The Kabiri or Girara District, , Papua. Author(s): A. C. Haddon Source: The Journal of the Royal Anthropological Institute of Great Britain and Ireland, Vol. 46 (Jul. - Dec., 1916), pp. 334-352 Published by: Royal Anthropological Institute of Great Britain and Ireland Stable URL: http://www.jstor.org/stable/2843397 . Accessed: 08/12/2014 22:56

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp

. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

.

Royal Anthropological Institute of Great Britain and Ireland is collaborating with JSTOR to digitize, preserve and extend access to The Journal of the Royal Anthropological Institute of Great Britain and Ireland.

http://www.jstor.org

This content downloaded from 128.235.251.160 on Mon, 8 Dec 2014 22:56:06 PM All use subject to JSTOR Terms and Conditions 334

THE KABIRI OR GIRARA DISTRICT, FLY RIVER, PAPUA.

[WITH PLATES XIX-XX.]

By A. C. HADDON.

THE Kabiri or Girara District is the low-lyingcountry between the Fly River and the Aramia affluentof the Bamu, and extendsfor a short distance beyond the Aramia. Roughly speaking,it extends from about east longitude142' 30' to 1430 15'. It is floodedduring nine months of the year,but thereare innumer- able low hillocksand ridges,some of considerableextent, on whichcoconut-palms are planted whereverthere is roomfor them. The villages are on hillocks,between which communicationis made by means of very long,narrow canloes of shallow draught. There are very large sago swamps; indeed, accordingto Mr. Beaver,' there is a plethora of sago and coconuts; he adds that yams,bananas and sweet potatoesare cultivated,but not to any great extent. Fish abound in the swamps, and are caughtin large conical traps made of cane, somethingafter the style of a lobsterpot [these may be at least 152 cm. (5 feet)in length]. Bird life is abundant; birds of paradise, goura pigeonlsand cassowariesare plentiful. The swampscontain many varietiesof waterfowl. Mr. Lyons2states that the Kabiri (as he terms them) live in some twenty scatteredvillages from Taitiarato in the east,on theFly littoral,to DomoriIsland and to just northof the Aramia River. He describesthe people as distinctlyof Papuan type and of finephysique. The men as a rule go nude, as do some of the women, whilstothers wear a narrowperineal band. Mr. Beaver says the majorityof the men wear a conical fibrehat, decoratedwith feathers, or else a skull cap ofnetwork (Fig. 2). The hair is shaved above the forehead,and a small corkscrewgoatee beard is worn. [Mr. Beaver illustratesthese factsby a photograph.] Mr. Lyons saysall the males adopt the diba, or conical hat, whichis affixedto the hair of the head by some glutinous mixture; the hat is not removed even when sleeping. Generally the outsideof the hat is coveredwith whitelime, and the top is made a receptacle forfeathers. Mr. Beaver says the womencover the head and bosomwith a veil of net. " This is stated to be a mourningdress, but so many of themwear the veil that I can hardly believe it is entirelymourning.3 Their other dress is rather scanty,and consistsof a wisp of grass drawnin tightlybetween the legs. . . . The Girara (as he terms them) emphaticallydeny being cannibals, but admit head- hunting.... They have fivetotems, which descend fromfather to son. These are the pig, the pigeon,the alligator[crocodile], the black snakeand the cassowary." They claim descentfrom a dog (Beaver, in a newspaperinterview).

1 Ann. Rep. Paputa,1911-12, p. 11. 2 Ann.Rep. Papua, 1913-14,p. 99. 3 Landtmanhas no doubt that thesecaps are solelyfor mourning.

This content downloaded from 128.235.251.160 on Mon, 8 Dec 2014 22:56:06 PM All use subject to JSTOR Terms and Conditions A. C. HADDON.-The Kabiri or Girara District,Fly -River,Pappua. 335

According to Mr. Lyons, the villages usually contain only one house, which is of immense size, mieasuringanything from 18-29 to 39-62 m. (60 to 130 feet) in length and from50 to 60 feet in breadth. [In a newspaperaccount Mr. Beaver is reportedto have said that the houses were sometiinesas much as 121P92to 152 4 m. (400 to 500 feet) long and 1829 to 24-36 m. (60 to 80 feet) in width.] These houses,which are built on high piles, are well constructedand last formany years. The roofsare high,arched, thatched with sago leaf,and extend to the groundalong each side. Compartmentsare made at the sides of the housesfor the accommodationof the womenand children,who are not permitted,except on certain occasions,to enter the men's portion of the house (genemna).The women enter their apartments from underneath the house. [Mr. Beaver says: "The centreof this huge buildingis a kind of commonhail, which is used only by the

:1

FIG. 1.-Chief post (1,3 m.; 4 feet3 inches)and chiefjoist (2 06 m.; 6 feet9 inches) of a Kabiri housein the Port MoresbyMuseum. Collectedby A. P. Lyons at Adiba. From a sketchby Mr. Lyonsand a photographtaken at Port Moresby. men,while the walls of the structureare divided into cubicles in three or four floors,access to whichis gained by means of ladders. The womenare not allowed to enter the building by the same door as the men."-Newspaper report.] Even the single men are restrictedto theirown end of the genema. Marriedmen alone have access to all parts of the house,and, in fact,to everythingconnected with the ceremonialrites of the tribe. The marriedmen's end of the house is called mana. The followinginformation is taken fromMr. Lyons' report,supplemented by his manuscriptnotes. When a suitable site has been selected fora new communial house, the village carveris deputedto preparethe chiefpost (tinmi),and the " chief bearer,"or joist (idadari). The formeris either a carving or a painting of the principalchief. The ends of the chief joist are carved to representthe heads of crocodilesswallowing either birds or numanbeings; usually the head or legs of a rnanare representedas stickingout of thecrocodile's mouth (Fig. 1). Whilst these

