-^v — iifc/ •?^^ ^-^-^ r^i it] '^.. .-i ^i& ?^^ It^- l£8 HARVARD UNIVERSITY. :ii::iii I^ I B K A R Y MUSEUM OF COMPARATIVE ZOOLOGY. h JWaxJIvAo, V^ii^-^i^A/v^xoxx. ^, \°\<:)<^ a. 2 o — oc ca oo — > 2 < I ^ I w -J <: X ''\\0 MEMOIRS OF THE BERNICE PAUAHI BISHOP MUSEUM OF POLYNESIAN ETHNOLOGY AND NATURAL HISTORY VOLUME II. - t. honolulu, h. i. Bishop Museum Press ^ I 906- I 909 \ '\ TABLE OF CONTENTS I. Hawaiian Mat and Basket Weaving: By William T. Brigham. Hawaiian Nets and Netting: By John F. G. Stokes. Issued February, igo6. II. Old Hawaiian Carvings: By William T. Brigham. Issued May, igo6. III. The Ancient Hawaiian House: By William T. Brigham. Issued June, igoS. IV. Kilauea and Manna Loa, X'olcanoes on the Island of Hawaii: By William T. Brigham. Issued Deeember, igog. LIST OF PLATES L Hawaiian Mat and Basket Weaving. — Hawaiian Nets and Netting. I. Micronesian Huskets. II. Micronesian Baskets. III. Hawaiian Sandals. IV. Hawaiian Fish-baskets or Traps. V. Corner of Pandanus Leaf Mat. VI. Plait in Hinai poepoe. VII. Cover of leie Hasket. VUI. Basket Covering of a gourd. IX. Hinai poepoe. X. Beginning of a Makaloa Mat. XI. Portion of Makaloa Mat. XII. Portion of Makaloa Mat. XIII. Portion of Makaloa Mat. XIV. Group of Modern Hawaiian Fans. XV. Native Scraping Olona. XVI. Makaloa Mat in Color. Frontispiece. 2. Old Hawaiian Carvings. XVII. Auniakua found in Cave. 3. The Ancient Hawaiian House. XVIII. Pago Pago Harbor. J. Martin. XXX. XIX. King's House, Mbau. Waitovu, Fiji. Liudl. XXXI. XX. Na Kali ; Fijian House. Lindt. XXXII. XXI. Maori Carved House. " XXXIII. XXII. Maori Carvings. XXXIV. XXIII. New Hebridean Huts. XXXV. XXIV. Communal Hou.se. New Guinea. XXXVI. XXV. High House, New Guinea. XXXVII. XXVI. Hawaiian House Framing. XXXVIII. XXVII. Hawaiian House Thatching. XXXIX. XXVIII. Hawaiian House Completed. XL. XXIX. Hawaiian Cords. Ipu holoi lima. Ipu aina with Inserted Teeth. Ipu kuha or Spittoons. Gourd Bottles for Fishing Lines. Hawaiian Stirrers and Knives. Hawaiian Carvings in British Museum. Decorated Gourd Vessels. Carved Coconuts. Hawaiian Unieke. Hawaiian Unieke. Hawaiian Unieke. 4. Kilauea and Mauna Loa: Volcanoes on the Island of Hawaii. XLI. Lava Fall, near Hilo. XLII. Descent into Halemaumau. XLIII. Sentinels and Overflow in Kilauea. XLIV. Pools in Halemaumavi. XLV. Cone and Pool, Halemaumau. XLVI. Drawn Lava from Fall. XLVII. Aa, Natural Size. XLVIII. Cave Stalagmite, Slender. XLIX. Cave .Stalagmite, Stout. L. Pool in Halemaumau. LI. Pool in Halemaumau from below. Lll. Pool in Halemaumau, from above. LIH. South Lake and Island. LIV. Cone in Halemaumau. LV. Double Cone in Halemaumau. LVI. Cone in Halemaumau. LVII. The Empt}' Pit. LVIII. On the Brink of the Pit. LIX. Cracked Banks of Lava Pool. LX. Dana Lake. LXI. South Lake. LXII. Floating Islands. LXIII. Flow of Aa. LXIV. .Source of Flow 1880-81. LXV. Crack in F^loor of Kilauea. LXVI. Kilauea in 1896. D. Howard Hitchcock. LXVII. Mokuawcowco in 1896. D. H. Hitchcock. LXVIII. Halemaumau, 1880. Furneaux. Frontispiece. Ciii) ILLUSTRATlONvS IN THE TEXT While eHL-h Memoir is puBt-cl iiulepeudeutl.v. tbe paginatiou uf the volume will be round at the bottom ot the page. Memoir 1. 1. Coconut Leaf Basket 3 2. Hawaiian Ki Leaf Basket 4 3. Coconut Leaf Baskets from Tutuila 5 4. Coconut Leaf Basket. Sliortland 1 6 5. Palm Basket. Ponape 7 6. Coco Leaf Basket. Caroline Iskiiuls 7 7. Closure of Basket Bottom 7 8. Fans from Solomon Islands 8 9. Hawaiian Coco Leaf Fans 9 10. Hawaiians I^ans, British Museum 10 11. Ancient Hawaiian Fan, Bishop Museum 11 12. Various Fans 12 13. Samoaii Fans 12 14. Structure of Fans 13 15. Fijian Palm Leaf Fans 14 16. Marshall Islands Fans 15 19. Caroline Islands Basket 16 20. Course of Strips 16 21. Solomon Islands Shield 17 22. Solomon Islands Rattan vShields iM 23. Gilbert Islands Coco Fibre Armor 19 24. Ponape Sennit Basket 20 25. Rolls of Sennit 21 26. Tongan Basket, British Museum 22 27. Tongan Open-work Mats 22 28. Coco Cords from Micronesia 22 29. Palm Leaf Hats. Guam 23 30. Palm Leaf Basket. Solomon Islan ' y ' n Fig. I. COCONUT leaf