germination) are indicated as ordinates, the wave-lengths of the spectral light being indicated as abscissae. able through the recent work of Burkholder and Pratt (see Burk- holder, 1936, p. 46), who found not only that radiation in the violet- blue region influenced the movement of Mimosa, but also that a band of radiation in the 7600 A region had a similar effect. In view of the foregoing the study of the relative effectiveness of the wave lengths of visible radiation promoting the germination of light-sensitive lettuce seed seemed to warrant the continuation of this work, which led to the results here reported. REVIEW OF LITERATURE Cieslar (1883) reported that yellow light favored the germination of certain seeds, whereas violet light retarded it. In later years the phraseology became more precise, with “* yellow to red” or “ yellow, orange, and red ” light as the promoting radiation, yet neither the wave NO. 2 IRRADIATED LETTUCE SEED—FLINT AND McALISTER 3 length range nor the relative effectiveness of bands of radiation within the range were established. Flint (1934) reported that the upper limit of the radiation inhibiting germination was at about 5200 A, while the upper limit of the seemingly analogous phototropic response was set as about 5400 A by Johnston (1934). The curves of the absorption spectra of carotin and xanthophyll as reported by McNicholas (1931) reach low points in the region 5000 A to 5500 A, with principal criti- cal wave lengths of maximum absorption at about 4400 A and 4800 A (when allowance is made for the shift in wave length associated with the solvent). These correspond closely with the regions of maximum effectiveness of this radiation in inhibiting germination and inducing phototropism. The curves of the absorption spectra of the chlorophylls as given by Zscheile (1934) are uniformly low in the green region and Shuck (1935) has held that there was a band of this radiation which had no effect on germination. On the other hand the curve of the action of radiation of specific wave lengths in relation to the germi- nation of light-sensitive lettuce seed as developed in the work of Flint and McAlister (see fig. 1) indicates an immediate transition of reac- tion in this region, though not as abrupt a change as that taking place at the long wave length end of the promoting influence at about 7000 A. Kommerell (1927), in studying quality of light and germination, had concluded that at any specified wave length the effectiveness of the radiation was proportional to the energy falling upon the surface of the seed. The work of Flint and McAlister, on the other hand, indicated that the effectiveness of the radiation was proportional to its absorption by the seed pigments. The investigations of Meischke (1936) support the latter viewpoint. METHODS The procedure and apparatus used in this investigation was essen- tially that described in detail in a previous paper (Flint and McAlister, 1935) except that blue light of controlled intensity was superimposed uniformly over the spectral region in which the seeds were being treated. By the well-known system of trial and error, using constant continuous spectral illumination, the blue light was so regulated in intensity that the maximum germination throughout the set of com- partments in 24 hours was precisely 100 percent in some one com- partment. With compartments 0.4 inch wide as used in the experiments previously reported, the conditions were then such that any increase in the intensity of the blue radiation gave less than 100 percent ger- mination in this compartment. All experiments reported were at 4 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 room temperature less than 29° C., for at higher temperatures the Arlington Fancy lettuce seed used in these experiments would not germinate, irrespective of the light conditions. On account of the narrow range of wave lengths represented by the critical maxima obtained, no correction for relative energy was applied to the graphic representation of the data. In the data as given in table 1, it may be noted that there was a more or less progressive falling off in the peak germination and that there was some variation between the different series. Under the conditions of the experiments, with continuous operation of the spectral and superimposed light sources, a differential reduction in the efficiencies of the lights was to be expected. RESULTS A large number of experiments were carried out incidental to the attainment of the prescribed conditions. Under conditions such that an additional intensity of blue, when superimposed upon the spectral light, would reduce the percentage germination in the compartment yielding 100 percent germination, the following results in table 1 were obtained : In conjunction with the results obtained as given in table 1, an examination of the absorption spectra of lettuce seed extract in acetone was undertaken. The results of this study are given in plate 1. The foregoing results establish the critical wave lengths of the radia- tion promoting the germination of the seed used in these trials and appear to indicate the nature of the material responsible for the reac- tion. The relative effectiveness of radiation with wave lengths within TaBLe 1.—Percentage germination of light-sensitive lettuce seed in spectral green to infrared radiation with superimposed blue light. Compartment width 0.4 inch. Percentage germination at approximate wave lengths at center Series of compartments 4950 5080 5180 5300 5440 5650 5710 5850 6000 6210 6400 6700 7050 7500 8200 A 0 0: 0 70) 10 —0) 50) 0) 40; 45:0) 27:3) 1000 GAO mmnee 0. B oo OO © © O © O O 307, 7353 e777. 0 C OO OO. © O BW © OW @ 0 S502) 304. ) D VO @ © © HO © WO @ Oo 0) 80.0 Oo (0) (0) E YT oO © O@® O W © O oO 65.0 oO (0) 0) Average’! 0:0 “0)) 0) 0) (0) oF 40) 0) e184) TRO 1807ens2 oso emo NO. 2 IRRADIATED LETTUCE SEED—FLINT AND McALISTER 5 the 5000-6000 A range has not been established in these trials, but in its entirety this radiation is obviously far less effective than that in the 6700 A region. DISCUSSION The results given in table 1 have been averaged and represented graphically as the full line curve in figure 2. It is obvious from this diagram that the radiation most effective in promoting the germination of the seed had a wave length of the order of 6700 A. The nature and position of this curve invites comparison 100 100 75 75 50 50 25 PERCENT GERMINAT/ON ABSORPTION COEFFICIENT +++++++« WAVE LENGTH /N ANGSTROMS Fic. 2—Curves of seed response and chlorophyll absorption. with the curve of the absorption spectrum of the chlorophylls in this region, and the analogy is impressive. Here is an instance, then, in which, within the limits of experimental error, the apparent critical wave length of radiation promoting germination coincides with the major absorption in this region by a pigment common to all green plants. The approximate order of correspondence is indicated in fig- ure 2. In Zscheile’s work the leaf pigments were extracted with ether. In the germination study the seed pigments were acting in their natural environment, which was an aqueous solution. Under these circum- stances the wave length shift apparent in figure 2 is about the amount to be expected. In plate 1 the upper spectrogram is taken from Willstatter and Stoll and represents the absorption of an acetone extract of leaf pigments. 6 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 The center spectrogram represents the absorption of an acetone extract of seed pigments. The lower spectrogram represents the emission of the mercury arc, and provides a wave-length scale for the center spec- trogram. It is apparent from this plate that the pigments absorbing the longer wave lengths of visible radiation in the seed are similar to those pigments found in the leaf by Willstatter and Stoll. Since the leaf pigments absorbing the longer wave lengths of visible radiation are for the most part identified as chlorophyll, it follows from the data represented in figure 2 and plate 1 that (within the limits of experimental error) the radiation most effective in promoting germination in the seed is that most effectively absorbed by chlorophyll in the same region. Unfortunately, these experiments were terminated before it was possible to make a positive identification of the pigments responsible for the relatively intense absorption for wave lengths of radiation shorter than 5400 A that is apparent on plate 1. It is certain that this is not due to the blue absorption of chlorophyll alone, for it is too intense compared to the red absorption and also extends too far toward the longer wave lengths. Thus, the presence of a yellow pig- ment is indicated, and the close agreement between the critical wave lengths of radiation inhibiting germination and those absorbed by the carotenes is impressive. However, a more extensive study of the ab- sorption spectrum of the yellow pigment found in these seeds is neces- sary before more can be said. The close analogy between the critical wave lengths of radiation influencing seed germination and the critical wave lengths of radiation absorbed by plant pigments places a distinct emphasis upon what is perhaps a new and promising viewpoint. In the seeds the violet-blue light promotes a set of physiological responses quite different from the set which is promoted by orange-red light. It follows that in green plants the physiological response to violet-blue light may be quite dif- ferent from the response to orange-red light. If such is the case, the study of quality of light offers increasing promise in relation to the problems of photosynthesis, photoperiodism, seed maturation, seed germination, plant distribution, and so on. In conjunction with an increasing interest in qualitative light effects the results here reported obviously emphasize the desirability of a more adequate knowledge of the quality of light at the earth’s surface and of the modification of that quality with time of day and year, with latitude, with altitude, and with water vapor in the earth’s atmosphere. NO. 2 IRRADIATED LETTUCE SEED—FLINT AND McALISTER Hf SUMMARY Radiation of wave lengths ranging from about 5200 A to about 7000 A characterizing the colors yellow, orange, and red, promotes the germination of light-sensitive lettuce seed; but within this range the longer wave lengths are by far the most effective. The critical wave length of the radiation promoting germination within the most effective range is approximately 6700 A. Within the limits of experimental error the radiation most effective in promoting germination is that most abundantly absorbed by chloro- phyll in the same region. The absorption characteristics of an acetone extract demonstrate the presence of chlorophyll in the seed. LITERATURE CITED BURKHOLDER, PAUL R. 1936. The role of light in the life of plants. Bot. Rev., no. 1, pp. 1-52, and no. 3, pp. 97-168. CIESLAR, ADOLPH 1883. Untersuchungen tiber den Einfluss des Lichtes auf die Keimung der Samen, pp. 270-205. Fuint, Lewis H. 1934a. Light in relation to dormancy and germination in lettuce seed. Science, vol. 80, pp. 38-40. 1934 b. Light-sensitivity in relation to dormancy in lettuce seed. Compt. Rend. Assoc. Internat. Ess. Sem., Copenhagen. Fiint, Lewis H., and McAttster, E. D. 1935. Wave lengths of radiation in the visible spectrum inhibiting the germination of light-sensitive lettuce seed. Smithsonian Misc. Coll., vol. 94, no. 5, pp. I-IT. 1936. The action of radiation of specific wave-lengths in relation to the germination of light-sensitive lettuce seed. Compt. Rend. Assoc. Internat. Ess. Sem., Copenhagen. JOHNSTON, Eart S. 1934. Phototropic sensitivity in relation to wave length. Smithsonian Misc. Coll., vol. 92, no. 11, pp. I-17. KoMMERELL, E. 1927. Quantitative Versuche tiber den Einfluss des Lichtes verschiedener Wellenlangen auf die Keimung von Samen. Jahrb. Wiss. Bot., vol. 66, pp. 461-512. McNicuoras, H. J. 1931. The visible and ultra-violet absorption spectra of carotin and xantho- phyll and the changes accompanying oxidation. Bur. Standards Journ. Res., vol. 7, pp. 171-193. MerscHkeE, D. 1936. Ueber den Einfluss der Strahlung auf Licht- und Dunkel-keimer. Jahrb. Wiss. Bot., vol. 83, pp. 359-405. 8 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 Suuck, A. L. 1935. Light as a factor influencing the dormancy of lettuce seeds. Plant Phys., vol. 10, pp. 193-196. ZSCHEILE, F. PAUL, Jr. 1934. An improved method for the purification of chlorophylls A and B; quantitative measurement of their absorption spectra; evidence for the existence of a third component of chlorophyll. Bot. Gaz., vol. 95, No. 4, pp. 529-563. SMITHSONIAN MISCELLANEOUS COLLECTIONS VOLE. 96,7, NO= 25 PE. ios) 7oB C 6001D E 5s00F G 400 —Normal spectrum | . . \_ Prismatic spectrum 6908— 4916— ho COMPARISON OF LEAF AND SEED PIGMENTS I. Spectrogram of acetone extract of leaf pigments. (Willstatter and Stoll.) pectrogram of acetone extract of seed pigments. lercury comparison spectrogram. 2) S 3. N 1 an é yl om SMITHSONIAN MISCELLANEOUS COLLECTIONS VOLUME 9%, NUMBER 3 BEOLOTROPIC RESPONSE AND CO, ASSIMILATION OF PLANTS IN POEARIZED TIGH) BY EARL S. JOHNSTON Division of Radiation and Organisms, Smithsonian Institution (PUBLICATION 3440) CITY OF WASHINGTON PUBLISHED BY THE SMITHSONIAN INSTITUTION JUL Yez9si7 } - _ i ~ _ 7 = - Wy gt i = . wi —— on \ \ \ TBe Lord Baltimore Press ‘BALTIMORE, MD., U. S. A. EPAQvOTROPIC RESPONSE, AND CO; ASSIMILATION OE PLANTS IN: POLARIZED LIGHT By EARL S. JOHNSTON Division of Radiation and Organisms, Smithsonian Institution From time to time articles both scientific and popular appear on the effects of polarized light on plants. Some years ago Semmens (1923) reported an increased velocity of seed germination in moon- light and suggested that the plane polarization of moonlight affected the diatase activity. Baly and Semmens (1924) reported an increased rate of hydrolysis of starches in plants exposed to polarized light. In a later paper Semmens (1930) characterizes plants grown in suc- cessive periods of darkness and polarized light by a disappearance of starch and other reserve products, such as glucosides, a temporary pho- totropism due to increased stem turgor, and a leaf fall with other signs of starvation. On the other hand, du Buy and Nuernbergk (1935) mention some unpublished experiments in which they found no differ- ence in the bending of Avena coleoptiles toward polarized and non- polarized light. Furthermore, Dastur and Asana (1932) indicate that the process of photosynthesis goes on as vigorously and regularly in polarized light as in ordinary light. Macht (1926) reports evidence of better growth of Lupinus, wheat, squash, and Helianthus seedlings in polarized light of a Mazda lamp than in his controls. Dastur and Gun- jikar (1935) report that leaves of 12 different species clearly show a larger amount of energy absorbed from polarized light than from nor- mal light of equal intensity. May (1924) conducted a number of ex- periments over a year to determine if there was a basis of fact regard- ing the seeding of crops during different phases of the moon. He found there was not enough difference in the general growth to be noticeable to the eye, certainly not enough upon which to found a theory. There is no evidence, as pointed out by Garner (1937), to show that the moon is capable of exerting any effect on crop plants other than those due to its action on illuminating conditions. In a number of plant growth studies conducted at the Smithsonian Institution it has been necessary to direct beams of light on the plants by means of mirrors. Judging from some of the discussions in the lit- erature one might raise the question as to the effect of polarized light on growth processes. Especially is this applicable to our studies on the growth of the oat coleoptile in monochromatic light reflected by a SMITHSONIAN MISCELLANEOUS COLLECTIONS, VOL. 96, No. 3 bo SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 mirror, and to CO, assimilation studies of wheat plants in which light intensities are increased by mirrors. The present paper discusses the phototropic response of the oat coleoptile and the CO, assimilation of young wheat plants in polarized light. The general method of illuminating the plant by two opposing beams of light was used in the phototropic experiments. This balancing @. LAMP \ “GLASS FILTER li Lei Mh ir a | \ y \ ! \ | \ | i ! \ \ ! \ | \ | \ | \ \ 4 \ | \ | \ , 1 /RON BOX \ | | i y | \ POLARO/DS Fic. 1—Diagram showing position of Avena coleoptile between two opposing beams of light which originate from a single source and are polarized at right angles to each other. action of light on phototropic response has been discussed by Johnston (1934) and by Castle (1931). In the experiments here discussed a single light source was used. Two beams were reflected by right-angle prisms in such a manner that the first 0.5 to 1.0 mm of the tip of a coleoptile was illuminated from opposite sides. The general arrange- ment of apparatus is illustrated in figure 1. NGI 3 PLANT RESPONSES IN POLARIZED LIGHT—-JOHNSTON 3 The light source was a 1,000-lumens, 1.6-amperes street series lamp with a heat-resisting, heat-absorbing extra light shade Corning filter (2.75 mm). After being reflected by the prisms two beams of light entered an iron box (34.5 x 18.5 x 30 cm high) through two oppositely located side windows. In each window was fitted a Polaroid disk (4 cm diameter). Each beam of light then passed through a hori- zontal 0.5-mm slit in a brass shield and fell on the tip of a coleoptile placed midway between these shields. These shields were 4 cm apart. The small Erlenmeyer flask (50 ml) in which the oat seedling was growing rested on a small shelf which could be raised or lowered by means of a worm gear. With this arrangement the tip of a seedling could be accurately placed in the path of the narrow opposing beams of light. The total length of each light beam from the lamp to the mid point between the brass shields was 107 cm. The oats, Avena sativa Markton, were germinated at approximately 25° C. between glass plates covered with moist filter paper. The plates were so placed in a moisture chamber that the seedlings grew verti- cally. A careful selection of the seedlings was made for straightness when they had attained a length of 1 to 2 cm. One was then trans- ferred to the small Erlenmeyer flask fitted with a cork stopper. It was supported by means of a little cotton 1n a small hole of the stopper. The flask was filled with distilled water so that the roots were entirely immersed. The seedlings were always handled in darkness or photo- graphically red light. One Polaroid disk was so placed that the plane polarized light was parallel to the axis of the seedling. The other Polaroid was placed to give a beam of light polarized at right angles to this. The setting at right angles could easily be accomplished by observing the lamp filament through the two prisms and two Polaroids and rotating the one Polaroid until the transmitted light reached its minimum visibility. At this point the filament appeared a dark purple red in color. By means of a specially constructed photocell a point was located between the slits of the two shields where the two beams were equally intense. At this point the seedlings either showed no phototropic bending or a very slight one after 2 to 3 hours. No change so far as the reactions of the coleoptile were concerned were noted when the Polaroids were rotated through an angle of go° in order to reverse the polarity of the light striking the two sides of the tip. The difficulty in this procedure was to locate accurately the mid point of equal intensities. The seedling was more sensitive to small differences of light intensity than the photocell. 4 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 In order to overcome this difficulty, a slightly different method was used. The mid point of equal intensities was approximately located with the photocell and the seedling placed slightly to the left of this point. A distinct phototropic bending then occurred toward the left. Another seedling was placed to the right of the mid point and the bending then occurred toward the right. The right or left displace- ments were never greater than 0.75 cm. Consistent results were obtained in a series of such experiments in which a fresh seedling was used each time, no matter whether the light was polarized parallel or at right angles to the axis of the plant. No difference could be detected in the phototropic response of the seedling in regard to the plane of polarization of the light impinging on its tip when the Polaroids were placed in the two positions mentioned. Calculations of intensities based on the lengths of light paths at points of maximum displacement of seedlings give a difference of slightly less than 3 percent as the maximum range. This clearly shows that if polarized light had a different effect on phototropism in one plane than in the other, as here used, such an effect is less than 3 per- cent. Crozier and Mangelsdorf (1924) found no difference in the phototropic efficiency of plane polarized and nonpolarized light of equal intensity on arthropods. The difference in phototropic effect of light depending on the plane of polarization which Castle (1934) found for the cells of Phycomyces is shown by him to be due to differences in the reflection losses at the cell surface. What Castle concludes for the growth of Phycomyces is undoubtedly true for the coleoptile of Avena, namely, that plane polarized light has no specific effect on its growth processes. The disappearance of starch and signs of starvation of plants grown in polarized light, as reported by Semmens, would indicate serious disturbances in the photosynthetic mechanism of such plants. This would undoubtedly involve the CO, assimilation process. Because of the disagreement between the results of Semmens and those of Dastur and Asana it was thought worth while to determine the CO, uptake of wheat plants in polarized and nonpolarized light, especially since this was the experimental plant used by Hoover, Johnston, and Brackett (1933) and by Hoover (1937) in their CO, absorption studies. In a series of experiments carried out by McAlister in which his recently described spectrographic method (1937, 1937a) for COs determination was used, little or no evidence was obtained that indi- cated a different rate of photosynthesis of wheat plants in polarized and nonpolarized light from a Mazda lamp. NO 3 PLANT RESPONSES IN POLARIZED LIGHT—-JOHNSTON 5 Young wheat seedlings were grown under controlled conditions in the growth tube and exposed first to nonpolarized and then to polar- ized light. The light source was a 1,000-watt Mazda projection lamp used with a suitable water filter. Light polarized parallel to the plant axis was obtained by inserting a Polaroid disk between the plant and the lamp. The Polaroid greatly reduced the light intensity and in order to reduce the intensity of the nonpolarized light to that of the polarized, a 200-mesh screen was inserted between the plant and the lamp. A small G. E. photocell foot-candle meter was used to approximate the intensities. Data from two such experiments are presented in table 1. TasLe 1.—Carbon Dioxide Assimilation of Young Wheat Seedlings in Nonpolarized and Polarized Light Light Character Intensity — A COs assimilation Thermocouple —=————— = Date Foot- reading with (mm? Cor- 1937 candles filter 2 IO min.) rected IWIEK i? 21562 Senor Nonpolarized 250 A Sas 482 588 Polarized 266 os 574 586 IMIR Se hip eee Nonpolarized 273 8.2 565 689 Polarized 282 9.8 670 684 8 With a Corning Aklo heat-resisting, heat-absorbing medium (2.46 mm) filter together with a 1o-cm water filter, the wave-length distribution was restricted to about 3800-7800 A, which includes the major portion of the spectrum active in photosynthesis. These thermo- couple readings were used in obtaining the corrected COs assimilation. In such experiments it is important that light and not the CO, of the air surrounding the plants be the limiting factor. In order to make sure this was the case, additional readings of CO, uptake were taken on May 5 at higher intensities in nonpolarized light. The CO, uptake at 700-800 foot-candles was found to be 1060 mm? per I0- minute interval. It therefore appears certain that at the lower intensi- ties used in these experiments light and not CO, was the limiting factor. Although the experiments on this phase of the work were not many and for 10-minute intervals, yet because of the extreme accuracy and quickness of this optical method of determining the CO, absorption by plants, it may safely be concluded that polarized light has no effect upon the uptake of CO,. There is the possibility, although it does not seem probable, that if the plants were grown for long periods in polarized light some secondary effects on the CO, assimilation might appear. The measurements indicate a difference of approxi- 6 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 mately 1 percent, but since it is assumed that the thermocouple-filter system measures the light effective in photosynthesis equally accu- rately in the two cases, it is well to bear in mind that a greater photo- synthetic efficiency may exist in one light than in the other. This may be the case because of the difference in color of the light after it passes through the Polaroid. Pollard (1936) has noted the wave-length dis- tribution characteristics of the Polaroid. Because the light transmitted is somewhat different in wave-length distribution from the screened Mazda light, accurate comparisons are exceedingly difficult. In summarizing, it may be concluded from the experimental evi- dence here presented that polarized light has no effect other than that of ordinary light in phototropism of Avena Markton. If there is a difference it is less than the 3 percent accuracy of the experi- mental technique. Also, there is little evidence that polarized light acts any differently from nonpolarized light of equal intensity and similar wave-length distribution in the process of CO, assimilation. EMT ERATURE Cllr Baty, E. C. C., and Semmens, E. S. 1924. The selective photochemical action of polarized light. I. The hydroly- sis of starch. Proc. Roy. Soc. London, ser. B, vol. 97, pp. 250-253. GAsina, 1B, S: 1931. The phototropic sensitivity of Phycomyces as related to wave-length. Journ. Gen. Physiol., vol. 14, pp. 701-711. CASTIE aS: : 1934. The phototropic effect of polarized light. Journ. Gen. Physiol., vol. 17, Pp. 751-762. Crozier, W. J., and Mancetsporr, A. F. 1924. A note on the relative photosensory effect of polarized light. Journ. Gen. Physiol., vol. 6, no. 6, pp. 703-706. Dastur, R. H., and Asana, R. D. 1932. Effect of plane-polarized light on the formation of carbohydrates in leaves. Ann. Bot., vol. 46, pp. 879-801. Dastur, R. H., and GunyjiKkar, L. K. 1935. Energy absorbed by leaves in normal and plane polarized light. Ann. Bot., vol. 49, pp. 273-281. pu Buy, H. G., and NueRNBERGK, E. L. 1935. Phototropismus und Wachstum den Pflanzen (dritter Teil). Ergebn. Biol., vol. 12, pp. 325-543. GARNER, W. W. 1937. Do the phases of the moon affect plant growth? (Mimeographed.) Bur. Plant Ind., U. S. Dep. Agr., February. Hoover, W. H. 1937. The dependence of carbon dioxide assimilation in a higher plant on wave length of radiation. Smithsonian Misc. Coll., vol. 95, no. 21, pp. I-13. NO. 3 PLANT RESPONSES IN POLARIZED LIGHT—-JOHNSTON 4 Hoover, W. H., Jounston, Eart S., and Brackett, F. S. 1933. Carbon dioxide assimilation in a higher plant. Smithsonian Misc. Coll., vol. 87, no. 16, pp. 1-109. JOHNSTON, EArt S. 1934. Phototropic sensitivity in relation to wave length. Smithsonian Misc. Coll., vol. 92, no. 11, pp. 1-17. Macut, Davi I. 1926. Concerning the influence of polarized light on the growth of seedlings. Journ. Gen. Physiol., vol. 10, no. 1, pp. 41-52. May, D. W. 1924. The moon theory. Porto Rico Agr. Exp. Stat., Agr. Ext. Notes (mimeographed), no. 65, Feb. 15. McAtutster, E. D. 1937. Spectrographic method for determining the carbon dioxide exchange between an organism and its surroundings. Plant Phys., vol. 12, PP. 213-215. 