This content downloaded from 128.235.251.160 on Mon, 8 Dec 2014 22:56:06 PM All use subject to JSTOR Terms and Conditions 336 A. C. HADDON.-The Kabiri or Girara -District,Fly River, Papua. are being made, the women and childrenprepare food fora feast. When all is ready,the people assemble; the chief (wasina) places the chief post in its hole, otherpiles are placed in alignmentwith it,and the chiefjoist is placed on themall; these will formthe middle supportsof the house. The wasina names the village, and thereis much feastingand dancing. When the houseis completed,which often takes many months,another big feastand dance are held. In a letterdated October,1910, and privatelyprinted, the Rev. B. T. Butcher says all the villages are built on the same principle-a great house (that at Mida, or Kubu, is about 200 feet long and 40 feetwide) with a central hall, which runs the entire lengtlhof the house and is reservedfor the use of the men; on either side are the walled-offprivate apartments. These consistof threestoreys. In the lowest of all the cookingis done,the middle one is forthe womenand children,and the top storeyfor the men. The women do not come into the central hall, but reach the cooking and sleeping apartmentsfrom underneath the building. The men reach the various storeysby means of notchedlogs whichrest against the side walls of the hall. The only light in the hall filtersthrough various hlolesand chinksin the walls and the small doorsat eitherend. Thereare no windows. The houses are very high and with an average width of 50 or 60 feet; they are raised 5 or 6 feetfrom the groundon piles. The flooris made of strips of palm trunks laid side by side and formingan undulatingsurface. Illumination is made by means of little lightedsticks (wiki), whichburn slowlyand need constantsnuffing, which is done by knockingoff the charredportion. The house at Barimu is about 450 feetlong, about 50 feetwide, and 30 to 40 feet fromthe ground at the gable. Here Butchersaw the men seated in a circle round their chief,who was beating time with a lightedwiki. Each man had his betel nut and lime gourdby his side, and a lighted wi7ciin his hand with which he beat time in unison with the chief, while boys were chewing "gummada" (kava) for the men's use. One very old man repeatedthe firstword or two of every line of a song before it was sung, then all joined in a weird,musical, deep-toned chant, which ended witha most quaint cry of " Wa ! Hi ! " twice repeated. Butchersays that along both sides of the house at Barimu a straightrow of finecrotons was planted,then came a broad road, 15 or 20 feet wide,and beyond this a line of nicely arrangedand beautifullykept gardens. Lyons furtherstates that a clear space of about 60 feetis leftall roind the genernafor dancing. The gardensare made beyondthe dancing ground; these are long,narrow, rectangular beds of heaped-up earth,finely broken up; paths 10 or 11 feet wide are left betweenthe beds. Coconuts,taro, and a yam called wisa, are planted. An old village usually presentsa prettysight, being surroundedby tall coconutpalms and the clean red clay of the dancingground. The approachesto the villages are also borderedby coconutpalms. The roads between the villages are wide and well kept. Here and therealong the sides of the tracks are to be seen trees marked with strangehieroglyphs, which denote the boundaries,either of the village or of an individual villager'sland.

This content downloaded from 128.235.251.160 on Mon, 8 Dec 2014 22:56:06 PM All use subject to JSTOR Terms and Conditions A. C. HADDON.-The Kabiri or Girara District,Fly River, Papria. 337

The womrenare occupied in making their huge fishingnets (?traps), in fishing,and preparingsago. The men ctultivatethe gardens, are fine handi- craftsmenand excel in carving(Lyons). Though a bush tribe, they have plenty of use for canoes during the wet season, when the hollowsare filledwith water; the waterwaysthus formedextend formany miles in all directions. Accordingto Mr. Lyons,' the principalceremony of these people is the moi- iata, five of which are held during the wet season. In connectionwith the ceremonythree large carved wooden crocodiles,known as Posia, Getehidi and Moi-ira,together with a number of minor articles,are used. When not in use all these objects are kept in a house,called the kakanipa,which only certain married men may approach. If anyoneelse did so he would be eaten by a crocodile,which is sacred to these people. When it has been decided to hold a moi-iata,a small boy suddenlydisappears fromthe village. It is given out that he has been taken by a crocodilewho was angrywith him for approachingthe kakanipa. There is much wailing,anid the bereaved parentsapproach the samui,or chief sorcerer,and ask him to implorethe crocodile to restore their lost boy, who promises to do what he can. All the unmarriedmen, women and childrenare sent away to make sago and collect coco- nuts fora big feast to propitiatethe crocodile. Whilst theyare away,the Posia, Getehidiand Moi-iraare removedto the mana, or marriedmen's end of thegenema. To the head of each of themis affixeda huge cane effigyof a crocodile'shead. All of themare then plasteredwith clay. The boy,who was supposed to have been taken away by a crocodile,but who actually had been blindfoldedand taken away by nightinto the bush, wherehe was hidden by the samin,is then placed insidethe cane mouth of the Posia. A fenceis erected round the effigiesto keep offtoo- ciurious persons. When all is ready, the sanbu and his assistants donigrotesque masks and beat loudly on drunis until all the villagersreturn. They are invited into the genema,when, with manyincantations, the lost boy is dragged fromthe mouthof the Posia. Of course,paynment is requiredfor the restorationof the boy, which the parentsare glad to make. Then followdancing and feastingin honour of the mightycrocodile, and marriagesare celebrated. The followingaccount of the materialculture of theKabiri is based mainlyon the- interestingcollection which Dr. G. Landtman generouslypresented to the Ethnological Museum at Cambridge. Most of these he obtained in 1910 when travellingwith the Rev. B. T. Butcher,L.M.S.; he also gave some specimensto the British Museum. The formerwill be referredto by an " L," and the number which followsis that in the CambridgeCollection. There are two main varieties of caps, the rigid and the netted conical cap (diba), Kubu (Fig. 2 A, B). The foundationof these caps is a strip of ratan,which is producedinto a continuousspiral, each whorlbeing lashed to that below and above