basket. makaloa {Cypents hcrigatiis^, a sedge from whose tender stems the fine Niihan mats were woven; olona {Toiic/iardia la/ifoh'a), invalnable for its tough and durable fibre, from which were twisted cords for fish lines, nets, etc.; mamaki {Piptunis albidiis), used mostly for felting into coarse, tough kapa, but also for cords and ropes; ieie {Frey- cinetia aniotti, and in Tahiti F. deiiiissa'), from whose aerial rootlets the choicest and most durable of the Hawaiian baskets were made; iwaiwa, several ferns from whose stems were plaited fish traps and baskets. Other fibrous plants of less extensive use will be noticed as they occur. Mat and Basket Weaznng. With this wealth of material, mostly growing wild, but in some cases, as waoke and olona, cultivated for convenience or to improve the quality, it is not strange that the islanders made their clothes (scanty indeed, but sufificient ), their beds and pillows, carpets and house linings, baskets, shoes and hats, fish lines and nets, sennit, cords and ropes of the varied form and quality we are to consider in this chapter. While the use of some of these fibres for felting in the interesting manufadure of paper (kapa), for which Polynesians are celebrated, must, from its importance, be treated in a separate chapter, the other uses of a textile nature will be treated here, arranged for convenience under the following heads: — Pal.m lSavKS and FiiiRHS. — Bris- kets, Fans, Sennit. Hats. Pai.m Stems. — Shields. Pandaniis. — Hats, Mats. Pillows, Baskets, Sails, Mat Garments, Cord (covered), Fijian Basket. FrEYCINETI.\ ROOT.S. — Baskets. Fern Stems. — Baskets, Fish Traps. Grass. — Makaloa Mats, Rush Mats, Cord, Bambu Fans, Combs, Spears aniioaii Mat Making. 41 "Upon the completion of one of these valuable mats. . . .all the women familiar with the mauufa(?tiire of these mats resident in the neighborhood were sxunmoned on a given day to bathe the mat. On the women assembling they proceeded to wash the mat in fresh water, and after well stretching it out to dry they adjourned to the house to partake of a feast provided by the hostess to celebrate the completion of her mat." How clear the piAure of these children of Nature assembling on the bank of one of the man}- Samoan streams to wash the mat which, after perhaps a year's work, one of their number had finished! They all rejoiced, for was not the work an honor to their village, to their sex, to their friend ? And if the customs of both children and domestic animals (dogs and pigs) were then as now, the washing must have been more than an idle ceremonj-. I am well aware that a Samoan house had a low fence across the doorways, intended to keep out pigs, for on this I have sat while chatting with the inmates of more than one Samoan grass house, both on Upolu and Tutuila ; but I also know that in modern times at least it is customary to take the growing mat out under a shad}' tree where both pigs and hardlj- less dirty children could scarcel}^ be kept from off it. The washing we may be sure was needed. Our author goes on to say : "There were also at least thirteen other kinds of clothing, sleeping and house mats made bj- the Samoans.' Various dyes were prepared from vegetables and roots of trees. A beautiful crimson was obtained by mixing the inner bark of the root of the noniifiafi'a {Eiigrtiia tnalaaensis) with sea water and lime. Yellow was prepared from turmeric and oil. It was also obtained from the bark of the iioii/i {Alon'iida citrifolia^ previously mentioned. A fine purple was obtained from the 3'oung shoots of the mountain plantain, soa\i {Miisa fc/ii)^ and a brown by mixing the inner bark oi the paiii (1 ) with sea water. A black colour was imparted to various articles by burying them in the soft mud of a taro patch formed in a swamp." Continuing our exploration of the technic of the islanders we find on Fiji, ac- cording to Dr. Berthold Seemann,'" a most trustworthy authority, that "Mats with which the floors of houses and sleeping places are thickly covered, are made of two kinds of screw pines ; the coarsest of the leaves of the