1937 a. Time course of photosynthesis for a higher plant. Smithsonian Misc. Coll., vol. 95, no. 24, pp. I-17. POrcarp;, A. F. C. 1936. Polarization of light and some technical applications. Nature, vol. 138, no. 3486, Dp. 311-314. SEMMENS, ELIZABETH SIDNEY 1923. Effect of moonlight on the germination of seeds. Nature, vol. 111 PP. 49-50. 1930. Hydrolysis in the living plant by polarized light. Bot. Gaz., vol. go, pp. 412-426. i) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO. 4, PL. 1 _ ald ceRS ae TRADE BEADS FOUND AT LEEDSTOWN, NATURAL SIZE SMITHSONIAN MISCELLANEOUS COLLECTIONS VOLUME 9, NUMBER 4 INDIAN SITES BELOW THE FALLS OF THE RAPPAHANNOCK, VIRGINIA (WITH 21 PLATEs) BY DAVID I. BUSHNELL, JR. (PUBLICATION 3441) CITY OF WASHINGTON PUBLISHED BY THE SMITHSONIAN INSTITUTION SEPTEMBER 45,1937 The Lord Battimore Press BALTIMORE, MD., U. 8. A. { a e { : 7 - , G A pe ony a CONTENTS PAGE np DE Ce a a Seas ee cy ew Oe I a enyr Os thie ap PalaANMOCk . o...,0 5 00% ds siviele noo de ard Geek ass 3 ao Rides 2 Acts relating to the Indians passed by the General Assembly during the second half of the seventeenth century.................0: 4 Movement of tribes indicated by names on the Augustine Herrman map, MOP tok ase hovers ote toh aie Ae lah ete ae EEN ee arn 10 SaemenrnTe (eit GELLIEMICIIES:. cc 5. sedis sts o's s sie ov ers 6.4 oc nuda ie ed widin ws Retopal esere 15 SO ae ers ech foig Aes ae ews loa ds ho Silaiew nisiok Aine AO e RC eee aie eee 16 RO LUCIA gas estes eyelet cores bse oik perbiay SRA AOE ee iT Ee ae 21 SOAD STONE geese err ere ortrele ae yates EG ee NER eR crenata 25 achepotatrade bead Siric.cm. ac ctsre css tome steels ete ce ita ie ores 2 DISCOVELRYT OF ENE. DEAUS es xs: cers serio et eee at Be 30 TENE RY GLORIES SIS he RCE pS ERA ers CRM et cee A Da a SS ih a 35 Nantel bane utarGertd waters acters cosy caaresa Cine torces Stoke ees eR a ee Cerne 36 Rortapacovdllagres TOSOka:sava ore «sieve cierecre Sei eather reiterci te at 39 Matenialerconm site ot Nandtanehtacunds..os-e acc aoe ene 42 [FACETS Latico eae ee Oo eo Omars Oricon sorcncisis so. cic. a. OM oetowie 43 SOAPSCOMEL eset cl cs creteteavests @ e SAGO RO eT Ee ROE OI Ete ae rer 50 iNoowe Jett ALG eieOn Sh epee ane Epa adios oad sto once oaee bo ado oe 51 Left bank of the Rappahannock above Port Tobago Bay.............. 2 EMMIO ME naOleVnllnamlc Greeleuth cao scoleicts cnccevein slate eeenekenstiene eels ork teers 55 CINGGOBISSONIEY ane ee nO re ton dod oapnon Hart ono mone 56 alniakeGnOmNUOUM tects coe hc cot eaitiee Cote ee EEE LR Eee 57 SeenON oars S PTTL TCNENI CY 050.5, Give whan, 3 al Sa, oom w: ein) ds Sud alee at og ROR RRERINC eae epAES At oP sags al 58 Waposttesthe mouth of Hough Creek... .%/4ieds. ae eae fee cess 60 Sra Fea O MPIC NAY. D5 ces aissno'c aseoroco vats. 0oe Sie oer RGIS ete A ROI ee GaN aes 60 Soak nwo ais AGO Oe ee En Donan cnndio otic cere coma censors 62 onciisionsestecested by certain SpeCimmenS...- «sem antec ses ein ee 63 ili ie i in at im re nn ILLUSTRATIONS PLATES PAGE 1. Trade beads found at Leedstown (Frontispiece) 2. North over the Rappahannock showing Leedstown and the site of | ETISSE ISIS oa ee een Se cies, MAG ie ee BX Sm, Brine extant ree 18 Sepeeatens from site Of Pissasecke 2-66 004.06 05+neclse vg aes os sieae cre oes 18 Mme pecinens trom site Of PissaseCks. c.e.as sce cc osuker vs oti ceueeoteene 18 SeSpecimensmromesite of issaseck:. +. . ces ases «cele ew eietioieasicecle we cites 18 Omen pPecimens trom site Of sPissaseck:; one csiccio ce occas ocemiocencccnenecs 26 Pee PEciuens from Site OL Pissaseck.m7. os. aa cles oasis eiteicla sae sieleielane 26 DMO VECIMeNS inom) Slike OL oISSASeEClos). siete cues) siete eee iecieteleieiele sie cieisioe oe ouavelore 20 g. 1. Specimens from site of Pissaseck. 2. Specimens from left bank of the’ Rappahannock above Greenlaws Wharf..................... 26 10. Site of Nandtanghtacund with Port Tobago Bay....................: 42 MeO peciinens trom site ot Nandtanghtacundi: «4... cece de ene ocaes. ese: 42 ianopecimens trom. site.of: Nandtanghtacund....2..00.20c0ssee5as cose 42 13. Fragments of pottery from site of Nandtanghtacund.................. 42 14. I. North over the Rappahannock showing mouth of Chingoteague Greek: 2: Pipes from) site at mouth of creek...3...........0-.-- 52 I5. Specimens from site at mouth of Chingoteague Creek................. 52 16. North over the Rappahannock, showing Cleve and Taliaferro Mount.. 52 iA SPECIMENS e LOM Mappaliatinock, Sites). a. HO ee 52 10. Point resembling the Folsom type from near mouth of Millbank Creek... 55 Hime SPECnenSELLOMETICar WOSUIS Utley. rtnteteeiteieteletoe tee tasiaeaicis x ercie erie ee 59 [INDIAN SITES, BELOW ‘THE FALLS OF THE RAPPAHANNOCK, VIRGINIA By DAVID i BUSHNELI: Jr. (Wir 21 PLaTeEs) INTRODUCTION During the summer of 1608, the year following the settlement of Jamestown, parties of the colonists made two successful and, considered in retrospect, very important expeditions along the shores of Chesapeake Bay. They discovered many streams that flow into the bay and came in contact with the natives who occupied the ter- ritory. The second party left Jamestown July 24 and returned late in August, after having explored the country to the northward, to the mouth of the Rappahannock. They entered the river and continued up the stream to the large island at the falls, near the present city of Fredericksburg, as far as their boat could be taken. The valleys beyond were occupied by the Manahoac tribes, a Siouan group ever enemies of their Algonquian neighbors; many camps and villages of the latter were then standing on the banks of the Rappahannock below the island and had been passed by the English when they ascended the stream. However, no villages were en- countered within a distance of 10 or 12 miles of the falls; that part of the valley appeared to have been recently abandoned by the Algon- quian tribes. The country claimed and occupied by the Manahoac tribes in 1608 has already been described,’ together with examples of material from various localities, and consequently, it is now thought desirable to present, for comparison, a brief description of sites that have been discovered on the banks of the river, beginning at Leedstown, the early colonial settlement some 40 miles below the falls, and con- tinuing up the valley. I am again indebted to Capt. H. K. Baisley, United States Army Air Corps, for aerial photographs of the areas described; and to F. M. Aldridge, of Fredericksburg, and R. G. Paine and E. G. Cassedy, of the Smithsonian Institution, for assistance in locating sites and material. * Bushnell, David I., Jr., The Manahoac tribes in Virginia, 1608. Smithsonian Mise. Coll., vol. 94, no. 8, 1935. SMITHSONIAN MISCELLANEOUS COLLECTIONS, VOL. 96, No. 4 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 tbo DISCOVERY OF THE RAPPAHANNOCK At the time of the establishment of Jamestown, May 1607, many settlements of Algonquian tribes stood on the banks of the Rappa- hannock, from the mouth of the stream far inland to the vicinity of the falls or rapids near the present city of Fredericksburg. Some were comparatively large villages with smaller camps nearby. The river was first explored by the colonists late in the summer of the following year, 1608, when they penetrated the wilderness as far as the falls and there came in contact with the Manahoac tribes who claimed the lands to the westward. As the English advanced up the Rappahannock from the Chesa- peake, they saw many native villages, and when near the settlement of the Rapahanock, probably in the present Richmond County, opposite the town of Tappahannock, they were treacherously at- tacked by the Indians but escaped injury. A short time before reaching the Rapahanock village the English were joined by the friendly Indian Mosco, who later served them as guide and inter- preter. They continued up the river and discovered other Algonquian camps and villages, and to quote from the narrative (p. 426) :* “ The Kings of Pissassack, Nandtaughtacund, and Cuttatawomen, used us kindly, and all their people neglected not any thing to Mosco to bring us to them.” The three villages will again be mentioned, and the sites described. The brief accounts that have been preserved of the encounter of the English with the Manahoac, who had gathered just above the large island, are of the greatest interest and importance. And as related by the Manahoac Indian from Hassininga, through the Algonquian interpreter, while in the vicinity of the falls they were near ‘‘ the bounds betwixt the Kingdome of the Mannahocks and the Nandtaughtacunds.” Thus it appears that the latter name was applied to the Algonquian group that occupied the country along the Rappa- hannock just below the falls, and therefore nearest neighbors of the Siouan or Manahoac tribes. Again when referring to the Rappa- hannock (p. 348) : Upon this river on the North side are the people Cuttatawomen, with 30 fight- ing men. Higher are the Moraughtacunds, with 80. Beyond them Rapahanock with 100. Far above is another Cuttatawomen with 20. On the South is the pleasant seat of Nantaughtacund having 150 men. This river ... is replen- ished with fish and foule. * Smith, Capt. John, The generall historie of Virginia, 1624. All references to Smith’s writings are quoted from the English Scholar’s Library edition, edited by Edward Arber, Birmingham, England, 1884. NO. 4 INDIAN SITES ON THE RAPPAHANNOCK—BUSHNELL 3 Many additional names appear on the Smith map, a section of which is reproduced in figure 1. Of the five groups mentioned in the preceding quotation the last two were within the limits of the region WO % rs, ws . 4 JA coun - Wau Lod > us Ca Ll ( LO} » — pe me, atacunt_ rom Fic. 1.—Section of the Smith map, 1624, showing the many native villages that stood on the banks of the Rappahannock from the vicinity of Leedstown to the falls at Fredericksburg. now being considered, which extends from Leedstown, on the left bank of the Rappahannock in Westmoreland County, up the river to the falls at Fredericksburg. 4 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 ACTS RELATING TO THE INDIANS PASSED BY THE GENERALE ASSEMBLY DURING THE SECOND HALF OF THE SEVENTEENTH CENTURY Great changes necessarily occurred among the native tribes of Virginia after the English entered the wilderness and occupied much of their land. Unrest developed, and it is evident that soon after the middle of the century the native tribes whose villages stood on the banks of the Rappahannock, beyond the English plantations, were regarded with grave apprehension. In November 1654, among “ Orders of Assembly ”, as transcribed by Hening, is one “Concerning the March against the Rappa’ Indians.” This reads in part (p. 389): “ Whereas divers complaints have bin made by the inhabitants of the counties of Lancaster, Northumberland and Westmoreland concerning divers injuries and insolencyes offered and done by the Rappahannock Indians, unto them.” The order then states the number of men to be furnished by each county, “‘ with armes, amunition and provisions, with boates and other necessaries for their voyage to the said Rappahannock townes.” But it is not known to which villages this referred. Late in the spring of 1656 occurred the sanguinary encounter between colonists and Indians, the latter thought to have been a Siouan tribe from farther up the Rappahannock, who had moved to the vicinity of the falls of the James. It was a period of much con- cern to the frontier settlements of Virginia where the native tribes were being harassed by their enemies from far northward. Many laws affecting the Indians were enacted and enforced in Virginia during the succeeding years. Some made specific reference to the people of the Rappahannock valley, others were more general and applied to all tribes alike. Trade was frequently the subject of the acts, and these now shed light on the economic problems with which the colonists were then confronted. Brief quotations from Hening should prove of interest: Act CXXXVIII of the “Grand Assembly Held at James City March the 23d 1661-2” was a digest of earlier laws relating to the Indians some of which reveal facts of historical importance.’ lirst concerning food: And be it further enacted that for the better releife of the poore Indians whome the seating of the English hath forced from their wonted conveniencies *Hening, William Waller, The Statutes at Large; being a collection of all the Laws of Virginia, from the first session of the Legislature, in the year 1619, vol. 1, New York, 1823. “ Hening, op. cit., vol. 2, p. 140. NO. 4 INDIAN SITES ON THE RAPPAHANNOCK—BUSHNELL 5 of oystering, fishing and gathering tuckahoe,’ cuttyemnions™ or other wild fruites by which they were wonted for a greate parte of the yeare to sub- sist, Be it therefore granted, enacted and confirmed that the said Indians upon *Smith wrote in 1612, when referring to the food of the Virginia Indians with whom he had come in contact (op. cit., p. 58): “ The chiefe roote they have for foode is called Tockawhoughe. It groweth like a flagge in low muddy freshes. In one day a Savage will gather sufficient for a weeke. These rootes are much of a greatnes and taste of Potatoes. They use to cover a great many of them with oke leaves and ferne, and then cover all with earth in the man- ner of a colepit; over it, on each side, they continue a great fire 24 houres before they dare eat it. Raw it is no better then poison, and being roasted, except it be tender and the heat abated, or sliced and dried in the sun, mixed with sorrell and meale or such like, it will prickle and torment the throat extreamely, and yet in summer they use this ordinarily for bread.” This was undoubtedly the same plant that was gathered by the Indians a generation later, and to which reference was made in the Act just quoted. But it is evi- dent that in later years many, if not all, edible roots became known as tuckahoe to the English colonists, although they may never have been known as such to the Indians. A century later, in 1749, Peter Kalm (Travels into North America, second edition, London, 1772) mentioned several plants to which the name tuckahoe was then applied. He wrote (pp. 387-388): “ Taw-ho and Taw-him was the Indian name of another plant, the root of which they eat. Some of them like- wise call it Tuckah; but most of the Swedes still knew it by the name Taw-ho. It grows in moist ground and swamps ... Nor did the Indians ever venture to eat them raw, but prepared them in the following manner: They gathered a great heap of these roots, dug a great long hole, sometimes two or three fathoms and upwards in length, into which they put the roots, and covered them with the earth that had been taken out of the hole; they made a great fire above it, which burnt till they thought proper to remove it; and then they dug up the roots, and consumed them with great avidity. These roots, when prepared in this manner, I am told, taste like potatoes. The /ndians never dry and preserve them; but always take them fresh out of the marshes when they want them. This Taw-ho is the Arum Virginicum, or Virginian Wake-robin.” The following note by Dr. Frederick V. Coville tends to identify the plant first called Tockawhoughe by the Virginia colonists: “ Peltandra virginica. The old Indian name, tuckahoe, appears never to have been adopted by botanists as the common name of this important Indian food plant. It has been called Virginia wakerobin because of its close relationship to Arum maculatum, the wakerobin of England. It bas also been called by botanists arrow-arum, be- cause of its arrow-shaped leaves.” Dr. Coville was convinced of the correct- ness of the conclusion, but desired to determine for his own satisfaction the edibility of the roots before the statement should appear in print. In the endeavor to collect roots of the plant, Peltandra virginica, we visited Hunting Creek, on the Virginia side of the Potomac just below Alexandria, Dec. 29, 10936, but were unsuccessful as the leaves had disappeared because of the lateness of the season, and the roots could not be identified. Dr. Coville was taken ill a few days later and died on Jan. 9, 1937, and thus the work remained unfinished. ** The wild fruit to which the name cuttyemnions was applied has not been identified, but the etymology of the word has been determined by John P. Har- 6 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 addresse made to two of the justices of that county they desire to oyster or gather wild fruite in, as aforesaid, they the said justices shall grant a lycense to the said Indians to oyster or gather fruites as aforesaid. But the time was to be limited, and the Indians were to have with them “only such tooles or implements as serve for the end of their comeing.” No guns were to be carried. The use of badges in identifying Indians when they visited the English settlements is mentioned in the same act. It was stated in part: ... because an intervall betweene the Indians and English cannot in the present neernesse of seating be soe laid out as may wholly secure the English from the Indians comeing in and pilfering things from them if a free inter- course be admitted, Be it therefore enacted for the prevention thereof and to the end that the nations may be distinguished and soe if they are taken in the manner of doing any injuryes the sufferers know to what kings to addresse themselves for remedy, that badges (vizt.) silver plates and copper plates with the name of the towne graved upon them, be given to all the adjacent kings within our protection. And that all the said kings give it in charge to their people that none of them presume upon what occasion soever to come within the English bounds without those badges upon them or one with a badge in their company, and if any damage or injury be done to any Englishman by them or any of them, that then the king or greate man of the place the badge denote shalbe answerable for itt; and if any shall notwithstanding this injunc- tion be found in our bounds without any such badge® or not accompanied with one that shall have them, that then it shall be lawfull for any Englishman to apprehend and carry him or them before any justice of the peace who shall keepe him or them in safe custody untill their king or greate man ransome them by paying one hundred armes length of rohonoake* for each Indian soe takenter. rington, of the Bureau of American Ethnology, who states: “The name ‘cuttyemnions’ evidently refers to some small black or dark-colored berry, the species of which has not been determined. The name is to be analysed as follows: cutty-, black, phonetically kate or makate; -min, berry of any kind; -an or -in, suffix denoting plural of inanimate objects, pluralizing the element -min. The name therefore means small black colored fruits or berries.” °Certain of the large metal plates discovered by Judge Graham in burials on the banks of Port Tobago River, Maryland, may have been used as badges; however, the surfaces of the specimens are greatly corroded, and if names had been engraved on them as mentioned in the Act, all traces have disappeared. 7 The writer is of the belief that this referred specifically to the small shells, Marginella virginiana Conrad, which occur in vast quantities in the waters of the Chesapeake. The shells were perforated and strung on threads of sinew to serve as ornaments. The belief is suggested by a brief reference found in the small printed catalog of the Museum Tradescantianum, published in London, 1656. On p. 47 of the catalog is mentioned: “ Pohatan, King of Virginia’s habit all embroidered with shells, or Roanoke.” This was one of the pieces associated with the famed coronation of Powhatan during the autumn of 1608, when the NO. 4 INDIAN SITES ON THE RAPPAHANNOCK—BUSHNELL Ii, The reply to a question submitted to the Governor of Virginia, by the Lords Commissioners of Foreign Plantations in 1670 was: “ The Indians, our neighbours are absolutely subjected, so that there is no fear of them.” * This referred to the Indians within the colony. However, if peace prevailed within the colony, there was always much to fear from beyond the frontiers. The Assembly of March 1675-6, anticipating an invasion of the colony and attacks on the outlying settlements, passed “ An act for the safeguard and defence of the country against the Indians.” This act, after mentioning the garrison to be located on the Potomac in Stafford County, continued: “one hundred and eleven men out of Glocester county to be garrisoned at one ffort or place of defence at or neare the ffalls of Rapahanack river, of which ffort major Lawrence Smith to be captain or cheife comander.” * The acts passed “ At a Grand Assemblie, Holden at James Cittie the fifth day of June, 1676’, included much of importance concerning the Indians of the colony and others who lived beyond the bounds. This assembly was held while the colony was dominated by Nathaniel Bacon and his followers, and consequently the acts became known as Bacon’s Laws. Act I was “ An act for carrying on a warre against the barbarous Indians ”, which referred to the Conestoga, who had harassed the Virginia frontiers and had later sought refuge near the Occaneechi far south in the present Mecklenburg County, Va. The act mentioned the number of troops to be raised in the several counties, supplies to be collected, and other questions relating to the proposed campaign. latter “ gave his old shoes and his mantle to Captain Newport” (Smith, op. cit., p. 125). The remarkable specimen is now in the Ashmolean Museum, Oxford, England. The ornamentation was formed by attaching great numbers of the shells, Marginella virginiana, to the deer skins of which the “ habit ” or “ mantle ” was made. As defined in the Handbook of the American Indians (Bureau of American Ethnology, Bull. 30): “Roanoke. A name applied, with several variants, by the Virginia colonists, to the shell beads employed by the neigh- boring Indians as articles of personal adornment or media of exchange; a case of substitution of a familiar word for one that was ill understood and probably more difficult to pronounce. Capt. John Smith (1612 and 1624) gives the Pow- hatan name for shell beads in the form of rawrenock and rawranoke, and Wil- liam Strachey defines rarenaw as ‘a chain of beads.’ The root rar means to ‘rub’, ‘abrade’, ‘smooth’, ‘ polish.’ The original word may have been raérenawok, ‘smoothed shells’, pl. of rarenaw.” ® Hening, op. cit., vol. 2, p. 513. ® Hening, op. cit., vol. 2, p. 327. 8 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 Act II was “An act concerning Indian trade and traders.” This prohibited trade with the Indians and then continued: Provided nevertheless that it shall and may be lawfull that such Indians who shall serve the English in the warr, and onely such be supplied to the vallue of their wages and pay for the takeing prisoners in such necessary things as they shall want, armes and ammunition wholy excepted, and it is hereby intended that our neighbour Indian ffriends bee not debarred from fishing and hunting within their owne limmits and bounds, useing bowes and arrowes onelie. Pro- vided also that such neighbour Indian ffriends who have occasion for corne to releive their wives and children, it shall and may be lawfull for any English to employ in fishing or deale with fish, canooes, bowles, matts or basketts, and to pay the said Indians for the same in Indian corne, but noe other commodities . . .” This is of the greatest interest in proving that the English made use of dugout canoes obtained from the Indians, and also pottery vessels, rush mats, and baskets made in the Indian villages. The use by the colonists of such material of Indian make was probably general throughout the English settlements. It is likewise evident that fish were obtained by the English from the Indians, and much game may have been procured from the native hunters, who continued to use their bows and arrows. During the meeting of the assembly, June 1676, when forces were being gathered for the intended expedition against the “ barbarous Indians’, it was ordered that “the forte in Rappahannack countie commanded by major Lawrence Smith, which was settled or in- tended to be settled by vertue of a late act of assembly, be forthwith deserted ...”’ The troops were to be ordered to other posts on the frontier where conditions were more serious. This was only 3 months after the order had been given for the erection of the frontier ‘“ ffort or place of defence ”, and if ever erected, which is doubtful, it would probably have resembled the palisaded structures so characteristic of the wilderness far beyond the mountains a century later. By the autumn of 1677 the growing importance of trade with the neighboring Indians was acknowledged by the burgesses, and acts were passed setting forth the manner in which it was to be conducted. On October 10, 1677, the Grand Assembly met at the house of Capt. Otho Thorpe, at Middle Plantation. Act III of that date, “ An act lycensing trading with Indians’’, was of great importance. It read in part: Forasmuch as the totall prohibition of tradeing with Indians is experimented and found hurtfull and prejudiciall to his majesties colony and the inhabitants * Hening, op. cit., vol. 2, p. 350. NO. 4 INDIAN SITES ON THE RAPPAHANNOCK—BUSHNELL 9g thereof; Bee it therefore enacted by the governour, councell and burgesses of this present grand assembly, and the authority thereof, and it is hereby enacted, that all Indians whatsoever being in amity and ffriendship with us from hence- forth shall have free and full liberty to come in amongst us and bring in any comodities whatsoever to the severall places and at the severall tymes here- after sett downe, and mentioned, and to trade with, sell or truck, for the same with the English, resorting thither, but noe where else for any comodityes what- soever, and that such marts of ffaires continue fforty dayes and noe longer, (that is to say) . . . the place for the marte or ffaire in Rappahanock river to be appointed and set downe by the justices of the peace in Lancaster county, being mett and satt in the court, the begining of which marte or ffaire to be yearely the twentyeth day of Aprill and the twentyeth of September... ‘ b A similar “ marte or ffaire”’ was to be held in Stafford County, beginning on April 30 and September 30 each year.” A description of a gathering at a ‘“ marte or ffaire’’, with Indians and colonists coming together to trade, would be of the greatest interest, but none is known to have been preserved. Nor is there any known record of when or where such a gathering took place. Two years later the protection of the outlying settlements again caused anxiety. The first act of the Grand Assembly “ begunn at James Citty the 25th of April, 1679” was “ for the defence of the country against the incursions of the Indian enemy.” This provided “that fower houses for stores or garrisons be erected and built at the heads of the ffower greate rivers.” The Rappahannock was one of the four. Two structures were to be erected at each chosen site, of the same dimensions, described as “‘ one store house to be strongly built, and well covered to be sixty foot long, and twenty two foote broad, and one small house of tenn foot square to be strongly built for ammunition.” Maj. Lawrence Smith of Gloucester County was in command at the falls of the Rappahannock. The frontier posts were maintained until the autumn of 1682, when it was ordered “ that the said severall forts and garrisons be dismantled . . .”’” The Indians were no longer feared as formerly, and to supply the garrisons proved a burden to the colony. Thus, just three-quarters of a century after Captain Smith had been conducted by his Indian captors to their settlements on the banks of the Rappahannock, the native population of the entire valley had been dispersed and the sites of many of their villages, with the surrounding cornfields, had become plantations of the English colonists. 1 Hening, op. cit., vol. 2, p. 410. ” Hening, op. cit., vol. 2, p. 498. IO SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 MOVEMENT OF TRIBES INDICATED BY NAMES ON THE AUGUSTINE HERRMAN MAP, 1673 The history of the Indians of the Rappahannock valley during the years from 1650 to 1670 would undoubtedly reveal a movement and shifting of many tribes from the localities in which they were encountered in 1608 and prove that others from a distance had entered the region. Although no references to the actual migration of native tribes from Maryland across the Potomac to Virginia have been discovered, it is evident that soon after the middle of the seventeenth century such movements did occur and that tribes or parts of tribes crossed to the right bank of the stream and thence continued to the Rappa- hannock, where they established new homes. This alone would explain the occurrence of certain names on the Herrman map, which was issued in 1673. A section of the map, representing that part of the Rappahannock valley now being considered, is reproduced in figure 2. The names that were recorded on that part of the river in 1608 have disappeared and have been replaced by some that were at that time applied to villages beyond the left bank of the Potomac. The official records of Maryland are replete with references to the native tribes with whom the settlers came in contact. Many lived within the Province, others came from beyond the bounds. Brief quotations from the records will make clear the causes which are thought to have led, ultimately, to the removal of many tribes from their earlier habitat. But such movements had continued through the centuries. On March 24, 1652, just 20 years after Cecil, Lord Baltimore, had received from Charles I the charter for Maryland, it was stated that: ‘Whereas there is Great Suspicion that the Indians intend Some Mischeife to the Inhabitants of this Province”, for which reason Thomas Gerrard was ordered to raise a sufficient force and “repaire to Portoback or Choptico and (if he think fitt) either dis- arme or Secure the persons of any of the Said Indians till Exami- nacon had touching the premisses or further Order.” Later in the same year a document of importance was recorded, being: “ Articles of Peace and freindshipp treated and agreed upon the 5th day of July 1652 Betweene the English Nation in the Province of Maryland on the one party, And the Indian Nation of Sasquesah- anogh on the other partie...” But this was not concluded for “ Archives of Maryland. Proceedings of the Council of Maryland, 1636-1667. Maryland Hist. Soc., Baltimore, 1885. NO. 4 INDIAN SITES ON THE RAPPAHANNOCK—BUSHNELL Te Fic. 2—Section of Augustine Herrman map, 1673. The Rappahannock from below Leedstown to the falls. 2 I2 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 some years (not until September 16, 1664), after other questions had been decided.” ‘“‘ Then was taken into Consideracon the Articles of peace with the Sasquesahanough Indians dated fift of July 1652.” During the same year that peace was made with the “ Indian Nation of Sasquesahanough”’ trouble developed with the natives on the Eastern Shore, across Chesapeake Bay. As recorded in the Proceedings of the Council, on November 25, 1652, at a Court held at St. Marys, an expedition was planned against the Indians of the Eastern Shore. It was that day “ordered and appointed that Sufficient forces be Speedily raised for a march against the Said Easterne Shore Indians, and for that purpose every Seventh man throughout the province is to be pressed for this Service ” Such an expedition was considered of the utmost importance for the safety of the people of the Province, but it was soon ascer- tained that the Indians on the Eastern Shore against whom the expedition was directed, as well as others living west of the Chesa- peake, had knowledge of the proposed invasion. Consequently, on December 18, 1652, “ with divers other very Materiall reasons as the danger of being frozen in and otherwise”’, it was decided to defer “the designe till a more Seasonable time and opportunity.” On May 2, 1661, the General Assembly, as recorded in the Pro- ceedings, feared the invasion by Indians from the north. Both houses assented to— An acte Impowring the Governor and Councell to Rayse forces and mayn- tayne a warre without the Province and to ayde the Sasquehannough Indians. Whereas it doth appeare to this prsent Generall Assembly that this Province is in Eminent danger by a warre begun in itt by some forreigne Indians as it hath been made appeare by credible informacon given of a person lately killd and of others that are probably cutt off by these forreign Indians. And that in humane probabillity our neighbour Indians the Sasquehannoughs are a Bull- warke and Security of the Northerne parts of this Province. And that by former treatyes with that nacon they have very much assured us of their affeccons and friendship And that they expected the like from us . That same month peace was renewed between the people of Mary- land and the Indians, but this was distinct from the document dated July 1652 and again September 16, 1664. It is presented in the Pro- ceedings of the Council, and reads in part: “ Articles of peace and amity concluded on the behalfe of the Lord Proprietary of the Province of Maryland and the Sasqsahannough Indians at a treaty begun at Spes Utia the 16th Day of May 1661 ... ” * Archives of Maryland. Proceedings and Acts of the General Assembly of Maryland, 1637-1664. Maryland Hist. Soc., Baltimore, 1883. NO. 4 INDIAN SITES ON THE RAPPAHANNOCK—BUSHNELL ae The agreement was for mutual help and understanding. The fifth article reads: “‘ That the Sasquesahannoughs shall finde the English Souldiers with sufficient fish and flesh, and bread ready beaten.” The last item is thought to have been tuckahoe, as described on page 5. Trouble continued within the Province and on November 17, 1663, the following statement was made to the Council: Whereas Sundry Complaints have been made unto mee by the Queene of Portaback in behalfe of her Selfe and Indians undr her, Intimating how that they have not only left their Towne standing by the water, but have removed themselves farther of even to their utmost bownds of their land—Leaving place to the English to Seate on theire ancient plantacons by the River side the English not being (as they informe mee) contented with what Land is allready freely granted Doe still take up land and Seate themselves very nigh unto the said Indians whose stocks of Cattle and hoggs doe and will yearely destroy theire Corne fields by which meanes they must of necessity come to ffamine they not knowing the way and meanes to fence in theire Corne ffeilds as the English doe will soon come to destruccon... The English were then ordered to remain away from the Indian lands. The northern Indians again invaded the Province and soon reached the banks of the Potomac and the shore of Chesapeake Bay. As recorded in the Proceedings of the Council: Att a Councell held the 27th June 1664: Then was taken into Consideracon the prservacon of the province agt the incursion of the Cinego Indians * who lately killed some English in Ann Arundell County & alsoe entred the County of St Marys, Ordred tht Warr be pclaimed agt the Cinego Indians tht a Reward of a hundred Armes lenght of Roan Oake be given to evry pson whether Indian or English tht shall bring in a Cinego prisonr or both his Eares if he be slayne .. . Many Maryland Indians may at this time have crossed the Potomac to seek new homes in Virginia, where they became estab- lished on more ancient sites that had been seen by the English in 1608 and which were shown on the 1624 map. However, it is evident that all did not abandon their lands beyond the Potomac in Maryland, and the names of many tribes were joined in a treaty made 2 years later. This was recorded in the Proceedings of the Assembly : Articles of peace & amity concluded agreed upon between the Rt. honoble Caecilius Lord & Propt. of the Province of Maryland & Avalon Lord Barron, of Baltemore & the Indians of Pascattoway Anacostanck, Doags, Mikiki- womans, Manasquesend, Mattawomans, Chingwawateick, Hangemaick, Porto- backes, Sacayo, Panyayo, & Choptico the 20th day of Aprill 1666. The treaty contains many interesting references to conditions and customs, and to the privileges accorded the Indians. * The Seneca Indians; the name Seneca, however, was usually applied to all members of the Five Nations. 14 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 Augustine Herrman finished gathering data for his map during the year 1670 and then forwarded it to London, where the plates were engraved by William Faithorne and where the map was issued in 1673." Consequently, the names that appear on that part of the map which is reproduced in figure 2 were the names of villages then inhabited—comparatively recent settlements which occupied more ancient sites. The different tribes or groups whose names they bore were not encountered on the banks of the Rappahannock by the first party of English from Jamestown, who ascended the stream in 1608. Portobacco and Doogs Indian, names on the map, are obviously the Portobackes and Doags mentioned in the treaty of 1666. The former was one of the best-known tribes of Maryland, and when the English ascended the Potomac in 1608, their village was mentioned as “ Pota- paco with 20 [men].’’ This was in the present Charles County, Md. The name which in English is bay or cove later became corrupted into Port Tobacco.” True, the name would have been applicable to the great bay, now known as Port Tobago or Port Tobacco, on the south side of the Rappahannock, but it does not appear to have been so designated in 1608, certainly not on the 1624 map. Beverley,” writing after 1692, referred to the scattered tribes then living in different parts of Virginia. He said in part: “ In Richmond Port-Tobago, bout five Bow-men, but Wasting.” This was Rich- mond County, on the left bank of the Rappahannock adjoining West- moreland. The small remnant probably numbered about 20 individ- uals and was decreasing. This particular group may never have lived south of the river, in which event they had retained the name by which they were known in 1608, when living on the Maryland side of the Potomac, although they had become separated from other members of the tribe. The identity of the tribe or tribes known as the Dogue or Doag Indians has not been clearly determined. The name does not occur in early records of Virginia and evidently does not appear until after the middle of the seventeenth century. The tribe is believed to have been part of the Nanticoke, on the Eastern Shore of Maryland, ** Phillips, P. Lee, The rare map of Virginia and Maryland, by Augustine Herrman. Washington, 1911. "Judge William J. Graham, Washington, D. C., who examined many sites on both sides of the Port Tobacco River, described them in a paper: The Indians of Port Tobacco River, Maryland, and their burial places, n.p., n.d. [1935]. * Beverley, Robert, The history and present state of Virginia, book 3, p. 62 London, 1705. ’ NO. 4 INDIAN SITES ON THE RAPPAHANNOCK—BUSHNELL 15 whence they may have removed, some having crossed the Potomac and continued to the banks of the Rappahannock. The tribal name is found in Maryland on early land grants and surveys:” “ Doegs Point in Mattawoman Creek”, Prince Georges County, 1699; and “Doags Neck”, Charles County, 1657, both counties bordering on the Potomac. Also several references to tracts in Kent County, on the Eastern Shore, dated 1704 and 1705: “Ona branch called Dogue Branch and on the north side of Chester River”, and again “ on the north side of a Branch called Dogue Branch just below a meadow called Chester Meadow.” The latter references are interesting as they apply to places within the region still occupied at that time by the Nanticoke. Chinquatuck ” of the Herrman map may be the Chingwawateick of the 1666 treaty. The occurrence of the names on the Rappahannock should be ac- cepted as proof that at some time before 1670, when northern Indians were invading Maryland and harassing the English settlements and native villages alike, many of the Indians whose lands had thus been invaded sought refuge across the Potomac in Virginia. The different groups may have been rather small, but of sufficient size to cause their old tribal names to become identified with the sites they occupied. SITES OF ANCIENT SETTLEMENTS The locations of many native settlements are shown on the Smith map of Virginia, often referred to as the 1624 map. These differed greatly in size and importance, and some were the villages of chiefs whose influence may have extended to the smaller groups outside the principal settlement. The names of such villages are accompanied on the map by a small representation of a habitation, the ‘ Kings howses ”’ as they were described in the legend. Three such relatively important towns stood on the banks of the Rappahannock within ap- proximately 40 miles of the falls. The first of these settlements to be reached in ascending the river was Pissaseck, on the left or north *T am indebted to William B. Marye, Baltimore, Md., for the references quoted. * According to information furnished by Judge Alvin T. Embrey, Fredericks- burg, Va., this small stream, which enters the left bank of the Rappahannock and which bounds Woodlawn, the old Turner plantation, was formerly known by the name Chingoteague, and was so recorded on several deeds, as in Deed Book 1, p. 620, and Deed Book 1A, p. 147, King George County, Va. The name at once suggests Chincoteague Bay on the Eastern Shore, being part in Virginia and part in Maryland. 16 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 bank ; next above was Nandtanghtacund, on the shore of a deep bay on the opposite side of the stream; and far above was Cuttatawomen, on the left bank. Other settlements, evidently of less importance, are indicated on both banks of the river, but none is shown within about to miles of the falls. It appears that in 1608 there were no Algonquian camps on the Rappahannock for some distance below the great island, westward from which the country was dominated by Siouan tribes, whose steady movement down the valley may have caused the aban- donment of the Algonquian villages. It is thought that some sites indicated on that part of the 1624 map, figure 1, have been identified. These will now be briefly described, beginning with Pissaseck, thence up the river, thus approaching the “bounds betwixt the Kingdome of the Mannahocks and the Nand- taughtacunds.” All are siown on the map, figure 3. PIsSSASECK This large settlement stood on the left or north bank of the Rappa- © hannock, and although not specifically mentioned in the early narra- tives as being one of the native villages encountered in 1608, during the exploration of the river, its position, as well as its importance, is shown on the 1624 map by a “ Kings howse.” It occupied a level area high above the river, below and adjoining an extensive marsh, known as Drakes Marsh, where even now much game and wild fowl is killed each year. It was an ideal site for a native village, and the vast amount of broken pottery and innumerable objects of stone that are and have been recovered from the surface prove it to have been one of great extent and one that had been frequented through genera- tions. The position of the village agreed with the general description of the environment of native settlements as presented by Strachey.” He wrote (p. 70): Theire habitations or townes are for the most part by the rivers, or not far distant from fresh springs, comonly upon a rice of a hill, that they may over- looke the river, and take every small thing into view which sturrs upon the same. Their howses are not many in one towne, and those that are stand dissite [dispersed] and scattered without forme of a street, farr and wyde asunder. The site of the ancient village is shown in the aerial photograph reproduced in plate 2. It stood on the far or left bank of the river -—now expanded and assuming the appearance of a lake—above the 4 Strachey, William, The historie of travaile into Virginia Britannia. Hak- luyt Society, London, 1849. “BINQSYIAIPII J VAOGR S][e} 2Y} 3B purysr ay} 0} UMO}Spaay] Wosy yOouueyeddey sy [—E ‘O17 17 LEEDSTOWN PISSASECK RAPPAHAN NOCK—BUSHNELL $ e kK = g 14 x RS x NS INDIAN SITES ON THE Ounasworusaagus tS NO. 4 18 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 small white arrow which points with the current. The few houses just within the cloud-shadowed area to the right are all that remain of Leedstown. During the days of Indian occupancy the greater part of the land was covered with forests which extended to the edge of the water, and here, not far distant from the river, were the scattered habitations with the fields and gardens nearby. English traders were probably established at or near Pissaseck by the latter part of the seventeenth century, and a “ marte or ffaire”’ may have been conducted in the vicinity. The old road leading to the Potomac undoubtedly follows the route of a more ancient trail that led to the native settlements. The region continued to become of greater importance to the colony and was more thickly settled with the building of warehouses on the bank of the river. Here was soon to be reared the town of Leeds.* Scant traces of the colonial town remain, and these are now encountered intermingled with the stone implements and bits of earthern vessels made and used by the earlier occupants of the region. However, the brick structures erected in the town covered only part of the land that had formerly been in- cluded in the native settlements, assuming the site to have been occupied and reoccupied through generations, long before the coming of the English. Again referring to the aerial photograph: A triangular tract is visible immediately above the white arrow. This is 15 acres in extent and is bordered by the river for more than 1,200 feet. The material now illustrated and described came from the surface of this tract, the greater part of it having been found in the extreme western part— the left as shown in the photograph—a short distance back from the river. Although having been discovered on the surface, exposed by the plow and erosion, it is within reason to believe it is material that had accumulated in refuse heaps, overgrown with vegetation, later to become scattered and spread as the ground was cleared and culti- vated. This hypothesis alone would explain the great variety of flaked objects, believed to indicate different periods of occupancy, ™ The General Assembly, convened at Williamsburg, May 1742, passed an act for the establishment of a town “on the north side of Rappahannock river, in the County of King George, where the church and public warehouses are built . . . The said town shall be called by the name of Leeds.” (Hening, op. cit., vol. 5, pp. 193-197). The area to be occupied, with the placing of streets and buildings, was described. Later, it was known as Leedstown, which became an active center of trade, with wharves and warehouses, whence quantities of tobacco and other products of the colony were sent to England, and where sailing vessels landed supplies for the rich plantations on the Northern Neck. HOASVSSId AO ALIS AHL GNV NMO.1SaGga7q ONIMOHS ‘MOONNVHVddvV¥Y AHL YSAO HLYON ‘sdioy any Away *S *qQ yde1s30j04 eg i Py Se RE: d G ‘Id ‘b ‘ON ‘96 “10A SNOILO31100 SNOANV1ISOSIN NVINOSHLIWS SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO. 4, PL. 3 SPECIMENS FROM THE SITE OF PISSASECK 4 natural size. a-d, U.S.N.M. nos. 378065-68. SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO. 4, PL. 4 SPECIMENS FROM THE SITE OF PISSASECK + natural size. Above, U.S.N.M. no. 378060; below, U.S.N.M. no. 378070. SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO. 4, PL. 5 SPECIMENS FROM THE SITE OF PISSASECK > Natural size. Above, U.S.N.M. no. 378071; below, U.S.N.M. no. 378072. NO. 4 INDIAN SITES ON THE RAPPAHANNOCK—BUSHNELL 19 and the vast amount of fragmentary pottery, now recovered from the site. And, of course, the variety of objects, rather than the quantity, is of the greater interest. To describe the material briefly : Plate 3—Specimens a. Seven examples of projectile points or knives, and two side scrapers, made of dark brown argillite, now weathered and bleached to a light yellowish color. The large broken blade in the lower right corner has recently been fractured, thus revealing the natural shade of the argillite and the extent to which the surface has been altered through long exposure. The condition and appearance of the specimens suggest great age. The specimens closely resemble certain material recovered from sites in the Dela- ware valley, artifacts belonging to the so-called argillite culture, which have been discovered in the yellow sand or soil below a stratum of black surface soil.” Specimens b. Twelve projectile points made of dark, very compact, argillitic slate. These also resemble specimens from the Delaware valley figured and described in the works just cited. Specimens c. Two specimens made of rhyolite of a light bluish- gray color, showing flow structure and slightly altered. A cache of 32 pieces, of the same form and material, many of which are in the United States National Museum (U.S.N.M. no. 756), were found in 1861 on the farm of Joshua Pierce, on Rock Creek, Washington, D. C. The farm surrounded the old Pierce Mill, on the bank of Rock Creek, where it is now crossed below Tilden and Upton Streets. Specimens d. Eight examples of comparatively large blades made of rhyolite. The majority are broken, but the portions remaining reveal the characteristic features. All are very thin and are beautiful examples of flaking. These are altered to a greater degree than are the two specimens c. The stone, containing crystals of feldspar, was evidently derived from quarries in South Mountain, “a high group of ridges extending from near the Potomac at Harpers Ferry to the southern side of the Susquehanna at Harrisburg, Pennsylvania.”” The quarries were discovered by Holmes in 1892,” one, very extensive, “is 75 miles northwest of Washington, and was readily accessible to the inhabitants of Potomac and Patuxent rivers.” Numerous scat- > Compare Hawkes, E. W., and Linton, Ralph, A pre-Lenape site in New Jersey. Anthropol. Publ. Univ. Pennsylvania, vol. 6, no. 3, pl. 17, Philadelphia, 1916. Also illustrations in Spier, Leslie, The Trenton argillite culture, Anthro- pol. Papers Amer. Mus. Nat. Hist., vol. 22, pt. 4, New York, 1918. ** Holmes, W. H., Stone implements of the Potomac-Chesapeake Tidewater Province. 15th Ann. Rep. Bur. Amer. Ethnol., pp. 73-77, Washington, 1897. 20 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 tered specimens and caches of blades have been found in the sur- rounding country. Plate 4——Above, several types of projectile points, knives, scrapers, and drills, made of quartzite. The material was obtained in the form of boulders and pebbles, so plentiful in the region, and is of various colors and degrees of coarseness. The two specimens a are drills which had evidently been inserted in shafts. The points are smoothed and rounded, the effect of the rotary movement of the drill when in use. The point of 0 is likewise worn from use. This and the four specimens to the left in the same row may be classed as scrapers. At the bottom of the plate are six examples of projectile points, or blades, made of diabase. No other similar pieces were found on the site. All are greatly weathered, including the fractured surfaces of the four broken specimens. The great similarity of form and size of the latter is remarkable. These have been altered to a greenish- brown color, are very rough, and resemble certain of the large imple- ments shown in plate 8. | Plate 5.