1 Loc. cit.,p. 100.

This content downloaded from 128.235.251.160 on Mon, 8 Dec 2014 22:56:06 PM All use subject to JSTOR Terms and Conditions 338 A. C. HADDON.-The Kabiri or Girara District,Fly River,Pap'aa. by a band whichis twistedtransversely across itselfbetween each whorl. The apex is hollow,and is producedinto a broad riing,the outer edge of which is protected by a plait of split ratan. The whole of the outsidewas originallycovered with black beeswax (?), and portions of the wearer's hair still adhere to the lowermost whorl. A smaller specimen (L 38) is coveredwith red clay; height, 18-5 cm. (71 inches); diameter,1141 by 12 9 cn. (4-- by 5 inches). A similar cap, gngu, was collected by J. Chalmnerssome twelve years ago, and is now in the British Museum,the ratan base of whichis servedwith bark fibre. Insertedin the apex are five mounted cockatoo feathers,the stems of which are threaded througha number of dark red lenticularseeds (Partingtonsays Pandanus seeds), and there are a couple of white featherscut into a series of V's. Height about 23 cm. (9 inches).' Netted cap for mourning,atima, Gaima (Fig. 2 c, D), of two-plystring made of the ininerbark of some plant. The cap is not finishedoff, and as thereis a small supply of untwistedstrips of the bark in the inside of the apex, it is probablethat the cap would have been somewhatlonger. The stitchis of a simple character,ancd is shownin Fig,.2 D. There are in the BritishMuseum two nettedcaps, " atima,, wornby bush tribesof Obere when fightingand dancing,"one of whichis said to be worninside the other; they are workedin horizontalbands of differentwidths, which are painted black, white,yellow, and red. The outside one has a tuft of cockatooor P. raggiana feathers,and is about 38 cm. (15 inches) in length; the other is about 31 cm. (12j inches).2 Partingtonalso figuresa " networkcap (atuna ata) fromObere, worn when in mourningfor parents or wite."3 These were also collectedby Chalmers. Three plumes forheaddresses were collectedat Gaima by Landtman. Fig. 3 A is a thin piece of light whitewood, 187 mm.long, cut in the shape of a feather, painted black and red at the base, mounitedon a quill, half of which has been cut away longitudinallyfor more than half its length,the whole end being stuck on a. wooden splinter; B, cockatoo feathersstuck in a disc of blackened pith; on dne side a red and green parrot feather,and on the other two notched similar feathers,mounted on a quill of a cassowary,part of which has been shaved away; total length, 35 cm.; greatest breadth, 17 cm.; c, two cockatoo feathera insertedin an elongated lump of black beeswax (?), in whichare inserted several Abrus seeds,and mountedon a strip of ratan,the upper part of whichis consider- ably thinneddown. Three water buckets(L 374-6), maripa,Gaima, of palm leaf (? Ptychosperma) folded in the usual nianner,with a stick passing longitudinallythrough the folded eilds and a loop handle of ratan. Water vessel made of one internodeof bamboo(Fig. 4); simpleincised design of three concentricincomplete circles within two ovals. Stopper of doubled-up palm leaf. Gaima.

1 Cf. AlbumII, Plate 192,No. 2. 2 Cf. AlbumII, Plate 191,No. 3. 3 Loc. cit.,No. 2.

This content downloaded from 128.235.251.160 on Mon, 8 Dec 2014 22:56:06 PM All use subject to JSTOR Terms and Conditions A. 0. HADDON.-The Kabiri or Girara District,Fly -River,Papua 339

FIG. 2. A, conical cap, Kubu (L 37): ht. 24-5cm. (91 inches),diam. 13,6x 1,47 cm. (5f x 53 inches); B, apex seen fromabove; the cap was cut offfrom the hair with great difficulty. c, mourningcap, Gaima (L 1): ht. about 36 cm.(144 inches),circ. at base about 52 cm. (20 inches); D, detail of the netting.

A~~~

FIG. 3. FIG. 4. Plumes forheaddresses, Gaima: Bamboo watervessel, Gaimna A (L 61), B (L 62), c (L 63). (L 227): 1. 44 cm.,diam. 98 mm.

This content downloaded from 128.235.251.160 on Mon, 8 Dec 2014 22:56:06 PM All use subject to JSTOR Terms and Conditions 340 A. C. HADDON.-The Kabiri or Girara District,Fly River, Papqta.

The canoes are verylong, narrow, of shallow draught, and withoutan outrigger> theyare carved and paintedin an elaboratemanner (cf. photo by Beaver).' Butcher refersto one about 21 m. (69 feet) long and 61 cm. (2 feet) broad," carved fromend to end and quite elaboratelypainted."2 Lyons says the canoes are dug-outs; one measured21P33 m. (70 feet) long by 1P14 m. (45 inches) in beam; the prows arem carved to representcrocodiles, and along its sides " they usually carve figuresof fishes,birds, or crocodiles. The paddles are exceptionallywell made. They have a spoon-shapedblade and long handles withmen's heads carved at the ends. The^ blades are markedin pigmentsof variegatedhues. Each familyhas its own mark so thatthe ownershipof a paddle is easily ascertainable."3Our paddle (Fig. 5 B) iS. of light-colouredwood. The elongatespatulate blade is 83 cm. long and gradually- widens from45 mm.above to 154 mmr.at a distance of 106 mm. fromthe end; one edge is veryslightly curved. The blade is slightlyconcavo-convex, and the stem, of the handle is prolongedon each side as a prominientkeel to the end of the blade.. The handle is approximatelyround, with a cylindricalhead whichis oval in section,. the narrowend being on the same side as the broadpart of the blade. The conlvex surface of the blade is painted black, white, and red; below, a human face is represented,and above are what appear to be the jaws of an animal. The under surfaceof the blade and the handle are coloured chocolate red, the head of the handle is white. I obtainedat Wododo,Dibiri Island, at the mouthof the Bamu. threeuncoloured paddles like those describedby Mr. Lyons. The spear,kibiri (Fig. 5 c), is of heavypalm wood,roughly made, blackened,, and flatoval in section, The barbedportion begins with a bead, and thereare about sixteen rows of small barbs which appear to be arrangedin an irregularspiral. It graduallyincreases in diameterfrom the point to the butt. The two wooden war clubs are of very differenttypes (Fig. 5 A,D): A is of heavy wood,length 1-54 in.(60 inches),greatest diameter of head 50 mm.,average- diameterof handle 22-25 mm.,approximately round in section,a flattenedbi-convex button'on the top. Upper portionwith eight grooves,with sharp edges between them,painted with simple designs in black, white and chocolate red,ungrooved- portionchocolate red; the groovedabout 63 cm. long. D, gabir-apira, is of dark heavy wood, length 121 m. (4713 inches); head 122 mm.long, 61 mm. diameter,. round in section,with fivewhorls of six teeth; a conical swelling at the handle- end withconvex base. Sling loop of red calico. The undecorateddagger from Gaima of casso wary leg bone is over 37 cm.in length. The two bows,gagi, Nos. 773, 774, are of black palm wood. L 773 is 1-89 m.. (74kinches) long,plano-convex in section,the convexitybeing produced into a slight keel near each end; betweenthis and the spike forthe stringis a slightswelling or imperfectbead. The spikes of 773 have a collar of plaited ratan. L 774 is 1-93m, (76 inches) long,and the flatside is slightlyconcave, there is a lenticularbead at each end below the spike. In both the flat side is at the outside of the curve- 1 A.R., 1911-12. 2 Loc. cit. 3 A.R., 1913-14,p. 100.