Balawa {Paudanus odoraiissinnis^ Linn.); the finest, of those of the Voivoi {P. can'cosns, Rumph.). The Balawa, or Vadoa, as it is termed in some distridls, is a tree twentj'-five feet high, indicative of poor soil, growing in exposed positions, and being one of the first plants appearing on newly formed islands. Its singular habit has often been dwelt upon. The smooth white branches, with their dense heads of foliage, not inaptly compared to the arms of a huge candelabrum ; the strong aerial roots, covered with minute spines, and serving as so many props ; the curious corkscrew-like arrangement of the leaves, the leathery, 'Some of these mats were laufau, lalaga, ie taua, ie 'ula, ie sina and si'aloa. '°Viti. B. Seemann. Cambridge, 1862, p. 554. 42 Mai a7id- Basket Weaving. sword-shaped leaves themselves, and their spiny edges ; the long spikes of male, and the shorter branches of female flowers, their delicions perfnme strongly recalling to mind that of the vegetable ivory of South America; finally, the bright orange-coloured drupes, formed into large heads of fruit, to say nothing of their insipid taste, appre- ciated onh' by natives, are all so essentially different from what a European traveler is accustomed to in his own country, that his attention is involuntarilj- arrested, and he hardlj' ever fails to record it. The Voivoi or Kiekie is a stemless species with leaves ten to twelve feet long, whicli delights in swampy localities of the forests, and is occasionalh' cultivated to meet the demand. Fans, baskets, and the finest mats — even those on which newl\- born babes, naked as thej' are for more than a twelvemonth, are carried — are made of its bleached leaves. Occasionally neat patterns are worked in by introducing portions of the material dyed black, whilst the borders of highly finished mats are tastefully ornamented with the bright red feathers of the Kula — a parroquet {Calliptihts solitariiis^ Latham) not found in the groups eastward of Fiji, and therefore highly esteemed by the inhabitants of those islands." Turning to the westerly Pacific region, we find in the Solomon Ids., according to Dr. Guppy," still another item in the technic of mat-making. I have not been fortunate enough to see any of the Solomon mats, of which there are a number in the Bishop Museum, that I could feel sure had been treated as Dr. Guppy describes; but his statement is interesting as adding to the methods used in preparing the pandanus leaves. He says: " Mat-making is one of the occupations of the women of the Straits, the material employed being the thick leaves of a species o{ pandaims which is known by the natives as the: pota. The leaves are first deprived of their thin polished epidermis by being rubbed over with the leaves of a plant named saiis7ifi\ which have a rough surface giving a sensation like that caused by fine emery paper when passed over the skin. The pan- danus leaves are then dried in the sun, when they become whitened and leather}-, and are then sewn together into mats." Evidently the kind of mat already described from the Solomon Ids., p. 32. In the Gilbert Ids. we learn from the narrative of the United States Exploring Expedition under Wilkes," "The mats are made of the leaves of the pandanus, slit into strips about a quarter of an inch wide, and woven by hand : these are of two colours, light yellow and dark brown : the former are made from the young leaves, and the latter from the old, which are prepared b}- beating them with a mallet to render them pliable. On the yellow mats the}' bestow a great deal more of their attention : the 30ung leaves "The Solomon Islands and tlicir Natives, London, 1SS7, p. 61. '-Narrative of the United .States Exploring Expedition during the years 1838-1842, by Charles Wilkes, U. S. N., Commander of the Expedition. Philadelphia, 1845. Vol. v, p. 94. Marshall Islands Mats. 43 are laid aside for two or three days after they are plucked, till they are withered : they are then roasted by holding them in the hand over the fire, and afterwards laid in the snn for three or four days to insure them