—A very large proportion of all the small flaked objects found on the site are made of white quartz, all of which had been derived from water-worn pebbles. Such pebbles occur here in vast quantities ; the majority are small and so provided an unlimited supply of pieces of suitable size for the making of arrowpoints and of a variety of small implements such as scrapers, drills, and knives. Examples of all are illustrated. Some forms are far more numerous than others, and although a great quantity of broken pieces has been recovered from the site, together with many entire specimens, only one example of the triangular type of point has been obtained. Several varieties occurring on sites farther up the valley, beyond the falls, have not been found here. A single specimen of much interest is shown at the bottom of the plate. It was found a few inches under the surface, some 50 feet back from the brow of the bank that slopes to the river about 15 feet below. It is made of gray flint * which is unaltered. The flaking is ** The specimen was examined by E. P. Henderson, of the Department of Geology, U. S. National Museum, who described the material thus: “This is a gray flint-like material showing banding due to the arrangement of small cavities rather than to a difference in composition of the flint. There are a few brown inclusions of limonite in the gray flint. A light brown siliceous area at one end has not been altered to the flint. In this there are also some brown limonitic inclusions. It appears that the original mass was a nodule of flint with a surface of this light brown material inclosing the gray flint. Since there is a similar inclusion in both portions, it is likely that both have formed at the same time; consequently, the difference is not due to weathering.” NO. 4 INDIAN SITES ON THE RAPPAHANNOCK—BUSHNELL 21 clearly defined, and the edges are sharp. A characteristic feature of the type is the increase in the width of the blade near the middle, and in many examples a few flakes are removed from one edge near the point evidently in the endeavor to make the latter sharper. Specimens resembling this from Michigan, Illinois, Indiana, and Kentucky are preserved in the collections of the United States National Museum. The specimen from Kentucky deserves special mention. It is made of a dark brown flint, so plentiful through parts of Kentucky and Tennessee, and was found in a cave a short distance northeast of Bowling Green, Warren County, Ky. It was found in 1888 and was sent to the museum at that time (U.S.N.M. no. 1062, loan). The specimen is embedded for a distance of about three-quarters of an inch in a fragment of human pelvic bone, which it had entered from the front through the body. Thus it had served as a weapon and had undoubtedly been attached as a point of a spear. This discovery suggests the manner in which all similar pieces may have been used. Nevertheless, if mounted as knives or daggers they would have been serviceable in many ways. The specimen from the site of ancient Pissaseck was undoubtedly made west of the Ohio. POTTERY A vast amount of fragmentary pottery has been recovered from the site. It is of a brownish color, often with coils clearly defined, and for the most part containing a tempering of crushed quartz or sand, some of which may have occurred naturally in the clay. With few exceptions the fragments bear on the outer surface the impressions of nets or single cords; a few reveal the use of closely woven fabrics. A single shard appears to have been decorated with straight, very regular lines, which had been made by impressing some hard material into the clay when the latter was in a plastic state. One fragment, appearing to be older than the majority, bears the impression of basketry.” Plate 6—The nine specimens forming the three upper rows are bits of rims of vessels and show several forms of impressions. The large piece on the top row is part of a rim that measured about 8 inches in diameter. It is coiled ware, and the difference in height of the two ends of the specimen represents the height of one coil of clay **T am indebted to E. G. Cassedy, by whom all drawings were made, for assis- tance in determining the nature of the textiles which were impressed on the pottery vessels. No i) SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 used in building up the wall of the vessel. The fracture has followed the line of contact of two coils. On the right may be seen the bottom of a coil of clay which had been worked down over the one below, but which had not become closely combined. The triangular fragment in the top row shows very deep, sharply defined impressions of single cords which continued over the top of the vessel. The cords passed at an angle over the rim, forming a very rough surface. It is flat and does not have the usual curved surface of a rim, which suggests a vessel with a square opening. The large, worn fragment on the left end of the second row is the only piece found on the site that bears the impression of what is believed to be basketry. The texture of the ware differs from that of the majority of specimens. It is more porous, and on the surfaces are many small cavities caused by the leaching away of the tempering material, thought to have been crushed shells. The rim was indented in a manner somewhat similar to a, next to be described, but had become smoothed from use. The vessel was evidently large. The two forms of rim decoration mentioned are found more clearly defined on other specimens. Specimen a. Coiled ware, with surface greatly weathered and now bleached to a light brownish color. On the surface are many very faint impressions of fine, twisted cords which may have been part of a net. The clay evidently contained a small amount of shell tempering which has decayed, resulting in the few cavities which are now visible on the surface and on the fractured edges. The rim is decorated as shown, exact size, in a, figure 4. Specimen b. Very hard, compact ware, which does not show any evidence of coils of clay in forming the vessel. Tempering very coarse sand, two pieces of which project from the right side of the fragment and are visible in the photograph. The surface bears the impression of a net which had been carried over the top and impressed on the rim. This is drawn, natural size, in b, figure 4. Many speci- mens found on the site were similarly decorated. As previously mentioned, many fragments of earthenware bear the impressions of nets. The meshes of some were small and regular, others were much coarser and made of heavier cords. Specimen c. A very hard ware that contains much sand, some of which is very coarse; one piece of quartz, being as thick as the wall of the vessel, is exposed on both the inside and outside. The surface of this bit of pottery reveals the impression of a beautiful example of net making. The cords are finer than the great majority, and the 23 BUSHNELL ON THE RAPPAHANNOCK INDIAN SITES NO. 4 Fic. 4.—Specimens from the site at Pissaseck. Textiles derived from impressions on fragments of pottery. Natural size. 24 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 knots are more evenly placed. The net, restored, and drawn natural size, is shown in ¢, figure 4. Specimen d. Coiled ware, porous but very hard. There are numer- ous small cavities caused by the leaching away of the tempering material, probably crushed shells. The impression on the surface was formed by a net in which apparently an extra series of small cords had been knotted across the original meshes, thus dividing each of the latter into two triangular sections. This unusual feature is more clearly shown in the drawing of the restored material, reproduced natural size in d, figure 4. Specimen e. A hard, compact ware, very dark in color, with no indications of a tempering material. The impression on the outer surface was made by a closely woven fabric. The warp elements were very slight when compared with the thickness of the woof. The latter had probably been made of the hair or wool of some animal—possibly the bison. Drawn and reproduced natural size in e, figure 4. Specimen f. Coiled ware, compact, with no trace of tempering. This differs from the preceding specimen in that the warp elements are very coarse, appearing to be twisted cords, possibly a vegetal sub- stance. The woof is composed of many small strands. This is restored and drawn exact size in f, figure 4. Plate 7—Specimen a. The ware is very hard and has weathered to a light brown color. It is extremely porous, containing many cavities caused by the leaching away of the tempering material, probably crushed shells. The deep, sharply defined lines on the surface were formed by impressing some hard, rigid object in the clay while the latter was still in a plastic state. This is the only example of this form of decoration encountered on the site. Specimen b is a small knob which appears to have projected from the side of a vessel. The ware is hard, of a reddish color, and contains a small amount of crushed quartz as tempering. The surface bears the impression of twisted cords, evidently a piece of cloth or net which had been pressed against the mass of clay. Nothing similar is known to have been found on the site. Specimens c, d, and e are three fragments, each showing the edge or beginning of the bottom of the vessel, with a small part of the wall attached. All are of a reddish brown color and contain a temper- ing of coarse sand. The two pieces c and d may be parts of the same vessel. On the surfaces are the impressions of cords and nets, the latter being clearly shown in the photograph of c. Specimen f. An example of the curious disks, of unknown use, made of bits of pottery vessels and found widely distributed, although NO. 4 INDIAN SITES ON THE RAPPAHANNOCK—BUSHNELL 25 no similar specimen was found on the sites examined along the Rappahannock. Much crushed quartz had been added as a tempering. The impression of tightly twisted cords remains on the outer surface. It is of a very bright, reddish color, different from that of any other specimen found on the site. The large fragment to the right is part of a massive vessel. It is coiled ware and in places is more than one-half inch in thickness. The surface is hard and slightly pitted, indicating the leaching away of part of the tempering material, although some small pieces of quartz remain exposed. Tightly twisted cords crossed and recrossed the surface, the impressions of which are clearly defined. The lower edge, as viewed in the photograph, is the bottom of a coil or band of clay which had separated from the one against or upon which it had been placed. This parting had occurred while the vessel was still in use, when, it is evident, the parts had been perforated and fastened together by means of a cord or thong. Perforations made for this purpose are frequently found in bits of pottery as well as in frag- ments of soapstone vessels. The specimens of earthenware just described are considered typical examples from the site of Pissaseck. SOAPSTONE The discovery of many fragments of soapstone vessels proves that at some time in the past much of the material was used by the occu- pants of the ancient site. The rather small pieces differ in texture and degree of purity, and consequently could have been derived from more than one quarry. Some may have been carried down the Rappa- hannock from the quarries far up the valley, well within the bounds of the territories dominated by the Manahoac tribes in 1608; other pieces were probably obtained from the extensive outcroppings in Amelia County, or from other sources in the tidewater region which have remained undiscovered. There are no known references by the early writers to the actual use of soapstone by the native tribes of Virginia at the beginning of the seventeenth century,.but it may have been used to a limited extent by some in conjunction with the far more numerous pottery vessels. This condition appears to have prevailed among certain related Algon- quian tribes farther north in New Jersey and possibly elsewhere. Peter Kalm, the Swedish scientist, who’wrote while in New Jersey in January 1749, mentioned the use of pottery and soapstone vessels by the Indians of the region at the time of the earliest European 26 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 contact during the seventeenth century. He wrote” (vol. 1, pp. 343-344) : The old boilers or kettles of the Indians, were either made of clay, or of dif- ferent kind of pot-stones, (Lapis ollaris). The former consisted of a dark clay, mixt with grains of white sand or quartz, and burnt in the fire. Many of these kettles have two holes in the upper margin, on each side one, through which the Indians put a stick, and held the kettle over the fire, as long as it was to boil. Most of the kettles have no feet. It is remarkable that no pots of this kind have been found glazed, either on the outside or the inside. A few of the oldest Swedes could yet remember seeing the Jndians boil their meat in these pots. They are very thin, and of different sizes; they are made sometimes of a green- ish, and sometimes of a grey pot-stone, and some are made of another species of apyrous stone; the bottom and the margin are frequently above an inch thick. The Indians, notwithstanding their being unacquainted with iron, steel, and other metals, have learnt to hollow out very ingeniously these pots or kettles of pot-stone. Four fragments of soapstone vessels, recovered from the site of Pissaseck, are shown in plate 7. The specimen in the upper right corner is a bit of a rim, worn smooth from use; it closely resembles the example from Nandtanghtacund, figure 8. The period of occu- pancy to which the specimens from the Rappahannock sites should be attributed has not been determined, but the statements by Kalm, although treating of the country farther north, suggest the possibility that soapstone may have been used in the Rappahannock villages as late as the beginning of the seventeenth century. Many of the heavier, cruder forms of implements are still to be found scattered over the site of the ancient village. Typical examples are illustrated in plates 8 and 9. Plate 8—Specimen a. An implement made of diabase, surface slightly weathered, and resembling many specimens found on sites above the falls. The groove had been formed by pecking and is smooth from long use. Specimen 6. Axe made of diabase, so greatly altered that it is not possible to distinguish the flaked from the natural surface of the stone. Specimen c. Pestle, made of diabase, with condition of surface similar to that of b. This was evidently a natural boulder which had been shaped by pecking rather than by flaking. Both sides are roughly pitted. Specimen d. Hammerstone; a natural quartzite pebble with the greater part of the surface showing effect of use. * Kalm, Peter, Travels into North America. 2d ed., 2 vols. London, 1772. SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO. 4, PL. SPECIMENS FROM THE SITE OF PISSASECK 3 natural size. U.S.N.M. no. 378073. VOL. 96, NO: 4, PEi7 SMITHSONIAN MISCELLANEOUS COLLECTIONS SPECIMENS FROM THE SITE OF PISSASECK below, soapstone, . Above, pottery, U.S.N.M. no. 378073 1ZE natural si al 2 U.S.N.M. no. 378074. SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO. 4, PL. 8 SPECIMENS FROM THE SITE OF PISSASECK 3 natural size. a-f, U.S.N.M. nos. 378075-80; g, U.S.N.M. no. 378080a. 1 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO. 4, PL. 9 1. SPECIMENS FROM THE SITE OF PISSASECK 3 Natural size. a, U.S.N.M. no. 378081; b-f, U.S.N.M. no. 378082. 2. SPECIMENS FROM NEAR LEFT BANK OF THE RAPPAHANNOCK ABOVE GREENLAWS WHARF 4, natural size. NO. 4 INDIAN SITES ON THE RAPPAHANNOCK—BUSHNELL 27 Specimen e. Hammerstone; a quartzite pebble, about 2 inches thick. Deeply pitted on the two opposite surfaces. Specimen f. Hammerstone formed of a piece of coarse sandstone. Both surfaces deeply pitted. Specimen g. Mortar made of an irregular block of coarse, dark brown sandstone. Plate 9 (upper part ).—Specimen a. Massive implement formed of a piece of coarse quartzite which had been struck from a boulder. It appears to have been attached to a handle and used as a hoe. Specimens b, c, and d. Three hammerstones made of diabasic rocks, with surfaces showing different degrees of weathering. Specimen e. Small quartzite pebble which had been used as a hammerstone. Specimen f. A very roughly chipped implement made of diabase, with surface deeply weathered. CACHE OF TRADE BEADS A cache of trade beads found at Leedstown a few years ago proved to be of great interest. Examples of the beads are shown in plate 1, but before describing them in detail and telling of the locality and manner in which they were encountered, it will be well to refer briefly to the manufacture and use of glass beads by the early colonists at Jamestown. Beads of many sorts were known to the Virginia Indians in pre- historic times ; consequently, those brought by the first settlers—being of a new and unknown material and often of a brilliant color—became of the utmost importance when trading with the natives. So important were they that it was soon decided to manufacture some in the colony, rather than to import all from Europe, and it is evident that quantities were thus produced at a house erected for that purpose not far from Jamestown. On April 30, 1607, before arriving at the site of future Jamestown, some of the English landed and visited the Indian village of Kecough- tan, about the position of the present Hampton, where they were treated kindly by the natives. The English were offered food and, so the narrative continues (p. Ixiv) :* “After we were well satisfied, they gave us of their Tobacco, which they tooke in a pipe made artificially of earth as ours are, but far bigger.” A dance followed and when it had ended “ the Captaine gave them Beades and other trifling Jewells.” ** Percy narrative in Smith, op. cit. 3 28 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 When trading with the Indians, during the summer of 1608 at a time when Powhatan was present, the colonists received a large quantity of corn in exchange for “a pound or two of blew beads.” About this time the first glass was made at Jamestown. The spring of 1609 found the colonists engaged in many endeavors ; houses were being erected, a well was dug, and they “ produced a tryall of Glasse.” This was done at “ the Glasse-house, a place in the woods neare a myle from James Towne ’”’,” which had probably been erected during the preceding year. To what extent beads were then made, or were produced during the years that immediately followed, is not known. The second attempt to establish glass making in the colony was a more serious consideration. On June 11, 1621, at a Court held for Virginia,” ‘‘ Intelligence was given y' one Capt: Norton made an offer & would undertake to pcure 6 straungers skillfull in makinge of Glasse and Beads to goe over to Virginia to be imployed in the saide worke for the Company for no other consideration then onely the halfe profitts of their labo’. . .” The proposition was accepted by the Company on June 13, and Norton agreed to “ carry. over with him 4 Itallyans and two servants of his owne . . . which six psons shall within three moneths after their Arivall in Virginia sett upp a Glass ffurnace and make all manner of Beade & Glasse . . .” But it soon became known that the cost of the undertaking would be greater than anticipated and that the Company “ were not able to goe through wth itt.” It was then deemed best to let it be “ free to pryvate Adventurers to undertake the same . . .” And so Captain Norton was released from his agreement. This was on July 16, 1621, and as stated by “‘certaine adventurers now present they did now acquainte this Courte that itt was not their intent therby utterly to exclude the Company from a buisiness of this speciall consequence unto them all (seeinge the Comoditie of Beads was like to prove the Verie Coyne of that Country) ...” In the “ Instructions to the Governor and Council of State in Virginia,” dated July 24, 1621, is included the following: “ Item whereas Capt William Norton and certaine Itallians now by the general Company and other worthy mynded adventures att a verie 2 Smithy Opa Clty Dao °° Tn The Records of the Virginia Company of London, vol. 1, p. 484. Library of Congress, Washington, 1906. * In The Records of the Virginia Company of London, vol. 3, p. 477. Library of Congress, Washington, 1933. NO. 4 INDIAN SITES ON THE RAPPAHANNOCK—BUSHNELL 29 great charge sent for the erecting of a glasse furnace in Virginia wee hartylie desire yow to afford them all favor possible.” It is evident the Company in London feared that bartering too many beads with the Indians would lessen their relative value, and in a letter to the Governor and Council in Virginia, dated August 12, 1621, stated (op cit., p. 495): “The makinge of beade is one of Capt Nortons cheife employmente w“" beinge the mony you trade w't the natives we would by no meanes have through to much abun- dance vilified or the Virginiane at all pmitted to see or und'stand the manufacture of them: wee therefore pray you seriously to con- sider what proportion of beade can be vented and their worth not abated, and intimate the proportion to Capt Norton and his Italians, and certifie the same to us in yo" next letters, that accordingly we may limitt the quantitie that shall from time to’ time be made.” The importance of beads, and of blue glass beads in particular, was thus acknowledged,” and it is evident that vast quantities were made and traded to the Indians. The use of beads was frequently mentioned in the early records of the colony. One such account of more than usual interest forms part of the Minutes of the Council and General Court.” At a court held at Jamestown, November 8, 1624, Robert Poole told that while on a trading voyage, “ wherein he was ymployed for Mr Thresurer,’” he had bought corn from the Indians, and had exchanged “ thirteene armes length of some beades for Another Tubb [of corn].” “ Further he sayeth y' Capt Croshow gave for a great Canoe w he bought 10000 of blew beades, saying yt he would give Mr Thresurer satis- faction for the beds. “Also he sayeth y' he paide for matts 20000 of blew beads, of w™ matts there was used to seele ye shipp 20, and “Further he sayeth that he gave to the great man of Potuxsea to be their guid to Pocotonck 6 or 800 of blew bead.” From this it may be assumed that innumerable blue glass beads, made at Jamestown, served the treasurer of the Colony as a medium of exchange when dealing with the Indians. Happenings in England: during the early years of the seventeenth century undoubtedly led to the endeavor to establish a glasshouse in * Strachey, in his “ Dictionarie of the Indian Language,” gave “ Blew beades, vnetagwushomon.” *“ The Virginia Magazine of History and Biography, vol. 21, p. 46, Virginia Hist. Soc., Richmond, 1913. 30 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 Virginia. In 1615” the importation of foreign glass into England was prohibited, but 5 years later permission was granted to import “rare and curious glasses.’’ About that time, 1620, “‘ an attempt was made to set up glass works in Scotland,” with workers from Venice. If beads were included among the articles that were not to be im- ported from a foreign country, which would have included Venice, it became necessary to make all that were required for the Indian trade. Such may have been the reason for the establishment of the industry at Jamestown in 1621. DISCOVERY OF THE BEADS Just east of the triangular tract, within the shadowed area in the aerial photograph (pl. 2), the river bank rose normally some 6 or 8 feet above the gravelly beach. At some time in the past the bank had been cut away for a distance of more than 150 feet, and the surface for a like distance back from the water had been graded, thus forming a level area approximately 150 feet square sloping gradually to the edge of the water. In the middle of the far boundary of the graded area, facing and paralleling the river and 8 or to feet higher than the beach, are the remains of an ancient brick foundation, and about midway between this and the river are other bricks which appear to be part of a wall. Locally, and traditionally, this graded square is known as the Old Arsenal, a term which during the seventeenth and eighteenth centuries would have applied to a place where boats were built, also where arms and all military equipments were manufactured or stored. Such was probably the site at Leedstown, with the customhouse nearby. The graded square has long been cultivated. When the ground was plowed early in the spring of 1925, a few beads were found on the surface near the center of the area. That night, as related by an old negro who lives nearby, a heavy rain fell and the following morning many beads were scattered over a very limited space. This caused a search for more, and soon great numbers of beads were encountered a few inches below the soil that had been disturbed by the plow. The 13 varieties of beads illustrated in plate 1 are thought to include examples of all that were discovered in the cache. There may have been others, but if so, they have not been traced. Their history is not known, nor has it been possible to determine when or for what reason ** Nesbitt, Alexander. Glass. New York, 1879. This is one of the South Kensington Museum Art Handbooks. es NO. 4 INDIAN SITES ON THE RAPPAHANNOCK—BUSHNELL 31 they were placed as found. All may have been contained in a wooden keg or box which decayed and disappeared, thus allowing the beads to remain closely embedded in the surrounding earth. The beads will be described and references made to identical or similar specimens from other localities that are now preserved in the collections of the United States National Museum, and to some that have been received at the Museum during the past 2 years for study or identification, but which have not remained in the collections. Many other kinds of beads occur on the sites, but only those resem- bling the specimens from Leedstown will now be considered. Two of the 13 varieties found in the cache prove to be of the greatest interest and later will be described in detail. The first three, top row of the plate, are types that were widely distributed by the traders ; they have been made for centuries and may still be obtained. As they are so numerous and so scattered, no specific references will be made to sites where they have been discovered. Top row.—Left, transparent or translucent glass, light green; middle, opaque glass, black; right, transparent or translucent glass, medium shade of blue. Second row.—Core transparent or translucent green glass, with thin glaze * of red glass covering entire surface. Over the surface of the red glaze are three groups of parallel lines, each group consisting of three lines, alternating white, blue, white. Similar beads: Tennessee, Sullivan County, U.S.N.M. no. 136810; Alabama, Madison County, Hobbs Island in Tennessee River ; Georgia, Bibb County, Macon; Pennsylvania, Bainbridge County, burial, U.S.N.M. no. 35773. Third row.—Core transparent or translucent green glass, with thin glaze of red glass covering entire surface. Over the surface of the red glaze are four groups of parallel lines, each group consisting of three lines, alternating white, black, white. No other examples traced. Fourth row.—Core translucent or transparent green glass, with thin glaze of red glass covering entire surface. * Beck, Horace C., Classification and nomenclature of beads and pendants. In Archaeologia . .. Published by the Society of Antiquaries of London. 