This content downloaded from 128.235.251.160 on Mon, 8 Dec 2014 22:56:06 PM All use subject to JSTOR Terms and Conditions A. C. HADDON.-Thte Kabiri or Girar-,aDistrict, Fly -River,Pcapua. 341

tT11 611

FIG. 6.-Details of bow and bowstring. A, B, methodof finishingoff bowstring; forthe sake ofclearness only two fibre strandsare shown; c, loop-knotfor stringing;D, end of (L 773); E, section of samne.

H E

C F

FIG. 5.-A, war club,Barimu = Kakimadase FIG. 7.-Carved woodenarrow-heads. A., B, E, (L 467); B, paddle, Barimu (L. 446): Gairna; C, D, F, G, H, Koabu. A (L 755), 1. 2-06 m. (81 inches); c, spear, Gaima 1.1P756 ni.; B(L 753),1. 1705 m.; c (L 566), (L. 771): l. 1-89m. (74i inches),barbed 1.1P418 m.; D(L 565),1. 1324 m.; E (L 754), portion 73 cm. ; D, war club, Gaima 1. 1-777m.; F(L 564),1.1.471m.; o(L 558), (L. 775). 1. 1-674 m.; E (L 762),1. P427 in.

This content downloaded from 128.235.251.160 on Mon, 8 Dec 2014 22:56:06 PM All use subject to JSTOR Terms and Conditions 342 A. C. HADDON.-The Kabiri or Gircra District,Fly River, Papua.

The ends of each bamboo bowstringare split and notchedon each side (Fig. 6). A strandof bark fibreis passed across the back of the bowstringfrom one notch to anotherof the firstpair, and on the frontside it is knotted,the free ends pass over the knot on the strandabove it, pass throughthe slit in the bowstring,diverge out- wardsand upwardson the undersideof the bowstring,each passing roundthe edge of the bowstringon to its frontaspect, where they converge upwardsto the centre of the bowstring.A freshstrand starts from the secondpair of notchesand pursues a similarcourse, and so on foreach pair of notches. All the strandsare eventually united to forma plait,which ultimately passes into a two-plytwisted string, which in No. 774 becomesgradually thinner and ends in a knot. The plait is formedinto a loop-knotfor stringing the bow (Fig. 6).1 The twenty-fivearrows in our collectionmay be grouped as follows:-A, bambooheads; B, severalbarbed prongs; C, woodenheads. The arrowsshould be comparedwith the descriptionsgiven in Vol. IV of the Torres Straits Reports. (This is referredto as" IV.") A. With bambooheads. The shaftmay have the rindentire (766), or it may be scrapedoff except on the upper internode. There is a wooden aftershaft. At the junctionof the two parts theremay be a plaited cane collar, or, as in 761, a smoothgummed lashing of vegetablefibre; the bamboo is so cut that the point of the head is fashionedout of a node,thus makingit thicker,stronger, and less liable to split. L 761, Waduru-uere,also labelled Ipidarimo-uere(i.e., arrowfrom Waduru or fromIpidarimo, on the oppositebank of the Fly River), Koabii, length1-645 m.; the palmwood aftershaft has two rows ofcarved deep zig-zaas; flattenedbamboo head. L 766, kowe,Gaima, length1-883 m.; aftershaftsquare in section with three rows of fourblunt barbs and two square beads above; none of the barbs is opposite the other, except in a longitudinaldirection; bamboo head concave. L 767, kow6,Gaima, length1-665 m. B. With morethan one point. These emenakatafrom Gaima agree in havingthe shaftscraped, except the upperinternode, which in 770 has longitudinalscrapings; the palm woodprongs are kept apart by a plug and strengthenedby ratan lashing (IV, Fig. 187, A). L 768, length 1-810 m., three prongs with 6-7 unilateral barbs; L 769, length 1-877 m., four prongs with three to five unilateralbarbs; L 770, length1610 m., fourprongs with threeunilateral barbs. C. With woodenheads. A selectionof the moreunusual typeshas been made in Fig. 7. In these all the shaft,except the upper 28-30 cm.,has the rind scraped offand is blackened,but in the upper portionlongitudinal and transversescraped bands occur,sometimes the bands of wholeskin are engravedwith simple patterns. A, B, ani-idi, Gaiina; head square in section,except lowerpart of B; each whorl has fourbarbs. E, batere,Gaima, flattened in section. These three have a collar of plaited cane and slender scraped shafts. c, D, F, haru-haru,Koabu, palm wood head, flattenedin section. c*,hirimca nepate (also labelled Ipidarimo),Koabu. H, I I collecteda bowwith this laced attachmentat Aimaha,Kikori delta, and it occursat Kumukumu,Aird Hill.

This content downloaded from 128.235.251.160 on Mon, 8 Dec 2014 22:56:06 PM All use subject to JSTOR Terms and Conditions A. C. HADDON.-The Kabiri or Girara District,Ply River, Papua. 343 tere-utere,Koabu; the upper part of the head is plano-convex. T-hesefive have a thin vegetablelashing, dyed red. The followingfrom Koaba agree so well with arrows that I obtainedin Mer, TorresStraits, that we may concludethat therewas a regulartrade in arrowsfrom this districtto the islands beforethe arrival of Europeans, L 559, 560, bobok7u, plain woodenhead. L 561, oto-oto,palm wood head,unilateral barbs. L 563, anidi, threewhorls of threebarbs roughlymade, L 567-571, nalu-Jpquru(IV, Figs. 181, 182); 568 is also labelled oto-oto. L 572, Jbaiopm,with a fusiformpoint. Most

FIG. 8.-Detail of the fishingnet, Gaima (L 400).