being sufficiently dried. During the latter part of the process they are brought every evening into the house to protect them from the dew or rain. When the leaves are sufficiently dry, they are left all night to bleach in the dew: they are then rolled up in balls and pounded with a mallet to render them soft and pliable, and when this is accomplished, they are slit with a shell and are read}' for use." It may be stated that the pandanus mats of the Marshall Ids. have long been an important article of commerce among the islanders, and the low coral islands of the T«*rr*" ^-^^^S^^-^M-J=?i>^^^^ Fig. 49. MARSHAI.I. ISLANDS MAT AS USUAI,I.V FOLDED. NO. 8o6l. group are largely dependent on the sale of their mats and the produAs of the cocouut for such things as their low and sand}' soil cannot produce. Hence these mats are found on most of the groups in the Pacific, and half a century ago they were often brought to Honolulu in the missionary vessels, and I have at times been led to sup- pose that they were of native Hawaiian manufacture. While the Hawaiians did make similar mats, those of the Marshall and Gilbert Islanders can generall}^ be distinguished from most other products of the mat makers. The mat figured is one of the better class, and measures 11 ft. 2 in. by 9 ft., with a fineness of four to the inch. The dark and light leaves mentioned b}- Wilkes are ver}- distinct, and by ingenious combinations 3neld very attradlive mats. The custom of keeping fine mats for a long time, using them only on great occasions, seems very ancient, and, if the word of some native Sanioans may be trusted, mats several hundred years old are now in existence. The Samoan legend of the origin of one of these mats is sufficientl}' quaint to warrant its translation here, although 44 Mat and Basket Weaving. it lias been translated into German elsewhere. '^ I do not pretend to be a Samoan scholar, but the similarity with the Hawaiian has warranted me in translating the song from the text of O. Steiibel.'^ There are expressions of which the meaning can onh' be guessed at, and even the learned translator into German, Dr. W. von Biilow, has not always been sure that he has guessed aright. There are several versions of this story ; the present one was colle^led at Safune and Dr. von Biilow gives another not very different, in the volume referred to. The stor}' is in genuine Polj'uesian form, and serves to explain the reverence with which certain fine mats are to this day preserved by the Samoans, wrapped in siapo and care- fully stored among their choicest treasures. O LR TAI.A I I,K IR. O Neefanua ma si ana tama o le pipili ma lona tuagane o Uu (Vaiuu) ma o Ololua (Lolua): ua sau la latou vaa mai Fiti ma lalaga mai i ai o le ie i le vaa. Ua maua o le igoa o le ie o le Lagavasa. Uua taunuu ai i Siuutu i Salailua, ua toe lalaga ona toe sauni lea a oo mai i le Itu o taiic. Ua lalaga o le ie. ua pipii ma Ic cleele. Ona faaigoaina lea foi o pipii ma le elccle. Ona fua lea. Ua sau o le vaa i gatai o .Safune. Ua fai atu Uu, sei nofo o le vaa i tuaau, sei asia o le nuu nei po e i ai o se latou ava. Ua alu ai Uu, ua tui le ava. Ona alu ifo lea ona fai utui, oomai ia ua maua o le ava. Ona alu ai lea o le vaa ua oC i uta. Ona fai atu lea Uu, oomai ina oo i uta : au fia nofo i si ou ava. Ona ua nofo ai Uu ua faatupu maa i totonu o le ava, e i ai o le maa i nei onapo i le ituava sasae. Ua alu ai o le vaa, ua tuuta i uta, na taunuu ia Tugaga ; ua tulia ; Ona alu ane e moe i le fuefuc. Ona toi igoa ai o le ie Ua moe i le fuefue. Ona latou oo lea i le Ituvai i sasae, ua latou oo i Asu, ua lalaga foi o le ie : ua oo i onapo o le tuaoloa, ua agimai o le matagi, ua sau o le oneone i le fale, oneone ai o le ie. Ona tata lai lea o le ie, ua asuasu, ona faaigoaina lea o le maota o Asu, i le asuasu o le ie. Ua oo o le malaga o le alii Lealatelc o Tanuvasava- samanaia, ua afe i le fale o le Pipili i Safune. Ua avaga o le Pipili ia Tanuvasavasamanaia, ua fanau o Tualafalafa o le teine. "Archives Internationales d'Ethnographie, xii, 136. "Die Leute des Tagaloa, p. 144. TIIR TALR OF