2d ser., vol. 27, Oxford, 1928. P. 55: ‘Glaze is a form of glass. It can vary very much in its composition, but it always contains silica and an alkali.” P. 56: ‘“ When a bead has been covered with a thin layer of vitreous enamel or glaze, it is called a Glazed bead.” 32 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 Similar beads: Tennessee, Sullivan County, U.S.N.M. no. 136810; North Carolina, Mecklenburg County, burials, U.S.N.M. no. 138808 ; Georgia, Bibb County, Macon, 1935, burials on plateau, also from surface of Mound D; Florida, Pinellas County, Maximo Point, Tampa Bay, beads somewhat smaller, U.S.N.M. no. 35775; Mary- land, Prince Georges County, burial near Piscataway, U.S.N.M. no. 5839; New York, Monroe County, near Brockport, U.S.N.M. no. 16685. Fifth row —Translucent or transparent dark blue glass, longi- tudinally striped with fine lines of opaque white glass. Similar beads: California, Santa Barbara County, Santa Rosa Island, U.S.N.M. no. 20236. Sixth row—Translucent or transparent dark blue glass. Similar beads: Tennessee, Sullivan County, U.S.N.M. no. 136810; Alabama, Cherokee County, site on Coosa River, U.S.N.M. no. 99217 ; Alabama, Elmore County, near junction of Coosa and Tallapoosa Rivers, site of Fort Jackson, earlier old French Fort Toulouse, U.S.N.M. nos. 91557 and 91564; Alabama, Madison County, Hobbs Island in Tennessee River ; Georgia, Bibb County, Macon, burial on the plateau, 1935; North Carolina, Mecklenburg County, burials, U.S.N.M. no. 138808; Florida, Hillsborough County, near Tampa Bay, U.S.N.M. no. 35335 ; Florida, Orange County, mound, U.S.N.M. no. 150100; Louisiana, Avoyelles Parish, U.S.N.M. no. 331724; Pennsylvania, Lancaster County, on Susquehanna River, U.S.N.M. no. 27048. Seventh row—Cane or tubular beads.” Core translucent or trans- parent green glass, with thin glaze of red glass over entire surface. Similar beads: Tennessee, Sullivan County, U.S.N.M. no. 136812; Alabama, Madison County, Hobbs Island in Tennessee River; Vir- ginia, Stafford County, average smaller, burial from site of Poto- mac village, at mouth of Potomac Creek. Eighth row.—Cane or tubular beads, opaque white glass. * Beck, op. cit., p. 60, described this type of bead: “Cane beads. To make these the glass was made into a rod or tube which was called a cane. These canes were sometimes made of one glass only; at other times they were made of different coloured glasses arrangd in a pattern. “To make a bead, a cane, usually tubular, was selected of approximately the same diameter as the bead required. A piece the length of the bead was cut off this cane. In some cases this was used as a bead without any further work on it. In other cases it was finished by either grinding or reheating. Beads made in this manner are called Cane beads.” —_—— NO. 4 INDIAN SITES ON THE RAPPAHANNOCK—BUSHNELL 33 Similar beads: Tennessee, Sullivan County, U.S.N.M. no. 136812; Alabama, Madison County, Hobbs Island in Tennessee River; Louisiana, Avoyelles Parish, U.S.N.M. no. 369258. Ninth row.—Cane or tubular beads, opaque light blue glass. No other examples traced. Tenth row.—Wire-wound beads,” opalescent white glass. Similar beads: Tennessee, Sullivan County, average smaller, U.S.N.M. no. 136812; Georgia, Bibb County, Macon, 1935, from surface of Mound D; Georgia, Whitfield County, U.S.N.M. no 15539; Alabama, Talladega County, U.S.N.M. no. 364574; Missis- sippi, Lee County, more globular, U.S.N.M. no. 209619; Louisiana, Avoyelles Parish, average smaller, U.S.N.M. no. 331724. Two bottom rows —Cut rock crystal, with eight facets on perimeter. Similar beads: Florida, Pinellas County, Maximo Point, facing Tampa Bay, U.S.N.M. no. 35775; Florida, Hillsborough County, 14 beads from burials near Tampa Bay, U.S.N.M. nos. 35334-35344; Florida, Orange County, U.S.N.M. no. 150100. The references in the preceding lists often apply to a single bead, seldom to more than two or three, which had been found with many others of different types. Some had been discovered in burials, others had been recovered from the surface of village or camp sites. In two instances no examples of similar beads have been traced in the collec- tions of the United States National Museum, although it is to be expected that some are preserved in other collections, both public and private. The relatively few specimens recorded may indicate the range of the several forms, rather than the frequency with which they have occurred. The history of the remarkable cache is not known, nor has it been determined from what country or countries the material may have come. It is doubtful if any of the beads were made at Jamestown; consequently very little can now be added to the brief descriptions already given. Two of the types (fifth row and two bottom rows) present problems which may be difficult to solve, but the solution of which would undoubtedly aid in determining the place of origin of other beads encountered with them in the cache. *” Beck, op cit., p. 60, referred to beads of this type: “ Wire-wound beads. A thin stick of glass heated until it had much the con- sistency of toffee was wound round a wire. During the process the glass was pulled out into a thread, and there is frequently a projection on the bead show- ing where this thread was broken off. When, however, as often happens, the bead has been reheated for subsequent decoration, this projection generally dis- appears. Beads made in this manner are called Wire-wound.” 34 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 Aside from the specimens from the Leedstown cache, the collec- tions of the National Museum contain only one series of the trans- parent blue glass beads with longitudinal stripes of fine threads of opaque white glass paralleling the perforation. These were re- covered from an ancient Indian burial on Santa Rosa Island, Santa Barbara County, Calif., by Stephen J. Bowers in 1876 (U.S.N.M. no. 20236). The islands were discovered by the Spaniards under Cabrillo in 1542 and were visited frequently thereafter by ships under the same flag. Although the beads from the bank of the Rappahan- nock are corroded, with their surfaces roughened, those from Cali- fornia appear as fresh and smooth.as when placed in the graves. This contrast in condition of the two groups may be attributed to the difference in the amount of moisture, the variation of temperature, and the composition of the earth by which they were surrounded. All are thought to be of the same age and to have come from the same source—some glasshouse in Spain. Cut rock crystal beads, similar to those recovered from the Leeds- town cache, have been discovered on the west coast of Florida, but are not known from any other part of the United States. They have been found in mounds and burials within a limited area extending southward from the north shore of Tampa Bay to the vicinity of Key Marco, a distance of about 175 miles. Other forms of beads and pendants, likewise made of rock crystal, have been discovered in Florida associated with the type occurring at Leedstown, but only the latter will now be considered. The specimens mentioned as having come from Pinellas and Hills- borough Counties were collected by S. T. Walker in 1879.* Many others were found during the progress of recent archeological ex- plorations under the direction of the Bureau of American Ethnology, supervised by M. W. Stirling. They were encountered in several localities, within the bounds previously designated, and all came from burials thought to have belonged to the years following soon after the first contact of Spaniards and Indians about the last half of the sixteenth century. Whether they had been traded or given to the Indians, or had been recovered by the natives from a Spanish wreck on the Gulf coast, may never be known; however, they were un- doubtedly brought to America in a Spanish ship. The crystal beads from the cache on the Rappahannock and those from the west coast of Florida are identical in form and size. They * Walker, S. T., Preliminary explorations among the Indian mounds in southern Florida. Jn Ann. Rep. Smithsonian Inst. for 1879, Washington, 1880. az: NO. 4 INDIAN SITES ON THE RAPPAHANNOCK—BUSHNELL 35 have the same number of facets on the perimeter. On all, the plane surface from which the perforation was begun is rough, suggesting the use of a saw in preparing the mass, and the opposite end of the perforation emerges in a distinct concavity which was made when the drill broke through the thin wall of crystal. The two distinct types of beads found in the Virginia cache—the blue glass with fine white lines and the cut rock crystal—should, with a degree of certainty, be attributed to a Spanish source, and this suggests that all beads in the Leedstown cache were of Spanish origin. During the sixteenth and seventeenth centuries Spain produced glass equal to that made in other parts of Europe ; consequently, it is inconceivable that beads required for trade in foreign lands would have been brought from other countries. Barcelona was the center of the glass industry, which would undoubtedly have included the manufacture of beads. Thus it had been for centuries, and in the words of Sefor Juan F. Riano, in a catalogue of Spanish objects in the South Kensington Museum: Jeronimo Paulo, who wrote in 1491 a description in Latin of the most remark- able things at Barcelona, says they also send to Rome and other places many glass vessels of different sorts and kinds, which may well compete with those of Venice. Marineus Siculus, who writes at the beginning of the sixteenth cen- tury, says that the best glass made in Spain is that of Barcelona; and Gaspar Baneiros in his chronographia, published at Coimbra in 1562, mentions that excellent glass was made at Barcelona, almost equal to the Venetian. These were the years during which Spanish vessels so frequently touched the coast of Florida and had intercourse with the native tribes. Although it has not been possible to trace definitely the source of the crystal beads, it is believed they were cut in Spain and brought to America during the latter half of the sixteenth century. The beads, crystal as well as glass, found in the cache on the bank of the Rappahannock, may have been carried to Virginia in a ship under the English flag, but when, where, and how they had been obtained by the English would form an interesting bit of history. KERAHOCAK A village bearing the name of Kerahocak is shown on the 1624 map several miles above Pissaseck. It stood on the left bank of the Rap- pahannock opposite Port Tobago and Green Bays, probably about the position of Greenlaws Wharf, which is visible in the aerial photo- graph, plate 10, and is also indicated on figure 5. Here as elsewhere, ” Quoted by Nesbitt, op. cit., pp. 101-102. 36 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 fragmentary pottery, arrowpoints, and other artifacts have been collected. Three specimens found a few years ago” a short distance above the wharf are illustrated in plate 9, figure 2. They were discovered several feet below the surface, not far from the river bank, and may have been associated with a burial. All are finished objects with edges smoothed from use. Specimen at left is an unusual form and well chipped. The material is dark rhyolite with many small phenocrysts of quartz and feldspar. Specimen at right is made of grayish quartzite, probably derived from a pebble found nearby. The edges are greatly worn, and the entire surface appears smooth from long use. Specimen at bottom is a piece of exceptional interest. It is made of fine-grained rhyolite, with flow structure, of a grayish color. As viewed in the photograph the right lower edge, near the point, reveals the effect of much use, being worn to a greater degree than is any other part of the surface. It had evidently been mounted and used as a knife or dagger, with the left end, for a distance of several inches, inserted in a handle. Length 8 inches. The surface is only slightly altered and in this respect resembles that of the two specimens c shown in plate 3. The rhyolite was probably derived from a quarry in South Mountain, south of the Susquehanna, in Pennsylvania. NANDTANGHTACUND As previously mentioned, this is the name of a large village that appears on the 1624 map. It is placed on the right, or south, bank of the Rappahannock, on the shore of a large bay, the early Indian name of which has not been preserved but which is now known as Port Tobago or Port Tobacco Bay. Just below the site of the village, which is indicated by a “ Kings howse”, is the name of a smaller settlement which may have been part of the larger village. Nandtanghtacund “ was evidently seen by Captain Smith late in December 1607, or early in January 1608, while he was held captive by the Indians and before he was taken to Powhatan at Werowa- comoco. He was conducted to many native villages and, as one narrative states (p. 398) :* they led him to the Youthtanunds, the Mattapanients, the Payankatanks, the Nantaughtacunds, and Onawmanients upon the rivers of Rapahanock, and Pa- “ Now in the private collection of F. M. Aldridge, Fredericksburg, Va. “The spelling of names often differed in the text from the forms appearing on the map; the latter is followed in the present article unless quoted literally. “Writings of Capt. John Smith. Arber ed. NO. 4 INDIAN SITES ON THE RAPPAHANNOCK—BUSHNELL 37 tawomek; over all those rivers, and backe againe by divers other severall Nations, to the Kings habitation at Pamaunkee: where they entertained him with most strange and fearefull Conjurations. This was before the exploration of the Rappahannock by the colon- ists during the summer of 1608 and may explain the presence of certain names on the map issued in 1624. Traces of a very extensive settlement, a site that had been long frequented or often occupied, have been encountered on the eastern shore of Port Tobago Bay, where much material, including frag- mentary pottery, is to be found scattered over the surface of a wide area extending eastward to the marsh bordering Green Bay. This was the site of ancient Nandtanghtacund, later occupied by the Porto- 2 Ss H S. GREENLAWS ONE MILE Fic. 5.—Map indicating position of Nandtanghtacund, later Portobago village. The area included within the dotted lines is shown in the aerial photograph, plate Io. 7 bacco, which is so clearly shown in the aerial photograph reproduced in plate 10, and indicated on the map, figure 5. Nandtanghtacund was a settlement of importance during the days of Indian occupancy of the valley, but as it lay far beyond the frontier of the colony until after the middle of the seventeenth century it may have been seldom visited by the English before the native population had become greatly changed from what it was in 1608, at which time Smith referred to “ Nantaughtacund having 150 men”’, indicating a total population of approximately 700. In October 1669 “ An act ffor destroying Wolves ” became neces- sary in the colony. This read in part: “ “ Jt is enacted that the Indian “ Hening, op. cit., vol. 2, p. 275. 38 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 tributaries be enjoyned and assessed to bring in a certaine number annually”, and in the list that followed, in Rapahanock County, were first the Portobaccoes having 60 “ Bowmen or hunters”, who were assessed 12 heads; and secondly the Nanzcattico and Mattehatique who together had 50 men and were assessed 10 heads. The Porto- baccoes are assumed to have occupied the site of ancient Nandtangh- tacund, with the Nanzcattico living on the opposite side of the river. The position of the Portobacco village is shown on the Herrman map, issued in 1673, a detail of which is reproduced in figure 2. Rappahannock County was created in 1656, at which time it ex- tended on both banks of the river, but in the year 1692 it was divided, and that portion that stood on the right bank became Essex County and that on the north, or left, bank, Richmond County. Caroline County was formed in 1728 and included part of Essex County. Tobago or Tobacco Creek, from the southwest shore of Green Bay, became the boundary between Essex County on the east and Caroline County on the west; consequently the greater part of the site now being considered is just within Caroline County. Another reference to the villages is found in a document that was prepared about 30 years after the appearance of the Herrman map. It occurs in a communication to the Board of Trade in London, dated Williamsburg, Virginia, July 18, 1702.“ The principal rivers of the colony are mentioned, together with the names of the Indian tribes whose villages stood on their banks. Following Rappahannock River: “ Indians and No. of.—Portobago or Nanzattico, 30 [men] ; Wicocomoco.” Thus at the beginning of the eighteenth century the remaining Portobago and Nanzattico were considered as one, and they probably occupied a single village. Jefferson, in his references to the native tribes of Virginia,” in- cluded the Nantaughtacunds—thus following the spelling found in Smith’s text—whose chief town had in 1608, so he believed, stood on Port Tobacco Creek in Essex and Caroline Counties. In this Jefferson clearly identified the ancient site with the association of the two names. Being able to trace the settlement from the earliest days of the Colony, when primitive ways of life prevailed, it is gratifying to present a brief description of the village as it was in the year 1686, soon after contact with the English had caused many changes in the “In The Virginia Magazine of History and Biography, vol. 1, no. 4, p. 363. Virginia Hist. Soc., Richmond, 1894. * Jefferson, Thomas, Notes on the State of Virginia, Philadelphia, 1788. NO. 4 INDIAN SITES ON THE RAPPAHANNOCK—BUSHNELL 39 customs of the native inhabitants. However, outwardly the group of bark-covered habitations, with their surrounding fields and gardens, and the dugout canoes drawn up on the sandy beach, did not differ in appearance from that of many small villages first seen by the colonists in 1607. PORTOBAGO VILLAGE, 1686 A brief narrative of a visit to an Indian village on the bank of the Rappahannock, in Virginia, is contained in the journal of a young French Huguenot who traveled through the colony during the autumn and winter of the year 1686. Durand reached Virginia in October 1686 and, after varied experi- ences in a new country where he had difficulty in making himself understood, continued northward through Gloucester and Middlesex Counties to the right bank of the Rappahannock. He then crossed the Rappahannock to Lancaster County, thence up the left, or north, side of the stream, through a country sparsely settled and with roads that followed Indian trails. “And so we entered into the county of Rappahannock. ... We went the next day to Portobago, for so is called Mr. Wormeley’s “ rich plantations in this vicinity.” This was an extensive tract, with 8 or 9 houses. “I perceived also that about two thirds of these lands were in timber, and others in prairies which I was told, were the plantations that the Savages occupied 5 or 6 years ago. Three of these Savages came to visit us as soon as we arrived, they brought him two large wild turkeycocks and one tame one, the wild ones weighed about 40 pounds apiece. We saw their village on the other side of the River, and the next day having expressed a wish to see them at home, Mr. Wormeley ordered three horses sent over the water.” That afternoon they crossed the river to the south, or right, bank where the Indian village stood, and having mounted “ rode ““The title page of the small book, in French, does not reveal the name of the author; however, it is stated in the text that he was a native of Dauphiné and a member of the Huguenot family of Durand. The title page reads: Voyages d’un Francois, Exilé pour la Religion, avec Une Description de la Virgine & Marilan dans L’Amerique. A La Haye, Imprimé pour |’Autheur, 1687. The volume, with the exception of the part treating of the Indian village which is now translated and presented, was privately printed in English by Fairfax Harrison, 1923. Ralph Wormeley (1650-1701), of Rosegill on the lower Rappahannock, where Durand had remained a few days. He was a burgess from Middlesex in 1674, advanced to the Council in 1677, and Secretary of the colony from 1693 until his death. 40 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 through all lands on that side of the river, which were greater in extent than those on the north side where we were lodged... . Having explored all of this region we went to the Indian village.” The Indian settlement which had been seen from the left bank of the river when looking across the stream is thought to have stood on the eastern shore of Port Tobago Bay, on the site of ancient Nand- tanghtacund, from which it may not have differed greatly in appear- ance. Durand’s brief description of the village and its people, as they were in December 1686, just two and one-half centuries ago, follows: These Savages have rather pretty houses made of tree bark the walls as well as the roofs, so well put together with thongs of deerskin that neither rain nor wind disturb them at all. They are a people darker than the Gipsies (Egiptiens) we have in Europe. They mark their faces with cuts in the shape of a snail shell, in which they put powder, and thus they are marked for life.* The women of the house wear only a deerskin to cover their least decent parts, in winter they wear the hairy side next the skin, and in summer put skin to skin. They build their fire in the middle of the house, their beds around it they inter-weave a certain strong grass which they find along the river resembling straw,” and they do it on four little forks (quatre petites fourches), these serve them for seats to sit upon. In the village the men wear only a wretched shirt of white or blue cloth, and from the time they put it on they never take it off, it falls off of them in tatters, for they never wash anything. Aside from the deerskin the women are entirely naked the rest of their bodies, their little children completely naked, no matter how cold it is. The men do nothing except hunting and fishing, the women raise Indian corn (Bled Sarrasin) which is common among them, any one takes it who needs it.” They make also pots and vases from earth and pipes to smoke, the Christians buying their pots or vases fill them up with Indian Corn “This refers to tattooing, but to what extent it was practiced by the people of the Rappahannock is not known. From Durand’s statement it may be inferred that he had seen both men and women who had followed the custom. But Smith did not mention men when he wrote some years before; he stated (op. cit., p. 66) : “They adorne themselves most with copper beads and paintings. Their women some have their legs, hands, brests and face cunningly imbrodered with diverse workes, as beasts, serpentes, artificially wrought into their flesh with blacke spots.” ” A sedge, Scirpus americanus Pers. Quantities of this grow in shallow water near the banks of the river. ° This is rather vague. Adair (History of the American Indians, London, 1775), who wrote of the tribes farther south, stated that (p. 430): “ Formerly, the Indian law obliged every town to work together in one body, in sowing or planting their crops; though their fields are divided by proper marks, and their harvest is gathered separately .. .” William Bartram (Travels through North and South Carolina ... London, 1792), referred to the same customs in these words (p. 510): “In the spring, the ground being already prepared, on one and the same day, early in the morning, the whole town is summoned, by the sound of a conch shell, from the mouth of the overseer, to meet at the NO. 4 INDIAN SITES ON THE RAPPAHANNOCK—BUSHNELL 4I and that is the price,” they all smoke as well as the men, but they do not raise any tobacco, it is given them in exchange for game or fish.” They marry among themselves but it is only to avoid confusion among the children, for as soon as a young man takes a wife he builds a little house, leaves his father and mother and retires to it. They have some knowledge, but a very imperfect knowledge, of the true God, they believe that he is creator of all that they see, and of the growth of what is necessary to life, but that he does not lower himself so far; that the demons which are inferior to him were created for that purpose, and so they fear them because they are from time to time tormented by them. They have no other ceremony in their marriages unless it be the assembling of the village, and the man having chosen she whom he wishes to take gives her a roe’s foot or a deer’s, and.she gives him an ear of corn, which signifies that the husband will keep the house provided with meat and the woman with corn.” The Ministers of this region take no pains to convert them to Christianity and instruct them, although the greater part of them know how to speak Eng- lish. When we left them they made a present to Mr. Wormeley of a dozen deerskins, and to Mr. Parker and to me a handful of pipes each. public square, whither the people with their hoes and axes; and from thence proceed to their plantation, where they begin to plant, not every one in his own little district, assigned and laid out, but the whole community united begins on one certain part of the field, where they plant on until finished; and when their rising crops are ready for dressing and cleansing, they proceed after the same order, and so on day after day, until the crop is laid by for ripening. After the feast of the busk is over, and all the grain is ripe, the whole town again assemble, and every man carries off the fruits of his labour, from the part first allotted to him, which he deposits in his own granary; which is individually his own.” But a large crib was provided in the town in which grain was placed, volun- tarily, and in such quantity as the individual chose, which served as “a public treasury . . . and to which every citizen has the right of free and equal access, when his own private stores are consumed.” Durand may have known of a similar custom prevailing at Portobago village. *t Evidently the value of the corn that a vessel would hold was the accepted value of the vessel itself. As previously recorded, an act passed the Assembly in June 1676 that allowed the English to obtain various articles from the Indians, mentioning “ canooes, bowles, matts or basketts, and to pay the said Indians for the same in Indian corne, but noe other commodities.” * Nicotiana rustica has been identified as the plant formerly raised by the Indians of Virginia. Discussed by W. A. Setchell in Amer. Anthrop. vol. 23, no. 4, 1921. 8 Adair (op. cit., pp. 139-140) referred to this ceremony: “ When the bride- groom marries the bride, after the usual prelude, he takes a choice ear of corn, and divides it in two before witnesses, gives her one half in her hand, and keeps the other half to himself; or otherwise, he gives her a deer’s foot, as an emblem of the readiness with which she ought to serve him: in return, she presents him with some cakes of bread . . . Formerly, this was an universal custom .. .” However, Adair was writing of conditions and customs in the country far south of Virginia and many years after Durand visited the village on the Rappahannock in 1686. 