A

B ~~~~~~~~D

AIL ~~~~~C

FIG. 9. Tobacco pipes, Gaima: A (L 150), 1. 46'5 cm., diam. 5'5 cm.; two holes; snake and an unrecognizablecreature. B (L 149), 1. 35 cm.,diam. 5 cm. ; two holes ; four hindlegs offrogs (?). c (L 148), 1. 38'5 cm.,diam. 6 cm.; one hole; labelled segia; fresh-waterturtle. D, end of c, showingperforated disc of nacreand Abrus seeds. have all the rindof the shaftscraped, except the upper internode,which may be decoratedwith a broad black band. The heads have sharp conical or fusiform points,and some have a zig-zagbeading (IV, Fig. 182, A,B). Hand fishingnet, nigisi, Gaima. The net is fastenedby a strip of bast in a continuous spiral to a circular ring of ratan, 61 cm. (24 iniches)in diameter. The net, whichhas a depth of about 217 mm. (8- inches),has a rim of continuous loopinginto which the loops of the outermostrow of the body of the net are worked(Fig. 8). The stringis two-ply. VOL. XLVI. 2 A

This content downloaded from 128.235.251.160 on Mon, 8 Dec 2014 22:56:06 PM All use subject to JSTOR Terms and Conditions 344 A. C. HADDON.- lte Kcabirior Girara District,Fly River, Paputa.

I know of nine tobaccopipes. Landtmangave sevento ourMuseum and one to theBritish Museum, and thereis one in the Port MoresbyMuseum, collected by A. P. Lyons at Iu. Theyall agree in consistingof one internodeof bambooand in having the backoroundof the design scrapedand colouredbrown, so that the light design shows up clearly. Five of them(L 147, 149, 150, 151 and Pt. M.) have the septum intact at both ends,and in additionto theordinary hole forthe bowl thereis a second one in a correspondingposition near the otherend; it is not possibleto distinguish betweenthese two,as theyappear to have been used indiscriminatelyfor the bowl. This is the firsttime.that pipes with two lateral holes and blind ends have been recorded. I find,however, that in our Landtmancollection there is one (164) from Tirio or Madiri (on the rightbank of the Fly estuary and oppositeto the Kabiri district),which is of similar type and technique of decoration,and though the designis not quite like any I have seen fromKabiri, therecan be no doubt that it is an importedspecimen. [There is also a fairlytypical richly engraved Kiwai pipe (L 161) of two internodesand a terminalhole, but a second small hole has been made near the open end, possiblydue to Kabiri influence.] L 147 has the centreof one end blackenedand surroundedby two rows of Abrus seeds,while in L 148 the hole at the open end is pierced througha small disc of pearl shell, which is surroundedby a single circle of Abrus seeds, the black ends of which are centri- petal (Fig. 9, C, D). Most of the designs on the pipes are representationsof animals or parts of animals-among these I thinkI can recognizea frog(152), hind legs ofa frog(149), fresh-waterturtle (148), snake(150), crocodile(B.M.), two newly- hatchedhornbills (Landtman's indentification, 151), a birdor turtle,and a fish(B.M.). A lime gourd,Gairna, with a pith stopper(Fig. 10), has on one side an enigraveddesign of a central circle and a spiral line eilclosingit, the outer whorl of the spiral with transversecarved lines; an attemptat the samnedesign on the opposite side. One lime spatula (L 294) is a sharply pointed splint of cassowary bone, 24 cm. in length; the other(L 295) is similar in form,but made of,the rind of the stalk of a palm leat, length 23-5 cmn.(L 147); alnother type is shownin Plate XIX, Fig. 3. In our Landtman collection there are two drums labelled "Kiwai," which, without any doubt, must have

, /X b come fromKabiri (Fig. 11). Their great lengthand large 0\'tJ, highhandles distinguish them from any other type of drum fromNew Guinea, and theirdecoration settles their origin. They are cylindrical,but are slightlyand gradually con- stricted towards the centre. The tympanumis made of the skin of a Tetrodontfish with extremelysmall dermal

FIG. 10. spines,and is cementedon; the lashing of A is rough and Limegourd with pith appears to be put on fora temporarypurpose. The plaited stopper,Gaima (L 542): 1. 23 cm. ratan band near the handle of A was put on to strengthen

This content downloaded from 128.235.251.160 on Mon, 8 Dec 2014 22:56:06 PM All use subject to JSTOR Terms and Conditions A. C. HADDON.-The Kabiri or Girara District,Ply River, Papua. 345

the drum,as it was cracking; the neatly plaited ring in the handle doubtless servedfor suspension.The handle in both is carved in simplepatterns. The main carvingis at the open end. In A thereis a double-headedfresh-water turtle (?) in low relief; betweenthe fore and hind limbs is a circular design,in the centreof which is a simalldisc of pearl shell surroundedby Abrus seeds (this has entirely disappearedon one side). In B thereis a bilobed designin the place corresponding to the hindliinbs of the tuirtleon the formerdruini; the annular designis replaced by concentriccrescents, the rest of the open end is occupiedby simple patterns. The intaglioparts have been painted red and white.

FIG. 11. Drums A (L 513), 1. 127 cm.,diam. 15 cm.at tympanum,14 x 16 cm. at open end; handle21 cm.long, 14.5 cm.high. B (L 510),1. 108 cm.,diam. about 13 cm. at each end; handle 25 cM. long,8-5 cm. high,

FIG. 12. Drum, Isagu, Port MoresbyMuseum (froma photographand sketch): 1. 145 m. (57 inches),diam. at tympanum20 cm. (8 inches).