THE MAT. Neefanua with her daughter the cripple (hunch- back) and her brothers Uu and Ololua came in their canoe from Fiji, and she plaited a fine mat in the canoe. Then was given to the mat the name l.agavasa (woven on the high sea). They made land at Siuutu in Salailua and wove the mat again and traveled to Itu o tane. She wove the mat and the dirt clave fast to it. For this was the mat named Pijiii ma le eleele. More leaves were to be collected. The canoe went on to the seaward of .Safune. Then spake Uu, "The canoe must rest on the other side of the reef and a boat-passage must be found in the reef." Then went Uu to pick out a boat-passage. He returned and said, "Come, there is a boat- passage at hand." Then went the canoe through the I'ecf to tlie land. "Now," said Uu, "Come, go yovi ashore while I stay by my boat-passage." Then stayed Uu there and raised up a stone on the inner half of the entrance, that verj' night, a stone on the east side of the passage. The canoe made land and was drawn up on the beach, it came to Tugaga ; they went onward. She went thence and slept in the bush. From this was the mat called "Ua moe i le fuefue." She went then to the east side of the stream, and turned to her mat-weaving again. Then arose in the night the .southeasterly wind and blew the sand into the house so that the fine mat was full of sand. Then was the mat shaken : thence came the name of the chief's house, "Asu," from the shaking of the mat. The chief Tanuvasavasamanaia of Lealatcle was journeying and came to the house of Pipili at Safune. Then Pipili married Tanuvasavasamanaia and bore him a girl, Tualafalafa. Samoan Talc of a Mat. 45 (Ua e i ai i nei onapo o le ala o le Pipili i Safune.) Usu mai o le alii o Faitala mai Atua ia Tualafalafa, fanau Sina fafaga i tua. Usu ifo o Aoee i luga i le lagi ia Sinafafaga i tua, fanau Sinamaa Aoee. Ona gasegase ai lea o Aoee, ona feauina ifo ka o Sina niaa Aoee e alu ai i le gasegase o lona tama; ua faaputu toga ua ootai o le teiiie i lona tama, aua o le gafa ua ta i le lagi. E i ai o le upii, "O le gafa ta i le lagi." Ua fai toga. A ua muimui Tualafalafa, aua ua leai soona faitoga ; a afiifii lava i le aluga o le ie, ua igoa tele. Ua oo i le aso ua tele ai toga, ona tatala loa lea o le ie: Ua uila ma pogisa ma faititili ma afa sasa ia o le ie. Ua ofo o le toatele o tagata, ua latou faaigoa ai o le ie ia Tasi e afe. O le uiga o lenei igoa : Ua tasio le nei ie, a ua afe toga ua mauai. Ua tea i le lagi o le ie. Sa tepa ifo nei o Sina maa aoee ia Tooalo i RIanase, ua to tigapula, ua sili i le tua o se ie ula. Ona tagi ai lea o Sina mia aoee i lona tama, ona oo ane i ai lea o tama toalua o Ale ma o Ua faafuasei late momolia o le teine i le alii, ua (ia fai tane ai. Ua to o Uafaafuasi, tago loa Tooalo, sei o le ie ula e tu i lalo o le maa, a tu loa o le teine o Sina maa aoee, ua tepa atu, ua leai o se ie ula ; ona fesili lea o le teine: Alii e, oifea o le alii, sa totoo lona tigapula? Ua tali atu Tooalo; O au lava lenei. Ona oso ai lea o le teine, o lea e au i le vai, a ua afisi i lona aoao o le ie. Ua tatala o le ie, faa ua susu. Ua iloilo, ua mago lava. Ona faaigoaina lea o le ie ia Matumaivai. Ua nofo Sina maa aoee ia Matilafoafoa, ua fanau o Sinataeoilagi. Usu Tooalo ia Sinataeoilagi, fanau Sinaautuimo. Usu mai Leuluialii ia Sinaautuimo. fanau Muliaga ma Matagitausulu. Usu mai Tuisamoaia Matagitausulu, fanau Nonu- maufele. ( There is to this ila\' in Safune the street of Pipili. ) The chief Faitala ni li .'