42 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 ¥ Since it was already dark we called the boat to take us away, because we needed considerable time, the River being very wide, for it carries even in this place vessels of six tons, although distant 30 leagues from the sea. The small Indian village, occupied in December 1686, is thought to have stood on the eastern shore of Port Tobago Bay, the region designated A on the map, figure 5, and so clearly shown in the aerial photograph reproduced in plate ro. MATERIAL FROM THE SITE OF NANDTANGHTACUND It is known that during past years innumerable arrowpoints and other small flaked objects, stone implements and weapons of many forms, as well as quantities of fragmentary pottery, have been re- covered from the site of the ancient village. But all does not belong to the same period of occupancy ; some specimens are thought to be far older than others. Much may even now be found, revealed by the plow as the surface is gradually lowered’or when the bank facing the water falls away. As a result of constant plowing and cultivation of the land through generations the pottery has been reduced to small fragments and many of the stone objects broken. Even in this con- dition, however, all prove to be of interest. In the endeavor to show examples of different types of objects that have been recovered from the area, several specimens found some years ago™ have been included and described with others col- lected during recent visits to the site. Nevertheless, it is realized that few of the many forms that would have been encountered in the village are included in the present sketch. Examples illustrated in plates 11, 12, and 13 will be briefly described : Plate 11.—Specimen a is a small axe or weapon, chipped and deeply weathered. Material, a diabasic rock. This is attributed to an early period of occupancy that preceded, possibly by centuries, the historic village. Specimen b. One of several similar celts found on the surface near Port Tobago Creek, area B on map, figure 5. Thick, oval in section, smoothed and polished through use. Material, a very dark, greenish diabase. Specimen c. Double-edged implement, worn and polished. Material, diabase. Specimen d. Grooved axe, found on the surface near shore of Port Tobago Bay. * Specimens b and c, plate 11, and a, plate 12, are in the collection of F. M. Aldridge, Fredericksburg, Va. 1457 NO AVG ODOVEOL LYOd HLIM GNNOVLHONVYLGNYN JO ALIS ‘sdio) ary Autry “S ‘-Q ydes30j04g OL “Id ‘bt ‘ON ‘96 “10A SNOILO3S1109 SNOANVTIISOSIN NVINOSHLIWS SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO. 4, PL. 11 SPECIMENS FROM THE SITE OF NANDTANGHTACUND 5 natural size. d, U.S.N.M. no. 378083. SMITHSONIAN MISCELLANEOUS COLLECTIONS VOES 916, NO.4, sPES 12 SPECIMENS FROM THE SITE OF NANDTANGHTACUND a and b, 4 natural size; c, + natural size. c, U.S.N.M. no. 378084. SMITHSONIAN MISCELLANEOUS COLLECTIONS FRAGMENTS OF stage we CE we VOL. 96, NO. 4, PL. POTTERY FROM THE SITE OF NANDTANGHTACUND at 2 natural size. U.S.N.M. no. 378085. 13 NO. 4 INDIAN SITES ON THE RAPPAHANNOCK—BUSHNELL 43 Specimen e. Grooved axe, found on area B, near Port Tobago Creek. Fractured, but the edges were later worn smooth by use. Specimen f. Hammerstone; a natural quartzite pebble with ends battered, showing evidence of use. Plate 12.—Specimen at top is an unfinished banner stone of excel- lent workmanship. The surface, now greatly weathered, had been partly polished, and the perforation had been started. Material, a very hard greenish hornblende gneiss. Found on the surface near Port Tobago Creek. The small white quartz objects shown in the plate include several characteristic forms, but the beautifully chipped specimen in the middle of the top row is unusual. It was found near the left bank of Port Tobago Creek, and within a short distance a similar specimen of the same size, made of a dark crystalline rock, was discovered. A massive mortar, found on the surface near the shore of Port Tobago Bay, is shown at the bottom of the plate. Material, a dark, reddish brown coarse sandstone. It is between 4 and 5 inches in thickness and about 12 inches in length. POTTERY Twenty-seven fragments of earthenware are illustrated in plate 13, of which number 22 were discovered near the shore of Port Tobago Bay, many being found along the face of the bank that rises from ' 5 to 7 feet above the gravelly beach. The remaining 5 pieces, includ- ing the first, third, and fifth from the left in the top row, and specimens c and g, were found on the surface near the left bank of ’ Port Tobago Creek, the area designated B on the map (fig. 5). It is evident that the shoreline of the bay has changed during the two and one-half centuries that have elapsed since the site was visited by Durand. The waters of the bay have encroached upon the land then occupied by the village, and consequently the ground that now forms the shore may, in 1686, have been far back from the water. Numerous pieces of pottery, stone implements, and other objects of Indian origin, have been found in the water some distance from the bank, and it is within reason to think that similar material occurs farther out, in deeper water, where it had been deposited when the soil was washed away. Therefore the 22 specimens shown in plate 13 came from a part of the settlement that had been many feet from the shore. The surface extending to the edge of the bank has been cultivated for generations, which readily explains the fragmentary condition of the pottery. The larger pieces, however, were found on 4 44 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 the beach and in the face of the bank below the level where the earth had been disturbed by the plow. The 11 specimens shown in the upper part of plate 13 reveal various kinds of surface decoration including the impression of a fabric, of twisted cords, and of another material, and deep lines. The frag- ments vary greatly in texture and are thought to have been made during different periods of occupancy. The five pieces forming the top row are bits of rims of vessels and may be described briefly, beginning at the left: 1. Coiled ware. Tempering, crushed quartz. iS) Coiled ware. Tempering, sand or crushed quartz. 3. No evidence of coils. Clay contains a quantity of fine sand which was probably natural. Vessel is a very light yellow with top or rim flat. Thickness at top 5/16 inch. 4. Coiled ware. Tempering, small amount of crushed quartz, and evidence of crushed shell which has decayed and disappeared, causing small cavities to remain. 5. No indication of coils. Much very fine sand mixed with the clay. Vessel of light color. Second row, beginning at left: 1. No trace of coils. Clay contains much fine sand or crushed stone. Rather fragile with surface weathered a light gray. NN No trace of coils. Tempering, sand or crushed stone. 3. No evidence of coils. Very dark inside with surface weathered light gray. Tempering, coarse sand. 4, 5, 6. Three pieces similar to others found higher up the Rappa- hannock opposite the large island at the falls. When describing the specimens found opposite the island it was suggested that a vegetal substance had served as the tempering. This belief had been suggested by the presence in one piece of small cavities filled with particles of carbon which, when leached away, would have caused the porous condition of the ware. However, the occurrence of a piece of shell, + inch in width, exposed in the middle of the edge of the largest of the 3 fragments, at right end of the second row, proves conclusively that crushed shell had been employed as a tempering material, either solely or in conjunction with some other substance, all traces of which have disappeared. The pitted surface, and the cavities throughout the ware, resulted from the disappearance of the tempering material, whatever it may have been. NO. 4 INDIAN SITES ON THE RAPPAHANNOCK—BUSHNELL 45 The indented lines on the surfaces of the fourth and fifth speci- mens from the left were formed by some hard, smooth, and very regular object that was impressed upon the plastic clay before the vessel was fired. The lines on the fifth example are sharply defined and reveal a median ridge extending the entire length. Several frag- ments of pottery bearing the same impressions were found on sites above the falls of the Rappahannock; it therefore appears to be a recognized form of ornamentation practiced by the potters of the region. The fragments of pottery illustrated in plate 13 have been chosen from many pieces discovered on the site. The 16 specimens desig- nated by the letters from a to p, inclusive, have been selected to determine as far as possible, with a degree of certainty, the form of basketry or type of fabric that had been impressed upon the plastic clay when the vessel was made. The pottery is shown one-half size, but the fabric and basketry that is revealed by the impressions is restored and drawn exact size. The letters attached to the drawings reproduced in figures 6 and 7 are the same as those on the specimens in plate 13 from which the impressions were derived. They may be described as follows: Specimen a. Coiled ware. Tempering crushed quartz. A piece of exceptional interest, as it proves the use of a rigid basket in shaping a pottery vessel. This fragment came from near the base of the vessel as is indicated by the convergence of the warp elements. This is clearly shown in the drawing at top of a, figure 6. As the basket widened, additional warp elements were inserted, to be held together by what may have been a continuous coil of a more pliable material. The inside of the fragment is next shown with the line of contact of two coils of clay indicated by the broken line. Below is sketched a view of the fragment from above, with a bit of the basket, restored, pressed against the outer surface of the vessel. The convex surface of a coil which is exposed in the fragment is indicated in the drawing, looking from above. Specimen b. No evidence of coil. Tempering, crushed shell or vegetal substance, all of which has disappeared. The impression of coiled basketry is clearly shown on the surface. Specimen c. No evidence of coil. Tempering leached away, caus- ing the ware to be very porous. The surface bears the impression of coiled basketry. Specimen d. Coiled ware. No trace of tempering. The impression suggests a coarsely made matting, having a foundation of flat splints 46 CPOO- Yi SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 Mp> =~ pos = nth. sale : Tro =, LIT . mae AIT FFF za oe ah a {1 \ im << LT awa Ny > p< Zz et a To we Ae Ftc. 6.—Specimens from the site of Nandtanghtacund. Textiles derived from impressions on fragments of pottery. Natural size. RAPPAHAN NOCK—BUSHNELL INDIAN SITES ON THE NO. 4 x hey Lh] LY \/'\ Z vy k-S SAG WOK RKYINKK/ DRS XO NRN ODS. K <= ———— \ \ —=— 4 — ¥ a3 \ SS Pr ss ie ‘ese SB ———— 4 { SSS t] \ a AS =) ) Led SoS 5 \ } J —— 4 | \) aS \ \\ S) \ >= >— —~ htacund. Textiles derived from Fic. 7.—Specimens from the site of Nandtang impressions on fragments of pottery. Natural size. 48 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 or of reeds that had become crushed and flattened, held together loosely by twisted cords. An example of twined weaving. Specimen e. No indication of coil. A small amount of fine sand mixed with the clay may be natural. It has not been possible to determine the type of vessel of which this is a fragment. The surface shown in the photograph is part of the inside of the vessel, and is very dark and smooth. On the right is the outer wall on which the impression appears. The outer surface is a yellowish brown color. Plain twine weaving. Specimen f. Coiled ware. No evidence of tempering. Very hard. The impression on the surface was made by a piece of fabric of twined openwork weave, the unusual detail of which is that the horizontal cords appear to have been plaited rather than twisted. Other examples of braiding or plaiting have been recorded. Holmes” illustrated a fragment of pottery from Georgia on which the decoration had been formed by impressing plaited cords upon the plastic clay, forming parallel straight lines. And during recent work in Manatee County, Fla, “ charred fragments of braided and woven hair ” of beautiful workmanship were found in a burial pit discovered in a mound not far from Little Manatee River.” Consequently, braiding such as is represented on the small piece of pottery from the Rappahannock site, although rarely encountered, is not unknown. Specimen g. Coiled ware. Tempering, a small amount of crushed stone or sand. The surface bears the imprint of an example of twined openwork weave, having a crossed warp, differing in this respect from f. Specimen h. Coiled ware. Clay impure and contains much fine grit. The impression of a fabric similar to that shown on f occurs on the surface, but the cords were twisted, not plaited. Specimen 7. No evidence of coil. Tempering, a quantity of fine sand or crushed quartz. The impression on the surface was formed by a piece of openwork fabric less regular in weave than the majority of examples and made of extremely fine threads. Specimen j. Coiled ware. Much fine grit mixed with the clay. This is a fragment of the rim of a vessel. The impression on the surface suggests coiled basketry, with twisted cords forming the woof passing over and under a foundation made of some firm but flexible material. * Holmes, W. H., Prehistoric textile fabrics of the United States derived from impressions on pottery. Jn 3d Ann. Rep. Bur. Ethnol., p. 423, 1885. * Stirling, M. W., Smithsonian archeological projects conducted under the Federal Emergency Relief Administration, 1933-34. Jn Ann. Rep. Smithsonian Inst.) 16345 pegoleLoss: NO. 4 INDIAN SITES ON THE RAPPAHANNOCK—BUSHNELL 49 Specimen k. No indication of coil. No tempering. The impression on the surface was made by a rather closely woven fabric formed of coarse, loosely twisted cords, probably bison hair. In texture this resembles charred fabrics found in mounds in the Ohio valley, as well as work of historic tribes.” Specimen /. Coiled ware. No evidence of tempering. The impres- sion on the surface is very unusual and may have been made by either a basket or a piece of woven fabric. Two groups of woof elements, which may have been twisted cords, crossed as they passed over and under parallel warp. The straight lines, which were added after the impressions had been produced, are similar to those on the fourth and fifth specimens on the second row already mentioned. Specimen m. Coiled ware. Tempering, small amount of sand or crushed quartz. At the crossing of many of the cords forming the net meshes there is a short depression in the clay, made when it was in a plastic condition. This is difficult to explain by any knot. It may not have been a part of the net proper but fragments of small feathers that had been attached to the meshes, an art practiced by the Indians of Virginia as well as by neighboring and related tribes both north and south. An account of the Delaware ” and their native arts during the years 1654-1656 sheds light on customs that may not have differed from those practiced by the Indians of the Rappahannock during the same years. Lindestro6m wrote in part when referring to their use of feathers (p. 221): ‘“ They also make very fine and beautiful quilts of painted bird feathers. In the first place they tie them with meshes like nets, yet very fine; then they fasten the feathers in the meshes, so neat and strong that not one feather can come loose from it; it would sooner go clear off.’’ This was also translated “ The feather would sooner break off than come loose.” Were small pieces of quill to remain ‘in the knots of the net, the impression in clay would resemble that on the surface of m. Specimen 2. No evidence of coil. Tempering, crushed shell or vegetal substance, all traces of which have disappeared. Very hard. The impression on the surface was made by an open, netlike fabric. This was made by holding a single woof element between two strands * Compare Holmes, W. H., Prehistoric textile art of eastern United States. In 13th Ann. Rep. Bur. Ethnol., p. 36, pls. 6, 7, 1806. 8 Lindestrom, Peter, Geographia Americae with an account of the Delaware Indians, based on surveys and notes made in 1654-1656. Translated from the original manuscript with notes, by Amandus Johnson. The Swedish Colonial Society, Philadelphia, 1925. 50 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 of the warp which were twisted as they advanced with the woof between. Specimen o. No indication of coil. Tempering, small amount of shell or vegetal substance, leached away, causing cavities. The very distinct impression of a coarse net appears on the surface. Specimen p. Ware similar to 0, but having the impression of a finer mesh. Part of the rim of a vessel. Thus it has been possible, by means of impressions on bits of pot- tery, to gain some knowledge of the art of weaving and textile making as practiced by the inhabitants of the ancient village. SOAPSTONE Many fragments of soapstone vessels are encountered on the site, and it is evident that the material was used extensively by the people Fic. 8—Fragment of a steatite vessel from site of Nandtanghtacund. 1 natural size. U.S.N.M. no. 378003. of the village. Here, as at Pissaseck, vessels made of soapstone may have been used as late as the beginning of the seventeenth century. The most interesting piece of any sort found on the site is a frag- ment of a steatite vessel. This was discovered partly embedded in sand at the edge of the water of Port Tobago Bay, near the position of the upper A on the map, figure 5. Two views of the fragment are shown in figure 8, one-third natural size. The specimen appears to have had a very narrow base—so narrow that it could not have stood in an upright position without support. One conical perforation near the edge indicated that it had been broken and the parts held together by cords before being lost or discarded. But the most unusual feature is the knob in the middle of the narrow end, and this, although greatly worn, suggests a human face or head, with lines on the surface of the bowl both above and below. Undoubtedly, a similar ornament or NO. 4 INDIAN SITES ON THE RAPPAHANNOCK—BUSHNELL SI handle was placed on the other end of the vessel in the same relative position. The greatest thickness of the wall is 1 inch, just below the knob. The entire surface, both inside and outside, is very smooth, resembling the surface of the rim fragment shown in plate 7, found at Pissaseck. The villages of Pissaseck and Nandtanghtacund were important centers of population at the beginning of the seventeenth century and so continued for some years after the settlement of the colony. The sites of both native settlements are very extensive, and it is evident that they had existed through generations, although the entire area of each as now recognized may not have been occupied at any one time. From year to year, with the return of the people to the villages, different places would be selected for the erection of their habitations, a custom that resulted in a very extensive tract showing evidence of having been occupied, thus causing the sites to appear to be those of very large, widely dispersed settlements. Such a moving about over a restricted area may have continued for a long period and readily explains the occurrence of varied types of objects in many localities. The material found on the two important sites just mentioned may be considered characteristic of the region. It is known, however, that certain forms of artifacts that would formerly have been encountered are not represented in the collections illustrated, but the deficiency will be partly supplied by showing specimens from sites farther up the river which will be briefly described. Nevertheless, were it pos- sible to create large, representative collections from other places on the banks of the Rappahannock above Pissaseck, there would be a marked difference in the appearance of the material as a whole, and some types that are numerous on one site would be lacking or seldom encountered on others. AxnovE Port Tosaco Bay The right bank of the Rappahannock, at and immediately above the beginning of Port Tobago Bay, is bordered by an extensive marsh. That the high ground adjacent to the marsh was once occupied by a native settlement is indicated by the presence of pottery, now reduced to very small pieces, scattered over the surface. Various stone imple- ments may also be found, together with a few pieces of soapstone vessels. The extent of the occupied area could not be determined, but it may have been comparatively small. Many of the shards discovered on the surface of the site are smooth, others bear the impressions of cords and textiles, but the majority are so worn that it is not possible to identify the kind of 5 52 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 material that had been impressed on the plastic clay. One fragment is a very hard, dark-colored coil ware, with no trace of tempering. The impression on this suggests that of a rather loosely woven basketry, with both warp and woof formed of firm material, not twisted cords. This is reproduced and drawn exact size in a, figure 9. The wide mouth of Mill Creek is located on the right bank of the river about 2 miles above Port Tobago Bay. A mile or more from the river the creek is crossed by the main highway from Fredericks- burg. A small camp site was discovered on the left bank of the creek immediately south of the road. Several argillite points and fragments of very coarse, net-marked pottery were found on the surface. The pottery appears to have belonged to not more than two vessels, and although worn and greatly disintegrated, several of the shards reveal _Fic. 9.—Textiles derived from impressions on fragments of pottery. a, from right bank of the Rappahannock near Port Tobago Bay. b, from camp site on Mill Creek 1 mile above its mouth. Natural size. the impression of a net which resembles specimen n from Nandtangh- tacund. This was made by the woof elements being held firmly between two strands which were twisted together as they advanced to form the warp. The net, as restored and drawn exact size is shown in b, figure 9. The argillite points are deeply weathered. Traces of many small camps similar to the one just mentioned— camps of hunters away from the larger villages—could undoubtedly be discovered near springs and on the banks of streams throughout the wooded region. Lert BANK OF THE RAPPAHANNOCK ABOVE Port Tosaco BAY The names of two settlements appear on the 1624 map on the left bank of the Rappahannock a short distance above the deep bay later known as Port Tobago or Tobacco Bay. The first is Papiscone and just beyond is Assuweska. One of these is thought to have stood at the mouth of Chingoteague Creek, but it is not possible to determine SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO. 4, PL. 14 Photograph U. S. Army Air Corps. 1. NORTH OVER THE RAPPAHANNOCK SHOWING MOUTH OF CHINGOTEAGUE CREEK 2. PIPES FOR SMOKING. MADE OF CLAY. FROM SITE AT MOUTH OF CHINGOTEAGUE CREEK Natural size. U.S.N.M. no. 378086. VOL. 96, NO. 4, PL. 15 SMITHSONIAN MISCELLANEOUS COLLECTIONS SPECIMENS FROM SITE AT MOUTH OF CHINGOTEAGUE CREEK + natural size. ‘TautoS yYysta toddn ur ‘ys019 IeUI0JOg JO YoU Avo “TOATY IMeUO}JOT ‘}Jo] OF MOIIB J9AO Yue zYst uo ‘uno Ossojeyey }YS4 0} Mosse daoqe ‘yueq jo] UO ‘AIT HOONNVHVddva AHL YSA0 HLYON ‘sdioy ary Autry *S ‘FA qdessojoyg a ae ne aes 9L “Id *b “ON ‘96 “10A SNOILO31100 SNOANVITISOSIN NVINOSHLINS SMITHSONIAN MISCELLANEOUS COLLECTIONS VOLES 9165) NOi4 Platz SPECIMENS FROM RAPPAHANNOCK SITES Above, from Gay Mont. Below, from Cleve. 4 natural size. NO. 4 INDIAN SITES ON THE RAPPAHANNOCK—BUSHNELL 53 the exact position of either. Traces of a village have been discovered on the narrow tongue of land which extends upward from the junction of the two streams and which is bordered by the left bank of the river and the right bank of the creek. This is visible in the aerial photo- graph, plate 14, figure 1. Several large springs issue from the river bank, which rises in places more than 20 feet above the gravelly beach. The site commands a beautiful view of the winding valley, and as evidences of occupancy are encountered for a distance of more than a quarter of a mile, it may have been occupied and reoccupied for many years. It is known that a vast number of small chipped objects, also many larger specimens and quantities of fragmentary pottery, have been recovered from the site of the ancient village. Much may even now be found on the surface, but as the area has been cultivated for many years, the pottery has been reduced to bits, and few entire arrowpoints or objects of any sort are to be discovered. Small tubular clay pipes have been found within a rather restricted area—just above the small arrow shown on the aerial photograph. One perfect and two fragmentary examples are illustrated in plate 14. All are made of clay without tempering, are of a light brownish color and very hard. The perforation, made when the clay was in a soft condition, had been produced by the insertion of a comparatively rigid, firm substance, probably a slender twig or stem of some plant. The discovery of pipes in considerable numbers at once recalls the statement made by Durand that when leaving the Indian village of Portobago, to return to the Wormeley house on the north side of the river, he and another member of the party had received as a gift “a handful of pipes each.” He had previously told that the women of the village were then making “ pots and vases from earth and pipes to smoke.” This was during the autumn of 1686, and it is within reason to believe that the pipes to which he then referred were similar to those now being considered. Consequently, this primitive form of pipe was made and used in the Rappahannock villages just two and one-half centuries ago. The bowls are so small that it is difficult to understand how the pipes could have been used unless the leaves of the plants were rolled and inserted in the opening, rather than being crushed and employed in the usual manner. Additional material from the site is shown in plate 15. Although the pottery now found scattered over the surface has been broken into small pieces by the constant cultivation of the land for many years, larger specimens have been encountered exposed on the face of the river bank below the upper stratum of soil that had been dis- turbed by the plow. One such piece is illustrated in plate 15. This 54 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 large fragment is part of a vessel that measured approximately 11 inches across the rim, with the wall + inch thick. The vessel had probably been about 15 inches in height, with a small, conical bottom. It is coiled ware, with a coarse sand tempering. The inner surface is crossed by many small, parallel grooves caused by the object or material used in smoothing the clay when the vessel was formed. The outer surface had first been marked by tightly twisted cords placed vertically, overlapping, and passing over the top of the opening, caus- ing the edge to be very rough and irregular. Six single cords, parallel, form a band about 1 inch in width just below the rim. A single cord is placed parallel to these 1 inch below. Groups of four cords extend diagonally across the intervening space, the groups of four being about 13 inches apart. All had been impressed on the clay while it was in a plastic condition and before the vessel was fired. The impres- sions of the cords are easily distinguished in the photograph. Three smaller fragments are shown to the right of the large speci- men. The second bears the impressions of cords placed horizontally as on the large fragment, but the rim had been indented, or rather scalloped, by means of a heavy cord pressed at intervals into the clay. The ware is similar to the large fragment, and, like it, has a tempering of coarse sand. The specimen at the upper right bears the impression of a net over part of the surface, but on the left there are several vertical impressions that appear to have been made by matting or basketry. This combination suggests that a piece of net was placed against some rigid material before the impression was made in the clay. The third of the small fragments bears the impression of a net, similar to specimens from Pissaseck and elsewhere. A vast number of small shards were found scattered over the culti- vated land, but only three were net-marked, the great majority being similar to the large piece first described. Four examples of the larger chipped objects, found on the surface during past years,” are shown on plate 15. All are finished specimens with edges worn and smoothed from use. The knife, placed hori- zontally at the top of the group, is made of dark, brownish rhyolite as is also the large blade at the left. The second piece is made of yellow- ish quartzite, and that on the right is diabase with surfaces greatly altered. These were selected for illustration from a number of similar specimens as being typical examples of rather unusual material. *° Now in the private collection of A. J. Jones, of Jersey, King George County, Va. NO. 4 INDIAN SITES ON THE RAPPAHANNOCK BUSHNELL 5 on At MoutsH or MILLBANK CREEK Millbank Creek flows into the Rappahannock from the north between 3 and 4 miles above Chingoteague Creek. The present Mill- bank Creek is believed to have been shown on the Herrman map as Omoy Cr., the first large creek below the ‘‘ Doogs Indian.” Jiles Point, at the junction of the streams, extends between the right bank of the creek and the left bank of the river. The land facing the river is high, but on the other side it is lower and is bordered by a marsh through which the creek flows. The point of land, well protected by the water, is thought to have been occupied by one of the villages indicated on the 1624 map. Much fragmentary pottery and numerous arrowpoints may even now be collected from the surface, and mor- Fic. 10.