In the Port MoresbyMuseum there is a drum obtained by A. P. Lyons at Isagu, which is of a differenttype (Fig. 12). It is cylindrical,with jaws at one end,and at the otheris a tympanummade of kangarooskin; the ratan bands which keep the skin in place have splintsof bambooinserted beneath them,which thus act as wedges. The druinis carved in the typicalmanner. We have no inforinationconcerning the followingobjects, most of which are doubtlessceremonial. Two human effigiesare knownto me: one is in privatehands at Daru, Papua the other,which came from Pigi, was given to me by Mr. J. E. N. Westwood (Fig. 13). The body and limbs are cut out of a flat board,the head, hands,and feet being carved out of a soft,pith-like wood, and are removable. On the head is the conical cap; the eyes are of pearl shell,with a circle of Abrus seeds with the black ends centripetal; the nose is rather narrow and the lips thin but very 2 A 2

This content downloaded from 128.235.251.160 on Mon, 8 Dec 2014 22:56:06 PM All use subject to JSTOR Terms and Conditions 346 A. C. HADDON.-The Kabiri or Girara District,Fly River, Papua. protuberant. There is a hole in the chest; the navel is a Helix shell surrounded. by Abrus seeds; the wristletsand armletsare of twistedfibre. The painting of the body is indicated in the figure, as are the pigments employed, viz.,. white,black, yellow ochre,and a reddishchocolate. At Daru I saw two small mlodelsof human heads,one painted red and black,. the other white with a large black ring round the eyes; in both the eyes are of pearl shell surroundedby Abrus seeds \\{ t B(Plate XX, Fig. 6). Theformer was very similar to a specimen (Fig. 14) from Kubu, and describedas "a head used for decoratinga shield-shapedmask." It is carved in the same softwood as that of' the effigy,and painted white, black, red and chiocolate;the niosehas a white tip,. and there is a white disc in the centre of the forehead; the eyes are of black

encircledcement,with with a centralAbrus seeds.plug of Landtmanwood,and also gave a verysimilar specimento the British Museum, which is 18 cm. long,. but it has a sagittal crest. In all these specimens the face is long and narrow, the nose is rathernarrow, and the tongue is protruding.

FIG. 13.-Human effigy,Pigi: ht. 121 cm. FIG. 14. (471 inches),br. 43 cm.(17 inches). Model ofhuman bead, Kubu CambridgeMuseum. (L 313): 1. 23 cm. (9 inches).

This content downloaded from 128.235.251.160 on Mon, 8 Dec 2014 22:56:06 PM All use subject to JSTOR Terms and Conditions A. C. HADDON.-The Kabiri or Girara District,Fly BRiver,Papta. 347

Landtmangave to our Museuima long ovoid mask(davi) fromKubu, made out of one piece of very light,pith-like wood (Fig. 15). There is a narrowstrip of ratan round the edge,which is sewn over with ratan. The back has two longi- tudinal ridgeson each side,into the outerof whichthe plumesare inserted; a stick, passed right across themat about the lower third; at the junction of the ridges near the top,a stopper-likepiece of wood is inserted; near the lower end is an orificewhich opens out on the frontside in the two eyeholesof the mask. The

FIG. 15. Mask,Kubu (L 312): 1.1 m.,br. 28 cm.; frontand back views. front is painted white,with a design in black, yellow,and chocolate. Near the centreis a raised disc of cement borderedby Abrus seeds; doubtlessoriginally it carrieda disc of nacre. At the lower end is a human head, 16 cm. long,painted white with a black border round the eyes and mouth; the large orbits are perforated,the upper borderand lobe of each ear has a tassel of red calico,the tongueprotrudes slightly. The peripheryis decoratedwith plumes of feathersof cockatoosand the hackle feathersof whitecocks, etc. The centreone is 9-shaped,

This content downloaded from 128.235.251.160 on Mon, 8 Dec 2014 22:56:06 PM All use subject to JSTOR Terms and Conditions 348 A. C. HADDON.-The Kcabirior Girara District,Fly River, Papua. the centre beinga solid piece of pith; the stemis a strip of ratan neatly served with bark fibrewhere it is perforatedfor the quills of the feathers. These are white,and cut offsquare. There was at Daru a painted truncatedoval board bound roundwith ratan, witha typicalhead wearinga conicalcap in highrelief in the lower portion; above this is an oval design surroundedwith white feathers,and above this again two concentriccircles. Each of these two designs has in its centre a disc of nacre

FIG. 16.-Cerenmonial tablet with two snakes FIG. 17.-Ceremonial hook, in high relief, Gaima: 1. 50 cm. (193 inches); Pigi ?: 1. 43 cm., diam. 27 mm. br. 34-5 cm. (13? inches). Cambridge Museum. Cambridge Museum. encircledwith Abrus seeds (Plate XX, Fig. 4). The Rev. E. Baxter Riley,of Daru,, gave me a very old shield-shapedthin slab of wood fromnGaima (Fig. 16), on which two snakes are carved in high relief. These are provided with teeth, pierced nostrils,and raised eyes of nacre and seeds; in the centre a Helix shell and rowsof Abrus seeds are imbeddedin a raised mass of cement; the background appears to have been black, the designson the shield and snakes being dark red and chocolate. In the museum at Port Moresby there is an oval slab, 53*5 by 38 cm. (21 by 15 inches),collected by A. P. Lyons in 1914; it is labelled " diwadi, a dancing ornamentworn on the head." There is the usual centraldisc of nacre and ring of seeds surroundedby three paintedinterrupted circles, the outermost

This content downloaded from 128.235.251.160 on Mon, 8 Dec 2014 22:56:06 PM All use subject to JSTOR Terms and Conditions A. C. HADDON.-The Kabiri or-Giragla Distritr, Fly River, Patpua. 349 of which has anigledprojections; the whole is painted white,red, chocolate and yellow. Lancltmangave to the British Museum a longCovoicl painted board, 91 by 28 cm.,from Kubu. The disc of niacreand ring of seeds is surmountedby a white,a yellow,and a red semicircle; near the lower end a mouth is painted, above whichis a tranisverserow of colouredspots. The otherdesigns are simple.

WS~~~~~~~~~V

FIGS. 18, 19. Models of crocodiles'heads: 1. 36-5and 20-4cm. (141 and 8 inches). CambridgeMuseun .

FIG. 20. Dance ornament,Kubu (L 314): 1. 103 cm. (40-sinches).