—Point resembling the Folsom type made of diabase. Found near mouth of Millbank Creek. Natural size. U.S.N.M. no. 378004. tars, axes, and other large specimens are often found, all tending to indicate the location of an ancient settlement. The most important piece known to have been discovered on the site is a Folsom point, made of dark gray flint, which has already been described.” It is a typical example of the eastern form and was found on the surface in a small field, at the extremity of the point of land, that slopes to the marsh bordering the creek. No similar speci- men is known to have been found in the vicinity, but nearby, in the same field, was discovered a point made of diabase which closely re- sembles the Folsom type. This is shown, natural size, in figure Io. It is crude when compared with the first specimen, but, considering ® Literary Digest, June 9, 1934. Also in Bushnell, D. I., Jr., The Manahoac Tribes in Virginia, 1608. Smithsonian Misc. Coll., vol. 94, no. 8, pl. 13, Oct. 9, 1035. 56 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 the material of which it is made, it is doubtful if a better piece could have been produced. The surface is altered to a far greater degree than are the blades made of the same stone from Pissaseck illustrated in plate 4. The two specimens just mentioned are thought to have belonged to the same period of occupancy or to have been lost by the same nomadic bands. The third field beyond the mouth of Millbank Creek is visible on the extreme right in the aerial photograph, plate 16. This is the lower part of Cleve, the old house having stood in the midst of the grove, near the river bank, a short distance to the left. Much broken pottery and innumerable objects of stone have been found on the surface of this field, which is just above the white arrow in the photograph. The two celts illustrated in plate 17 were found here. Both are oval in section, symmetrical, and are made of a hard, greenish diabase. The small hammerstone shown with the celts was found in the field nearer the mouth of the creek. It is a natural quartzite pebble battered at both ends. The fragments of pottery are mostly small, few being more than an inch in length. One such piece bore the impression of a net, several were smooth, all others were cord-marked. The ware resembles in all respects that discovered near the mouth of Chingoteague Creek, on the same side of the Rappahannock several miles below. CHECOPISSOWA Much material has been collected from the surface of the fields bordering the right bank of the river, an area clearly shown in the lower left quarter of the aerial photograph reproduced in plate 16. It is part of Gay Mont. This was the locality of a large native settle- ment, and although it is not possible to be positive in the identification of the ancient sites, it may have been the village of Checopissowa, as indicated on the 1624 map. The three small grooved axes and the cylindrical pestle, the latter made of fine-grain quartzite, which are illustrated in plate 17, were found on this site near the river bank. Many of the smaller objects known to have been collected here, and which have been examined, are exceptionally well chipped and represent a variety of forms. The majority of the small pieces, including the projectile points, are made of white quartz. Fragmentary pottery is plentiful, and many pieces bearing the impression of what is thought to have been rigid basketry are found intermingled with ware that obviously belonged to a recent period of SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO. 4, PL. 18 Photograph U. S. Army Air Corps. 1. East over the Rappahannock. Ancient village site to right of black arrow which points with the current. The Potomac River in the distance. 2. Above, from right bank of the Rappahannock north of Taliaferro Mount. U.S.N.M. no. 378087. 4 natural size. Below, axe from below mouth of Dogue Run. 4 natural size. SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO. 4, PL. 19 ican Tet Ft ee _ OL aa eee Army Air Corps. 1. Looking across the site, up the valley of the Rappahannock, with Lamb Creek in foreground. Photograph U. S. Photograph U. S. Army Air Corps. 2. Junction of Lamb Creek and the Rappahannock. SITE OF CUTTATAWOMEN SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO. 4, PL. 20 SPECIMENS FROM THE SITE CF CUTTATAWOMEN ! natural size. U.S.N.M. nos. 378088-91. SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96, NO. 4, PL. 21 Two SPECIMENS, SHOWING RECHIPPING, FROM SITE ON LEFT BANK OF THE RAPPAHANNOCK, 1 MILE ABOVE MOUTH OF LAMB CREEK 2 natural size. a, U.S.N.M. no. 378092. NO. 4 INDIAN SITES ON THE RAPPAHANNOCK—BUSHNELL 57 occupancy. Some vessels had been decorated with a roulette, and other shards are similar to the three illustrated on the right, second row, in plate 13. As a whole the pottery from this area differs from that discovered at Pissaseck, but resembles much of the ware found at Nandtanghtacund and at sites farther up the Rappahannock beyond the falls. TALIAFERRO Mount Again referring to the aerial photograph of the region, plate 16, Taliaferro Mount is on the right bank, heavily timbered, immediately above the white arrow which is pointing downstream. The mount rises abruptly from the water and is rather steep on the north, but slopes gradually to the south and west. This commanding spot was evidently occupied by the house of Robert Taliaferro, the same to which Lederer referred when he wrote :” On the twentieth of August 1670, Col. Catlet of Virginia and my self, with nine English horse, and five Indians on foot, departed from the house of one Robert Talifer, and that night reached the falls of Rappahanock river in Indian Mantapeuck. A few years later Beverley mentioned” “ Tolivers Mount, upon Rappahannock River”, as one of the elevations from which he had “view’d the Country all around over the Tops of the highest Trees, for many Leagues together.’ It was evidently a place of renown and had been frequented by the Indians long before the coming of the colonists. John Fontaine, who accompanied Governor Spotswood on his journey to the mountains during the summer of 1716, left Williams- burg November 9, 1715, for the German settlements on the Rappa- hannock above the falls.” Three days later they reached the home of Robert Beverley, “this Beverley is the same that made the History of Virginia.” There they remained until the 19th, and that day, so the journal continued: “ About three we came to a place upon Rappa- hannock River, called Taliaferro’s Mount, from whence we had a feeble view of the Appalachian Mountains, and a fine view of the river, which is navigable for large ships and has several fine islands * Lederer, John, The discoveries of . . . Begun in March 1669, and ended in September 1670. London, 1672. Reprint 10902. “ Beverley, Robert, The history and present state of Virginia, book 2, pp. 11-12, London, 1705. * Journal of John Fontaine. Jn Memoirs of a Huguenot family, by Ann Maury. New York, 1853. 58 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 in it... We saw upon the river abundance of geese, ducks, and water- es: Although a “ feeble view ” of the high ground and hills to the north- west may have been obtained, the Appalachian} Mountains, as now designated, could never have been seen. Much material of native origin has been recovered from the area bordering the right bank of the river and extending northward from the foot of Taliaferro Mount. Projectile points and other small chipped objects are plentiful, and typical examples are illustrated in plate 18, figure 2. All shown are made of white quartz. Small pieces of pottery may be found in the same fields near the river bank, indi- cating the site of a native settlement. “ Doocs INDIAN ” About 2 miles above Taliaferro Mount the Rappahannock makes a wide bend to the eastward, passing fertile fields which are frequently bordered by low marsh lands. This is shown in the aerial photograph reproduced in plate 18, figure 1, with the camera pointing east. Dogue Run, flowing between heavily timbered banks, joins the Rappahannock _ from the east (above the white arrow) opposite the end of the large marsh. The name of the run was derived from that of the Doeg tribe after they are thought to have come from Maryland and entered the valley of the Rappahannock. In 1673 this was the country of the “ Doogs Indian” as indicated on the Herrman map, figure 2. It is a beautiful section of the valley and was once a vast forest where game abounded. Traces of Indian occupancy are encountered in all parts of the region, and some of the stone implements and weapons which have been found on the surface are unusual forms and many reveal superior workmanship. The axe illustrated in plate 18, figure 2, was found in the field on the left bank of the river visible in the right center of the photograph. It is a natural pebble of diabase which had been slightly worked into shape. The shallow groove is very smooth, a condition which suggests considerable use. Larger, rather massive specimens of the same type have been found on the site. Innumerable arrowpoints and other chipped objects have been found on the surface of the field. Some rather large specimens are often discovered. Typical examples are sketched in figure 11." Of the five specimens shown, four are made of quartzite, and one, that ° These are in the private collection of William Howard, Fredericksburg, Va. NO. 4 INDIAN SITES ON THE RAPPAHANNOCK—BUSHNELL 59 in the upper right, is made of diabase. All are well chipped and symmetrical, and resemble certain specimens from the site at the mouth of Chingoteague Creek, previously mentioned. The small grooved axe shown second from top in plate 17, is similar to the specimens from the field on the left bank of the river Fic. 11.—Specimens found near Dogue Run. 4} natural size. below Dogue Run, just described. It likewise was fashioned from a natural pebble of diabasic rock, and the groove has become smoothed and worn from use. This distinctive type of weapon or implement was not encountered elsewhere on the Rappahannock sites. It is believed to have belonged to a late period of occupancy, as may also many of the projectile points, small chipped objects, and fragments of earthenware found in the vicinity. 6 60 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 OpposITE THE Moutu or HoucH CREEK Hough Creek enters the left bank of the Rappahannock about 1 mile above Dogue Run. The Indian name of the creek is not known. The mouth of the creek, bordered by brush and timber, is visible in the aerial photograph, plate 18, figure 1, to the left of the black arrow which points with the direction of the current. On the opposite side of the river, to the right of the arrow, is the site of an ancient settle- ment which appears to have been abandoned before the coming of the English in 1608. No village is indicated in this locality on the 1624 map. The extent of the occupied area has not been determined. It was probably a permanent village as distinguished from a tempor- ary camping ground. The land is comparatively low and during the freshet of April 1937 was under several feet of water. The material discovered on the site includes many projectile points and several types of small chipped objects, the majority of which are made of white quartz ; consequently, it is impossible to determine the relative age of a specimen by the condition of its surface. But other pieces formed of diabasic rock and a very fine-grain quartzite have become greatly altered and in this respect resemble the two axes found a short distance above on the opposite side of the river, shown in plate 21. A few bits of pottery that bear the impression of bas- ketry, and several fragments of soapstone vessels found near the river bank, likewise suggest an early period of occupancy. CUTTATAWOMEN The position of the village of Cuttatawomen, standing on the left bank of the Rappahannock, is indicated on the 1624 map by a sketch of one of the “ Kings howses,” proving it to have been the home of a chief. However, in 1608 it was not a large settlement, as it was then described as having but 20 bowmen. Mooney was of the belief that the ancient village stood at or near the mouth of Lamb Creek. Traces of a native settlement have been discovered at the junction of the creek and the Rappahannock and may indicate the location of Cuttatawomen. This beautiful site, on the left bank of the river just above the mouth of the creek, is shown in aerial photographs repro- duced in plate 19. The land has been cleared and cultivated for many years, much of it for two centuries or more, before which it had been occupied by native settlements with small clearings for gardens and cornfields. It is known that during past years large collections of stone imple- ments and numerous artifacts of various forms have been recovered NO. 4 INDIAN SITES ON THE RAPPAHANNOCK—BUSHNELL 61 from the surface of the site. Some may even now be found, and specimens discovered during recent visits are illustrated in plate 20. All were collected in the fields bordering the left bank of the river, rising 20 feet or more above the water, extending to the vicinity of the old wharf which is visible in the lower of the two photographs (pl. 19). The pottery found on the site has been reduced to small fragments ; characteristic shards are shown in plate 20. Many of the fragments are believed to be ware that belonged to an early period of occupancy, preceding that of the historic villages of 1608. The impressions on the surface of some pieces appear to have been made by rigid bas- ketry which had been used in shaping the vessels. Crushed shell had evidently served for a tempering material. No example of net-marked ware, so plentiful at Pissaseck and elsewhere, was found on this site. The two specimens shown below the pottery are made of diabase. To the left is a knife or scraper, formed of a flake struck from a pebble and revealing secondary chipping. To the right is a well-made celt with a sharp cutting edge. The surface is slightly altered. All pieces illustrated in the lower half of the plate are made of white quartz. The beautiful triangular points are of special interest, as few specimens have been found on other sites, a single example having been discovered at Pissaseck. At the bottom are six roughly formed scrapers or implements that had been used for different pur- poses, representing several types, all of which bear evidence of much use. A short way above its junction with the Rappahannock, Lamb Creek makes a sharp bend to the left (pl. 19, lower figure). This is caused by a narrow tongue of land. sloping down from the high ground, which rises some 20 feet above the water. The sloping ridge is covered with trees and brush. A large tree, standing separate and apart from others and with a very light spot on the ground to the left of it, at the edge of the cultivated land, may readily be dis- tinguished in the photograph. This large tree is growing from an embankment and ditch which may be traced from the foot of the high ground at the beginning of the marsh bordering the river, over the ridge and down to near the edge of the marsh through which the creek flows. The embankment and ditch follow a straight course. As viewed in the photograph, the embankment is on the far side, facing the level, cultivated fields. The best-preserved section is near the large tree, on the highest part of the ground, where the embank- ment is approximately 18 inches in height and the ditch about the same in depth. 62 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 Nothing is known of the origin of the embankment and ditch. They are very old and may have been constructed during days of Indian occupancy, when the embankment would undoubtedly have been surmounted by a palisade. Nothing similar has been encountered on other Rappahannock sites. SOCKBECK About 1 mile up the Rappahannock from the mouth of Lamb Creek, on the same side of the river, is the site of a settlement which gives the impression of being very old. It is near the river, on cultivated land just beyond the cloud-shadowed area shown in the upper of the two photographs reproduced in plate 19. It is a beautiful location which had probably been occupied and reoccupied through genera- tions and may have been the position of Sockbeck as placed on the 1624 map, but, as previously mentioned, it is not possible to be certain of the identification of any site. The extent of the site was not determined. Some very interesting materal was discovered on the surface, in- cluding arrowpoints and other small chipped objects, crude hammer- stones and pestles, and two axes which are the most unusual speci- mens found on any site mentioned in the present work. A few bits of pottery are to be found scattered over parts of the site, being more plentiful near the river, but all have been reduced by the plow to very small pieces. The two axes are shown two-thirds natural size in plate 21. They were found in the same field a short distance back from the river bank. Both are made of fine-grain, dark gray quartzite. Specimen a. When originally made, this was a grooved axe, or axelike implement. It was probably a quartzite pebble that was shaped by pecking or battering the surface until it was reduced to the desired form and the groove made. Then it was lost or abandoned for a long period, during which time the surface became much altered. Later it was partly rechipped, but the groove no longer served its original function, although the implement was again hafted and used. The rechipped portion is slightly altered and is easily distinguished from the more deeply weathered surface of the original axe. A small portion of the natural surface of the pebble remains and is visible far to the left in the photograph of the side of the implement. Specimen b. Made of a thin quartzite pebble. It had been formed by pecking, but part of the natural surface of the stone remains on both sides of the axe. The surface became greatly altered after having NO. 4 INDIAN SITES ON THE RAPPAHANNOCK—BUSHNELL 63 been unused for a long period. Evidently, it was later discovered and resharpened by the removal of several large flakes. The exposed surface thus produced is worn and smoothed from use, and even though it has been in this condition for three centuries or more, the surface remains unaltered, in strong contrast to the appearance of the older portion. Quartzite weathers or alters so slowly that the present condition of the two specimens is suggestive of great age. But the surface of the secondary chipping of the two pieces has not altered to the same degree, and it is evident the first specimen has been exposed much longer than has the second. The surface of the later chipping on a has become altered, whereas that of b has remained unchanged. The first may be the older of the two, and both periods during which it was used may have preceded that represented by the historic Algonquian tribes, whose villages stood in the valley of the Rappahannock in 1608. Beyond this site the left bank of the Rappahannock is bordered by an extensive marsh, and on the opposite side are cliffs rising nearly 100 feet above the water. From the brow of the cliffs it is possible, when conditions are favorable, to see the Blue Ridge Mountains far to the westward. Evidence of Indian occupancy is encountered throughout the region, which had been occupied and reoccupied through centuries. But it is evident that the greater part of the valley from here to the falls had been abandoned by the Algonquian tribes before the summer of 1608, at which time the English from Jamestown entered the country. CONCLUSIONS SUGGESTED BY CERTAIN SPECIMENS Early in the seventeenth century, when the English ascended the Rappahannock, that part of the valley now being considered was comparatively thickly settled, with camps and villages standing on the banks of the river. But, as proved by the discovery of certain objects, the native tribes then encountered were not the first to in- habit the country. The specimens recovered from the sites vary in form and material, and obviously belonged to several distinct periods of occupancy. The occurrence of a typical eastern Folsom point near the mouth of Millbank Creek, on the left bank of the Rappahannock, presents an interesting problem. It is a beautiful example of the type that has been discovered widely dispersed throughout the country east of the Mississippi, being more numerous in some localities than in others 64 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 but all possessing the same characteristic features, similar in many details to the western points. It is a highly specialized form and one believed to have developed from a single center rather than to have arisen independently in different localities at different times. The discoveries made by Roberts” at the Lindenmeier Site in northern Colorado during the past 2 years have established the great antiquity of the Folsom culture in that part of the country, placed by some at from 8,000 to 10,000 years; by others these figures are considered conservative. But not until a similar site has been un- covered in the eastern part of the country will it be possible to suggest, with a degree of certainty, the age of the eastern Folsom points, as represented by the example from the valley of the Rappahannock. It will be necessary to find specimens resting in an undisturbed, strati- fied deposit that will permit the geologist to determine the approximate age of the superimposed or surrounding mass. The discovery of such a deposit would be of the utmost value in the attempt to determine the antiquity of man in the eastern part of the continent. Next in importance to the Folsom point are the two axes illustrated in plate 21. Both bear evidence of having been used during two long-separated periods, and in this respect they are similar to another specimen found some years ago on the bank of the Rapidan.” And as has been suggested in the description, both periods during which one of the specimens from the Rappahannock site was used may have preceded that represented by the tribes encountered by the first English colonists. Some material recovered from the site of ancient Pissaseck may likewise have belonged to an early period, although this will be more difficult to prove. This refers in particular to the argillite points and scrapers illustrated in plate 3, which are similar to specimens recovered from the stratum of yellow sand, below black soil, in the Delaware valley. The objects which have been mentioned—a Folsom point, the two axes, and the argillite material—afford conclusive evidence of the existence of several distinct periods of occupancy in the valley of the Rappahannock preceding the coming of the historic tribes, those whose villages stood on the banks of the river at the beginning of the seventeenth century. * Roberts, Frank H. H., Jr., Additional information on the Folsom complex. Smithsonian Misc. Coll., vol. 95, no. 10, 1936. * Bushnell, David I., Jr., Evidence of Indian occupancy in Albemarle County, Virginia. Smithsonian Misc. Coll., vol. 89, no. 7, pl. 2, 1933. NO. 4 INDIAN SITES ON THE RAPPAHANNOCK—BUSHNELL 65 Earthenware vessels continued to be made in the Rappahannock villages long after the arrival of the English colonists. There is a reference by the French traveler Durand to the making of “ pots and vases from earth” at the Portobago village during the autumn of 1686, and the art may have been practiced into the eighteenth century. Much fragmentary pottery is found on all sites along the river, but it is difficult, often impossible, to distinguish the more recent from the early forms of ware. A few scattered pieces bear the impression of basketry. These are attributed to an early period of occupancy and may have been contemporaneous with the axes previously mentioned. The impressions of textiles on fragments of pottery from Pis- saseck, Nandtanghtacund, and other localities serve as a means of determining the kind of nets, mattings, and coarse fabrics made and used in the native villages. Much of the ware so marked and now encountered on the surface was made during the late period and indicates the skill of the people as weavers and potters. All material described in the preceding pages was discovered on the surface, and the greater part of it may have accumulated in refuse heaps in and about the villages, later, when the land was cleared and cultivated, to be scattered by the plow. It is believed that much re- mains hidden beneath the surface, implements and ornaments, pottery vessels, fireplaces and traces of the habitations, and burials. The recovery of such material would make it possible to form a clearer conception of the manners and customs, the ways of life, of the in- habitants of the Rappahannock valley long before the settlement of Jamestown, and would aid in determining whence they came. (From “ Explorations and Field-Work of the Smithsonian Institution in 1937”’) PNCIENT SITES ON THE BANKS. OF THE RAPPAHANNOCK IN VIRGINIA By DAVID I. BUSHNELES JR: Collaborator in Anthropology, U. S. National Museum During the spring of 1937 a great freshet swept down the valley of the Rappahannock from the foot of the Blue Ridge, where heavy rains had fallen for several days. The low grounds were inundated, and when the waters had receded it was discovered that many areas had been greatly altered, gullies had been formed, banks of sand had been deposited and, in some instances, the surface soil for a depth of a foot or more had been washed away, causing the heavier masses to settle and remain exposed. Traces of ancient camps and villages were thus revealed. The valley of the Rappahannock below Fredericksburg is more open, the low grounds more extensive, and the river is wider than above the falls; nevertheless, the force of the great flood was felt far down the stream. On May 9g, when the ground was becoming dry, several places be- low the falls were again visited to see the effect of the flood and to endeavor to recover some of the material thought to have been de- posited by the swirling waters. Fortunately, a site of much interest was encountered. It is on the right bank of the Rappahannock in Caroline County, Va., a mile or more below the mouth of Lamb Creek on the opposite side of the river and is shown in the aerial photograph reproduced in figure r11, to the right of the black arrow which points with the current. This is now a rich and fertile section of the valley, but visualize the same region as it was in 1608, when first visited by English colonists from Jamestown—a dense forest with small groups of mat- and bark-covered lodges dotting the river banks, trails travers- ing the wilderness, game and wildfowl to supply the wants and re- quirements of the native hunters armed with bows and arrows, and streams teeming with fish. However, the earlier settlement, traces of which were uncovered by the freshet, is thought to have been aban- doned before the year 1608. When the site was visited on May 9, the surface for a distance of a hundred yards or more from the river bank, and extending to the beginning of the wooded area, was sand and gravel, all vegetation had been swept away, and in places it was deeply gullied. Fragments of 8 107 108 SMITHSONIAN INSTITUTION pottery and objects of stone, some broken and others entire, were scat- tered over the surface, indicating the location of a native village which had been occupied centuries ago. Examples of the specimens col- lected at that time are shown in figure 112. Above are 15 objects made of white quartz, so plentiful in the valley. Projectile points, knives, and scrapers are included in the group. The four pieces to the left in the second row may have been mounted as knives and their simi- Fic. 111.—Looking down the Rappahannock. Site of the ancient settlement to the right of the black arrow. (Photograph by U. S. Army Air Corps.) larity in form and size is remarkable, but being made of quartz the condition and appearance of the surface does not aid in determining their relative age. Below are 9 pieces representing a variety of forms, all made of diabasic rock and with surfaces equally altered as a result of long exposure. At the bottom is a cylindrical pestle, with a short, shallow groove clearly shown in the photograph. Two forms of scrapers may be recognized. All specimens, quartz and diabase, are thought to be of approximately the same age. Some fragments of earthenware found on the site bear the impression of coiled basketry, and this is considered the oldest form of pottery occurring in the Rappahannock valley; other pieces are cord-marked and some are smooth, porous, and deeply pitted through the leaching away of the 109 / XPLORATIONS, 193 1d SMITHSONIAN ) size. natural G settlement. the ancient of Specimens from the site IIo SMITHSONIAN INSTITUTION crushed shell which had served as the tempering material. Bits of soapstone vessels were also found. The types and condition of the objects discovered suggest that this was a permanent village rather than a temporary camp, and the uniformity of the weathering makes it appear that all articles of stone were made and used about the same time. Later in the year several sites farther up the river, which had like- wise been exposed by the spring freshet, were visited and examined. The material discovered, much of which differs from that now figured, will be described and illustrated at another time. Floods have been recorded ever since the country was settled by the English, when much of the heavy timber was cleared away and the ground was cultivated and leveled. The loosened earth was often inundated and gullied, as during the spring of 1937, and although the masses of refuse which had accumulated in and about the native vil- lages during different periods of occupancy were once distinct and stratified or separated, all became intermingled by the force of the waters. This readily explains the variety of objects, made of various materials, often encountered on the same site. Such conditions pre- vailed not only on the Rappahannock but in the vicinity of other streams as well. Some years ago a beautiful example of the eastern form of Folsom points was discovered near the left bank of the river a few miles below the site just described. Unfortunately, it was found on the surface, not beneath it, but this is not significant because, as explained above, the clearing and cultivating of the land enabled the periodic flood waters to change the contour of the land rapidly, and the Folsom point may therefore have once been well below the surface. The occurrence of the point in this region may be accepted as proof that man was here many centuries ago, although just how early he reached the country eastward from the mountains will be impossible to deter- mine until more evidence is available. During the year 1937, as for several preceding years, a superficial examination was made of many sites both above and below the falls of the Rappahannock. The results were interesting and satisfactory, and have led to the belief that an intensive investigation, including the excavation of certain areas, would prove of exceptional value and shed light on the manners and ways of life, and possibly reveal the identity, of the early inhabitants of the valley. SMITHSONIAN MISCELLANEOUS COLLECTIONS VOLUME 96. NUMBER 5 THE MALE GENITALIA OF ORTHOPTEROID INSECTS BY R. E. SNODGRASS Bureau of Entomology and Plant Quarantine U. S. Department of Agriculture (PUBLICATION 3442) CITY OF WASHINGTON PUBLISHED BY THE SMITHSONIAN INSTITUTION SEPTEMBER 25, 1937 The Lord Baltimore Press BALTIMORE, MD., U. & A. Py Transfer Coast & Geodetic paler YAN 24 41939 Sark ists ’ at Ae ae ae MALE. GENITALIA OF ORTHOPTEROID INSECTS By R. E. SNODGRASS Bureau of Entomology and Plant Quarantine, U. S. Department of Agriculture CONTENTS Whew WTREGOTICEITY «8s 1d 5 he eon, ea Poe otha ee Sot Oe ne STs ee ete ere I Development of thesmaleszonadseess. wisest oe aoe aeeine aaa. 2 mesma eh: eeriita lid WCtSp ges tae eect iecs teers eee a Gas Ge hee. 6 Whevexternalls Senitaliaen. yn ivr eretreicie sl meieittaeie cient cee oe 9 10, ESO See AO Rein aot re icc Gi cos ht Sener Rn Rare MO ETS ore, con Sn Me Eive, I2 UNM FSET TESECHIS LEICA tee occ els 6 aay ie oe Meee maar e as Fania s, Sa A Oe 14 UW, ZZAOTEEND LUG Ce gn caer ah ioc IR sae aad an eee Ne inte Gee Le 17 WieeeGrtay LOD att@icl Gal) «<7 clac crt eeee Mayda sie eae Neth etre eA Acie 19 VAIMMEEIASIMALOIGEA: Ai. cukaxen.« rites me are Seles ras he es nee 23 BIBER TESTES: secrets ds Sicha eevee Oe Re Ee OAS 23 sihesaccessony, enitalyelands'. cri sete tices cee ee eee 25 The external genitalia and associated structures................. 25 WALL, WIG aNONG Ker agit Seis mera anich yeehn Sec ieemniene cing ia re Re aH Mie ceria ba meniG Glee 20 Mihesiaternalycenitaliay sen cwecte sates ete cana eh erate eee 30 iickexternalacem#tall ie sees, 25r) telat eee as Baca rs THe alah eevee eee 31 Matingenhabitsiands SpetinatOpaOLesar meena ancient eieais 35 WHIT, IBIBO (aCe Rear oo ae Get mn eters ion sec eREeS oto sd soa obore 36 EINER TCSLEST aris ie sree orske ers or ersvoke olay Pore eee SIC tarsi seo oiieee Caer nores 37 there enitalmducts ao aner c otc oie cae ce ee Cee See 37 iiliesaccessonymcenitall glandseenay- 0) sce eo eee eee 37 Hehe sSemlinial WVeSIGlESe io fat cc aselss pelt ect ae ee eae 4o se gpliallic. olan spay reso ctece cers & nee ceieuel porici ist stented I ere cent: 41 ihespuallic Oreansaseooectias aoe ce eek cate sth ee ee 41 Miepremale =cenitaltai wind wcat bcc sromjern eats a eee rer Ge ce ees 2 SGnMAaON ‘And AISETMMTATION. oc) srt ey 2 kaye we ae here secs eres 55 NONCeM kt TOMTOIM Eales ;.c6 oe sit crate cichotaincte a necro vate a aicis Me ee ee A eee 58 Gy phodertis' monsthosama stoi Seinen. aes oi ee ee Te 61 ANGyamexoynueta bes mer ei aeanarreey sinc em oat oil COREE CRT eere CaS MET OR or ams Wie Rapido paOniniae's stacks wee tera atans Meenas cE re Coens ce 74 PIPEMOPCIMALIS, -THSEIS .saiu 5 ) 85) O30) 73) 9 i 84 See 211) 335) W545 (1387. MER. L544 4 OX One 70 4358 26 40 66 SI 16 67 66 II 77 79 Te 25 38 63 52 16 68 66 II 7 25g BOs.) SOUP RAZ Se 000m G0 8 68 |) 73> 25. 337° 63). (50. “35> G5 (65 (10. 74 5461 20 20 152). 30) AES) SAN 03 8 7t.) (7a 14 220 0352 0.).33) Ss AG 157, 7. 04) 70 2O'NH | h28) fae ho gS NGSs ASE et 4AS 7 $5) 175) aes 20; 30°) SO) 943) 7) OOM) S5y) LOy) -fOSin eo 19 30 49 38. 15) 53. 56 8 64 5780 17 49 66 45 36 81 61 10 71 79 78 TO AZ Sa 00) AZ area Oe oO 8 58. (80. aaa 20 46 66 45 28 ae 67 10 77 78 78 19° 47) (G6) 46° (32 °°77. 150 9 69 The average total seedling lengths for the three series are 58, 62, and 69 mm, corresponding to the exposed, “ diffused” control, and totally dark control, respectively. Here again the effect of light is noticed. The total length is greatest in complete darkness, less in the presence of minute scattered light, and least in the illuminated series. Although the total length is not greatly changed by light, as was true in the first series of experiments, yet there is a distinct indication that total growth is less and that the activity of growth substance or regulators is slightly decreased even by very weak light. No. 6 GROWTH OF AVENA COLEOPTILE—JOHNSTON 13 The marked differences in relative lengths of first internode and coleoptile in the dark and exposed series indicate some drastic re- distribution of growth regulators. The ratios of first internode and of coleoptile to total length of seedling are shown as percentages in table 4. These percentages were determined for each seedling and then averaged for each experiment and again averaged for each wave length. This last average is shown TABLE 4.—Percentage Length of First Internode and Coleoptile of Total Length of Avena Seedling for the Designated Wave Lengths and Temperatures [Average Data in Boldface Type. ] In same moisture chamber Wave — A Separate chamber length Exposed Dark control Dark control Temperature (° I.) ——— — ss Sue A Int. Col. Int. Col. Int. Col. Max. Min. Average 4047 36 64 68 32 oe oe 75 74 74.5 a 69 69 31 Le bic 79 76 FTAs 51 49 5 ae 88 12 79 76 7726 42 58 76 24 87 13 7 77 77-5 40 60 Fis 29 87 13 sha 7 76.8 4358 38 OI 70 24 86 14 70 77 78.0 40 60 7 24 86 I4 78 76 77.0 43 57 78 22 88 12 78 77 775 41 59 77 23 87 13 .- .. 77-5 5461 38 62 73 27 88 12 77 70 76.5 39 61 73 27 89 II 76 76 ~- 76.0 41 59 74 26 87 13 75 75 75.0 34 66 72 28 85 15 79 78 78.5 38 62 73 27 87 13 Sc wa 76.5 5780 25 75 56 44 86 14 79 78 78.5 29 71 60 40 87 13 80 79 79-5 30 70 61 39 87 13 78 78 78.0 28 72 59 41 87 TZ oa aC 7 hsiey | in boldface type. Temperature data are given in the last three columns. But little difference is to be seen in the percentages of lengths of.first internodes and coleoptiles in the first three wave lengths. The values for 5780 A are, however, different. A comparison of these per- centage values in the two series of experiments for the exposed coleoptiles is shown below. The wave-length values between the two series are not strictly comparable but are approximately correct. Each set of values shows about equal lengths of coleoptile for the first three wave-length values. The value for the 5780 A region of © 14 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 each series is higher. All values in the second series of experiments are higher. If it is assumed that the small temperature differences between the various experiments exert but little influence, then it must be concluded that the increased intensity in the second series of ex- periments has increased the percentage length of the coleoptile. This TABLE 5.—Comparison of Percentage Length of Coleoptiles in the Two Series of Experiments for Corresponding Wave Lengths Wave length First series Second series Percent Percent 4047 49 60 4358 48 59 5461 51 62 5780 57 72 is in agreement with the observation that the coleoptiles were longest in the exposed series (greatest intensity), shortest in the completely darkened series (zero intensity), and intermediate in length in the dark control exposed to very diffused light. The reverse is true regarding the first internodal values, which are the coleoptile values subtracted from 100. DISCUSSION In determining the sensitivity curve (Johnston, 1934) of Avena coleoptile, only the tips were illuminated. In the experiments here reported entire seedlings were grown in restricted wave-length regions of the spectrum. Went (1926) has pointed out two distinct light responses in the coleoptile, one the tip response, the other the base response. Where the entire coleoptile is illuminated, both responses are concerned. Van Overbeek (1936b) calls attention to a third type of response described by Tollenaar and by Van Dillewijn, which is designated as the “ dark growth response.” It is the increased re- sponse to growth substance in darkness and has nothing to do with bringing the plants back into darkness. Plants continuously exposed show the response. In the present experiments the time period was exactly four days, and in that length of time the first internode as well as the coleoptile attained considerable length. The light response of both organs must be considered as well as their interactions. Also the seed and roots in the exposed series received a certain amount of radiation. The problem becomes further complicated when the action of plant growth substances with their production, inactivation, and transport is considered. No. 6 GROWTH OF AVENA COLEOPTILE—JOHNSTON 15 Du Buy and Nuernbergk (1934) report an experiment in which monochromatic light of higher intensity (170, later 80 ergs/cm’*/sec. ) was used. The mercury are and incandescent lamp with appropriate glass and liquid filters were used. After six days the following lengths of first internodes and coleoptiles are given: TaBsLe 6.—Length of First Internode and Coleoptile in Monochromatic Light (After Du Buy and Nuernbergk) Wave length First internode A (mesocoty]) Coleoptile Total length 3063 13.5 39.5 53.0 4358 18.3 44.3 62.6 5461 19.7 51.2 70.9 6000 19.6 46.4 66.0 Dark control 22.5 50.3 72.8 At the end of four days the average length of mesocotyl under each of the five different treatments was given as 1.5. Under the four conditions of illumination, the coleoptile varied in length from 22 in the ultraviolet to 26.6 in the red. The dark value given for the coleoptile was 12.5, approximately half that of the exposed values. At the end of the third day no growth of the mesocotyl was detected. By the sixth day in these experiments the primary leaf had broken through in go to 100 percent of the seedlings. In the Smithsonian experiments the primary leaf was usually at or near the top of the coleoptile but had never broken through in the plants of either the exposed or the dark series. From the above table it will be noted that the first internode or mesocotyl of the dark control was longer than those of the illuminated seedlings. The coleoptile was also longer with one exception than those of the exposed seedlings. The reason for this disagreement with our experiments is not clear unless it may be due to a difference in oat varieties used. In the Smithsonian experiments the coleoptiles of the dark controls were always shorter than those of the exposed plants. This was also true on the fourth day in the experiments of Du Buy and Nuernbergk. At the end of six days, with the exception of the group exposed to 3663 A, there is but little difference in total length of seedlings. Here also the total length of their dark control is slightly greater than those of the exposed series. , Du Buy and Nuernbergk (1935) give two explanations as to why growth of the mesocotyl is depressed. First, according to their theory, protoplasmic streaming is correlated with the activity of growth regu- lators. Temperatures above 25° C. and illumination retard proto- 16 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 plasmic streaming. The cells of the coleoptile will, therefore, retain more growth regulators than are transported to the mesocotyl cells at high temperatures and when illuminated. Hence, under such con- ditions the coleoptile grows faster than the first internode or meso- cotyl. The length of path that the growth substance must travel to reach the cells of the mesocotyl is thus increased and less growth in this latter tissue will result. Their data (1934) and those of Hamada (1931) show, as one would expect from this theory, the mesocotyl cells of illuminated plants to be smaller than those of dark- grown plants. A second explanation is that mentioned by Van Over- beek (1935), who states that a temporary increase in temperature increases the use of auxin, thereby decreasing the amount available to the mesocotyl. According to DuBuy and Nuernbergk (1935), it is very probable that a decreased protoplasmic streaming and an increased use of auxin together cause an increased growth of the ’ coleoptile and a decreased growth of the mesocotyl. The reaction capacity of cells to growth substances may also enter into the prob- lem since such reaction capacities appear to be different for light and for dark exposed plants, as shown by Du Buy (1933), and for plants erown under different humidity conditions, as shown by Gorter and Punike (1037) The results of several experimenters show that the growth of the first internode may be depressed by light, high temperatures, and decapitation of the coleoptile tip. Van Overbeek (1936a) depressed mesocotyl growth of maize seedlings by exposure to a temperature of 48° C. for half an hour. The amount of growth substance taken from the tips of similarly treated seedlings was less than that from the controls. That there is a linkage between the amount of growth substance obtainable from the tip of the coleoptile and the growth of the first internode seems evident. It is reasonable to suppose than any condition that would inactivate the growth substance in the tip of the coleoptile or that would interfere with its transport to the first internode would depress the growth of the first internode. From the data here presented, it would appear that radiant energy, even as low in intensity as 1.2 ergs/cm’/sec., has a very definite influence. Whether this influence is exerted through a direct effect on the growth sub- stance itself or on its mode of transportation through protoplasmic streaming or other means cannot be stated. There also appears to be a wave-length effect. But whether this is a direct effect on the first internode or a secondary one through its action on the coleoptile, which in turn would control the growth substance reaching the first inter- node, cannot be concluded from these data. Under one set of con- NO. 6 GROWTH OF AVENA COLEOPTILE—JOHNSTON 7 ditions (darkness) the first internode is long and the coleoptile short. Under other conditions (illumination) the first internode is short and the coleoptile long. It would appear that in the first case the growth substances were not utilized for growth in the coleoptile, but transported to the first internode where the greater amount of growth was found. In the second case considerable more growth took place in the coleoptile. Here it would appear that some growth substance was prevented from acting on the cells of the first internode. In these experiments light was evidently the controlling factor in the distribution of the growth substance. Reference to figure 4 shows that wave lengths 4o50 A to 5500 A slightly depressed the growth of the coleoptile. In this region growth of the first internode was somewhat increased. In the wave-length re- gion 5500 A to 7200 A the reverse effect is noted. Here the coleop- tile growth is increased and that of the first internode decreased. Since there is very little difference in total growth of the entire seedling between light and dark conditions, one is led to the conclusion that light of this low intensity functions mainly as a distributing rather than an inactivating factor on growth substance. However, since there was actually a slight difference in total growth, there may also be a slight inactivating influence of light, for the experimental data show slightly greater average growth for the dark control plants than for the illuminated ones. SUMMARY Two series of experiments were conducted in which the growth of the coleoptile and first internode (mesocotyl) of Avena sativa, Mark- ton variety, was studied in different wave-length regions of the visible spectrum. Radiation of low intensity greatly depressed the growth of the first internode as compared to that in darkness. On the other hand, growth of the coleoptile was depressed in darkness. Total growth of the entire seedling (coleoptile plus first internode) for the 4-day periods of all these experiments was slightly greater in darkness than in light. Although the total length of seedling was not greatly dependent on intensity of illumination, the ratios of first internode and coleoptile to total length were extremely critical indices to in- tensity of illumination. It is extremely difficult to get conditions dark enough to avoid light effects on these ratios. All wave lengths give much the same tendency to decrease the ratio of first internode to total length and this effect increases with light intensity. Yet there is a wave-length effect yielding a minimum ratio at about 6500 A. 18 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 96 It would appear that growth of the coleoptile retards the growth of the first internode and that the growth of the one tissue takes place at the expense of the other. Light probably acts more as a redistributing agent of the growth substances than an inactivating agent. However, even at the low intensities here employed, there was a slight indication that some inactivation occurred, since the average total length of the illuminated seedlings was slightly less than that of the dark controls. LITERATURE ClhiEDp AVERY, GEorRGE S., JR. 1930. Comparative anatomy and morphology of embryos and seedlings of maize, oats, and wheat. Bot. Gaz., vol. 9, no. 1, pp. 1-39. BoysEN-JENSEN, P. 1936. Growth hormones in plants. English translation and revision by G. S. Avery, Jr., and P. R. Burkholder. 268 pp. New York. Du Buy, H. G. 1933. Uber Wachstum und Phototropismus von Avena sativa. Rec. Trav. Bot. Néerlandais, vol. 30, pp. 798-925. Du Buy, H. G., and Nurernserck, E. 1934. Phototropismus und Wachstum der Pflanzen. Ergebn. Biol., vol. 10, Pp. 207-322. 1935. Phototropismus und Wachstum der Pflanzen. Ergebn. Biol., vol. 12, PP. 325-543. Fiint, Lewis H., and McAtister, E. D. 1935. Wave lengths of radiation in the visible spectrum inhibiting the germi- nation of light-sensitive lettuce seed. Smithsonian Misc. Coll., vol. 94, no. 5, pp. I-IT. 1937. Wave lengths of radiation in the visible spectrum promoting the germination of light-sensitive lettuce seed. Smithsonian Misc. Coll., vol. 96, no. 2, pp. 1-8. Gorter, C. J., and FuNKE, G. L. 1937. Wachstum und Wuchsstoffproduktion bei Keimpflanzen von Raphanus sativus in trockener und fluchten Luft. Planta; Archiv. wiss. Bot., vol. 26, no. 4, pp. 532-545. Hamapa, H. 1931. Uber die Beeinflussung des Wachstums des Mesokotyls und der Koleoptile von Avena-Keimlingen durch das Licht. Mem. Coll. Sci. Kyoto Imp. Univ., ser. B, vol. 6, pp. 161-238. Jounston, Ear S. 1934. Phototropic sensitivity in relation to wave length. Smithsonian Misc. Coll, vol. 92; no! 11, pp. 1-17: Norem, W. LuTHER. 1930. Mineral nutrition and seasonal growth of ageratum in sand cultures with auto-irrigation. Amer. Journ. Bot., vol. 23, pp. 545-555. NO. 6 GROWTH OF AVENA COLEOPTILE—-JOHNSTON ide) VAN OVERBEEK, J. 1935. The growth hormone and the dwarf type of growth in corn. Proc. Nat. Acad. Sci., vol. 21, pp. 292-299. 1936a. Growth hormone and mesocotyl growth. Rec. Trav. Bot. Néer- landais, vol. 33, pp. 333-340. 19360 b. Growth substance curvatures of Avena in light and dark. Journ. Gen. Physiol., vol. 20, no. 2, pp. 283-300. WENT, F. W. 1926. Concerning the difference in sensibility of the tip and base of Avena to light. Proc. Kon. Akad. Wetensch. Amsterdam, vol. 20, pp. 185-101. jit } ; Let 7 nd —* in Beroy oiiga Pe y al q a es he yyy LYE Yyyyy yt oo wyyyy yoy: he U | »> eB: s >) DD» = 5 > >» y DY >» > WS he » VO bubba hehe Dy et hhed bee! Via bee Wali hy eS D »> ) ») ») ») » ») ») ») )) ») ») »») ) )) ») ») ») ») » J NS WS Sie Juyey yey GRAY ISS Nad wiv wee hy OS > Mi ]\e Vv JSST TATA Berce JS ae >> w >) o> SIDE EP SD ae 2 I» > p> >») DD» D DY » > >») »»> » > >») » >> >» »> DD» Ye. >») ee D »» 33 333 > >>». + $5 222 ») »» > » ee >. > ahs 33 » > > yy yy )) >». D> >> > ee D»D >») > > 2 2 >> =>. > ee 5 > > Py 2 > »> DS » ) »» yd D py» » yy »y : 2! 5 : »>»d ) _)»» Ps , Dy»yp) D >Dyyy)> >>> » »» » >» » ¥ > Dy) Ip >> w» >; ») >) yp » » 5 Dy) > BW > P32 > > »> Dw» > > DD») 92> > D Dy) » > » Dp - D>») >> »> p> D>: ] y) yy> e yy»). »> > D> 2») yy» J @ yy») > D> PID) 2» ) D>» > > » ) yp) > > ') ) yz > : >> > 2 >> > ») » J » > » ))) > D by) » 33% ; “ » >» > PD Dy) yp) Y | i) \ Ne PAD AHOWS ll 3 9088 01421 OIC