Mr. Westwood also gave me a barbed wooden hook anid two carvings of crocodileheads. The hook (Fig. 17), which does niotappear to serve any practical purpose,has its upper part carved as a human head, wearing a conical cap and havingprominent nacre eyes,a piercednose, and slightlyprominent yellow tongue; the rest is painted in black,white, yellow and red bands. The crocodile heads (Figs. 18,19) are carved out of a verysoftlight wood; the largerone has tuftsof a vegetable fibreknotted throug(rh the nose and ears. Some of the fibrespassing

This content downloaded from 128.235.251.160 on Mon, 8 Dec 2014 22:56:06 PM All use subject to JSTOR Terms and Conditions 350 A. C. HADDON.-The Kabiri or Girara District,Fly River, Papua.

throughthe ears are knottedon the top of the head to keep the tassels in place. The eyes are raised and are providedwith Abrus seeds and probablyoriginally with nacre. In the smaller one the eyes are painted,and a strip of ratan,to which small white feathersare lashed, passes along the top of the snout. Both are painted in the usual way, and the teeth are white. A " dance ornament" fromKubu coinsistsof a log of lightwood witheach end carved to representthe head of a crocodile,or possibly of a boar (Fig. 20), in which,as in Fig. 18, the undersideof the jaws is entirelycut away. A noticeable featureis the pair of large tusks. The ground colour is black, the designs being leftin the natural pale colour of the wood,or painted white,yellow, or chocolate; the eyes are painted. I saw at Daru some smiallcrocodile skulls in whichthe lowerjaw was lashed to the upper; the eyes were discs of nacre with a ring of Abrus seeds,and the tip of the snout of one was adornedin a similarmanner (Plate XX, Fig. 5).

THE AFFINITIES OF THE KABIRI. Despite our imperfectinformation, I think the followino deductions are warranted. Judging fromphotographs taken by Landtman,there are two well- markedphysical types among the population: the one with a narrow face, thin lips, narrownose, and apparentlya bulbous forehead,the otherwith a broaderface, thickerlips, and typicalplatyrhine " Papuan " nose. The firsttype is that carved in the human effigiesand heads,and we may thereforeregard it as the racial type of the stock; the second type is that characteristicof many Western Papuans (Plate XIX, Figs. 1, 2). I have not yet been able to followup the Kabiri type,it is not representedby Neuhauss in the T7dlker-Atlas,Deutsch Ne-GCuinea, II. The wearingof nettedcaps is commonin amongthe Kai of Hupe (in the interior,N. of Huon Gulf),and at the coast about 60 S. lat.' Conical caps, or "head-baskets,"of variouskinds are worn from south of 40 S. lat. to about 143' 30' E. long., and on the neighbouringislands2; and about 30 miles up the river3; but 0 Reche says that this head-gear is widely spread along the coast, but quickly disappearsand is rare up the rivers.4 I am not yet in a position to discuss the distributionalsignificance of the various shapes of drumsin New Guinea, but the two main typesmet withhere and in the whole Fly River districtand in TorresStraits seem to indicate two cultural influences. I do not lay any stresson the single example of the type withjaws having a kangaroo-skintympanum (Fig. 12), but the occurrenceof mammal-skin tympanais of interest. Van der Sande5gives various instances oftympana other than the ubiquitouslizard (Varanus) skin. Out of twenty-fourdrums described by

1 Neuhauss II, Plates 198-201,238, 239. 2 0. Finsch, atmoafahrten,1888, pp. 292, 299, 302, 306.- 3 NeuhaussII, Plate 277. 4 Kaiserin-Augusta-Fluss,1913, p. 84. 5 Nova Guinea, III, 1907, p. 305.

This content downloaded from 128.235.251.160 on Mon, 8 Dec 2014 22:56:06 PM All use subject to JSTOR Terms and Conditions A. C. HADDON.-The Kabiri or GiraracDistrict, Fly River, Papua. 351 de Clereq and Schmeltz,eighteen have skinof Varanus, one of ray,one of cassowary, two of opossumfrom Misol and Salwatti,one of tree kangraroofrom Mar (Wewe, the mostnortherly point of New Guinea), and one of deer obtainedat Waigiu, but importedfrom Tidor; these all point to Indonesianinfluence. MacGregor says that the Banga who live fiftyor sixtymiles up the Moreheadriver, " use the skin of the wallaby on the drum."' Seligmaninand Strongstate that "the tympana of their [of the Toro, thirty miles up the ] drums consist of kangaroo .skin,and the drumsthemselves were largerthan those we had previouslyseen in this district. It should, however,be noted that far largermammal-skin covered ,drumsare said to be used some distanceup the ."2 This last statement ,can scarcelyrefer to the Kabiri district,as it had not then been visited,but doubt- less these tympana are more widely distributedin the Bamu district. J. W. R. Koch refersto a drumfrom South New Guinea witha kangaroo-skintympanum.3 When in TorresStraits in 1914, Mr. C. H. Walker gave me a drum of the well- known Tugeri type,with a kangaroo-skintympanum; it was stated to come from the Bugi, whose home seems to have been mainlyto the west of the mouthof the Mai Kasa. Landtman obtainedseveral drums of the " Tugeri" type in a village a little inland of " Budji " (Bugi), but he was informedthat they are not manufacturedby the coast people, who obtain them from furtherinland. The same holds good for the beautifullydecorated arrows which are also sometimes describedas " Tugeri." This termseems to him to be somewhatuncertain, as it is used in rathera vague sense forthe unknown tribes who in formertimes raided the Britishterritory at the frontier. The Kabiri drum(Fig. 12) is the only one, so faras I am aware, fromBritish New Guinea withwedges. This methodof tauteningoccurs on the south side of MacCluer Gulf,on Salwatti,and at Wewe (de Clereq and Schmeltz)4;but the skin is stretchedin a morecomplicated manner,similar to that in Indonesia; Van der Sande recordsit fromMios Korwar,an island in GeelvinkBay,6 and 0. Reche figures fourspecimens from various places on the Sepik,6one being 252 km. up the river; his fig.450 is the only drumdescribed by Reche with a tympanumof intestine, it came fromMandanamn, 194 km. up the river; all the othershave Varanus skin. The decorativeart has several distinctivefeatures. The prevalence of spiral and curvedlines seems to be characteristicof the great series of migrationsfrom the interiorwhich reached the sea at Merauke,the Fly estuaryat this spot,and the Papuan Gulf furthereast. The interruptedcircles are characteristic,as are also the discs of nacre surroundedby Abrus seeds. The use of variously coloured earthsis also noticeable. The ceremonialobjects, like the decorativeart, of the Kabiri are unlike any otherspreviously recorded from New Guinea, and offera markedcontrast to those

I Ann. Rep.,1895-96 (1897),p. 43. 2 Geog.Journ., xxvii, 1906, p. 229. 3 De ZuidwestNiew-Guinea-Expeditie, 1904-5, Leiden, 1908,p. 598. 4 Nos. 653,655, 658, pp. 154, 155. 6 Loc. cit.,p. 306. B Kaiwerin-Agusta-Fluss,1913, Figs. 448, 450, 456, 458.

This content downloaded from 128.235.251.160 on Mon, 8 Dec 2014 22:56:06 PM All use subject to JSTOR Terms and Conditions 352 A. C. HADDON.-The Kcabirior Girara District,Fly River, Papua. of the coastal tribesof the Fly estuaryand of the rivers to the east. Neither is' thereany close resemblancein the materialculture to that of the Sepik (Kaiserin, Augusta river),so faras our presentevidence goes, though the resemblancesin social, and religiousculture may prove somewhatcloser. In mnypaper on "Kava-drinking in New Guinea,"' I suggestedthat the Tugeri and the Kabiri have both come from the interior of New Guinea. 0. Reche2 draws attention to the remarkable similarities between the inhabitantsof the S6pik and the TaLgeri,instancing mourning caps wornby women, tobacco cultivation,use of sago and betel, piercing of the ale nasi, head hunting, spiral patterns,and even the physicaltype is strikinglysimilar according to Poch.3 It is probablethat the people of the middleregion of the Fly, MurrayLake, and partof the Stricklandriver beloing to the same culture group as the Kabiri, this area is characterizedby stuffedhuman heads.4 Skulls with painted clay faces occur in abundance,not only on the Sepik, but fromthe Fly to the Kikori in the Gulf. W. Foy5 firstnoted that cane masks of very similartype are found on the- Sepik and somewhatinland fromthe estuaryof the Fly River; they also occur on Goaribariisland.6 Farther east, especially amongthe Namai of the Purari delta. and the Ele,ma tribes of the Papuan Gulf,is anotherdistinct culture which also appears to be in some way related to part of the cultureof Central New Guinea. I amigreatly indebted to Dr. Gunnar Landtnmani,of Helsingfors,for his kind permissionto reproducehis photographs,Plate XIX, Figs. 1-4, and Plate XX, Figs. 1-3. Figs. 4-6, on Plate XX, were photographedby my daughter,Kath- leen Haddon, at Daru. I have also to thank Miss Estelle Canzianifor the illustra- tions in the text. APPENDIX. SlNCE, the foregoingwas in type I have heard fromLt.-Governlor J. HI. P. Murrayof Port Moresby,who has recentlyvisited the Aramia district,that the real name of the people is Gogodara-niotGirara or Kabiri. They live in a large house called ganamna,of whichthere is. only one in a village ; the women live in compartmentsunder the same roof,but with a differententrance. They have a rough sort of lamp, nakeo,made generallyof stone, but. sometimesof wood,hollowed out in the centreto receivea resinoussubstance, gagaba, which burnssteadily and gives a good light; but it burns ratherfast. It has a pleasant smell and drives away mosquitoes. The wiki is also called garabo; it gives a good light and burns brightlyand well; if you sticka piece in the trunk of a coconutpalm you can see to climb. up and selecta nut by night. They have very finegardens, which are manuredwith wallaby- droppings. Thereis plentyof gamada.

DESCRIPTION OF PLATES. PLATE XIX. Fig. 1.-Man of the narrow-facedtype with conicalhat, Kubu. ,, 2.-Bearded man of the broad-facedtype with mourning netted cap and fringedepending; fromneck, Gaima. ,, 3.-Man sittingand chewingbetel, inside a house,Gaima. ,, 4.-Woman wearingmourning netted cap, Gaima. PLATE XX. Fig. 1.-Entrance to the commonhall forthe men at the gable end of a house,Gaima. 2,-Entrance forwomen through the side of the roofof a house, the loop-holesevidently- serveas windowsfor the women'sapartments, Gaima. 3.-House withhand fishingnets hanging from the verandah,Gaima. 4.-Painted board with a woodenhead and conicalcap in high relief(Kubu ?). ,, 5.-Four small decoratedcrocodile skulls (Kubu ?). ,, 6.-Two carvedwooden heads (Kubu ?).

1 Man, Oct., 1916, vol. xvi, 87, p. 145. 2 Loc. cit., p. 480. 3 Zeits.f. Ethnol.,xxxix, 1907, p. 391. 4 L. M. D'Albertis, NewGuinea, II, 1881, fig.,p. 134. 5 Globus,lxxxii, 1902, p. 380, figs. 2, 4. 6 J. H. P. Murray, Papua, or British New Guinea, 1912, p. 189, Plates, pp. 187, 204.

This content downloaded from 128.235.251.160 on Mon, 8 Dec 2014 22:56:06 PM All use subject to JSTOR Terms and Conditions Journalof the Royal AnthropologicalInstitute, Vol. XLVI, 1916, Plate XIX.

FIG. 1. FIG. 2.

3.:Xt, FIG. 4.i S. FiG.

T^'' KAB'RI OR GIRARA DISTRICT.

This content downloaded from 128.235.251.160 on Mon, 8 Dec 2014 22:56:06 PM All use subject to JSTOR Terms and Conditions Journalof the Royal AnthropologicalInstitute, V'o0. XL VI, 1916, Plate XX.

FIG. 1.FIG. 2.

5*FG.6 FIG.P _~~Tl KIRORGRRDSTC.

FIlG 3. FIG. 4.

This content downloaded from 128.235.251.160 on Mon, 8 Dec 2014 22:56:06 PM All use subject to JSTOR Terms and Conditions