\tua m irrietl Tualafalafa and begot Sina fafaga i tua. Aoee came down from heaven, married Sinafafaga i tua and begot Sina maa Aoee. Aoee was ill, and Sina mia Aoee was sent down from heaven to cure the illness of her father: fine mats were collected, and the daughter goes to her father; so it was the genealogy towered toward heaven. Thence comes the saying, "The genealog}' is from heaven." Mats were collected and Tualafalafa murmured be- cause her mat had not assisted. Therefore was the mat with many names made into a pillow. As now many fine mats were collected, the mats were lying unfolded. Lightning and darkness, thunder and hurricane were the omen of the mat. A.stonished were the people, and they called the mat "One of a thousand." The meaning of this name is: One is this mat, a thousand would balance it. The mat so came to heaven. Then looked Sina maa aoee upon Tooalo in Manase, he was planting a taro top and wore on his back a red (feather) mat. Then did Sina maa aoee call to her father, and he brought two youths. Ale and Uafaafuasei, for the maiden longed to get the chief for a husband. Then fell a sudden rain and Tooalo caught hold of his feather cloak and hid it under a stone ; then at once stood the maiden Sina maa aoee before him and beheld him without the red cloak ; then asked the maiden : "O Chief, where is the chief who plants taro? To her replied Tooalo, "I am he." Then plunged the maiden into the water, and swam through the stream, taking the mat in her armpit. Then was unloosed the mat, lest it be wet. It was examined and found quite dry. Then was the mat called Matumaivai (drj- in the water). Then Sina maa aoee dwelt with Matilafoafoa and Sinataeoilagi was born. Tooalo married Si na tae iolagi and begot Sina- autuimo. Leuluialii married Sinaautuimo and begot Muli- aga and Matagitausulu. Tuisamoa married Matagitausulu and begot Nonu- maufele. We can pi6lure to ourselves the Samoan woman sitting beneath the shade on the bank of a stream singing this ancient song as she, too, tries to weave the fine mat which may be kept in memory of the maker long after her earthly pilgrimage had ended. The red mat in the song is supposed to be covered with the feathers of the Fijian CaUiptilns solitarius^ Latham, feathers greatly valued but not found on Samoa. The father of Sina maa aoee was the rain god, hence his sending of the lesser gods to pour the sudden shower which caused Tooalo to hide his precious cloak from the rain. f 46 Mai and Basket Weaving. Mat Sails.— In the Diredlor's Report for 1899 (p. 25) is an essay on the Mat Sails of the Pacific, i by Mr. John F. G. Stokes, now Curator of Polynesian Eth- nology in the Mu- seum. This is a very complete and accu- rate compendium of ^vhat is known of the ancient sails, and as these were generally made of paudanus mat by a method of weaving differing from that used in the construction of the broad sleeping mats, we may quote here that portion re- lating to this work: "The sails through- out Micronesia were alwaj'S made in strips var3nng in width from four inches to three feet, the Micro- nesians being par- ticularl}- apt in this form of mat-making. The Marshall Isl- anders, who are among the most ex- pert canoe builders and sailors in the Pacific, use a lap- fig. 50. i.aphoard and strip of mat sail. board, cut from breadfruit wood, on which the mat is woven (Fig. 50). The board is arched and sets very comfortably in the lap of a person sitting on the ground. The I Mat Sails nf Pandainis. 47 strips of matting as woven are passed from the board and neatly rolled np. The strip of mat has four dark strands of dyed hibiscus fibre woven in on top of the usual strands of pandanus ; this is a favorite method of ornamentation among the Marshall Islanders. The weaving commenced on the left side, and the strands were cut to about twent}- inches in length, being long enough to pass around the three strands of pandanus used to form the border at the right and reach the left edge again, where after being woven in about half an inch, they were trimmed off. It might be noticed that at regular intervals along the left-hand border some strands were allowed to protrude ; at this edge, as stated, the fresh strands were applied, and when secured four ends out of every seven were trimmed off, the three remaining butts being left to guide the weaver in inserting the black ornamental strands. This strip is 4.7 in. wide, while the breadth of the strands varies from 3/32 to 1/8. "Having woven a great length of sail mat the strips were placed together with edges overlapping and sewed with a thread made from coconut fibre or twisted pan- danus, the edges of the strips, on the edges of the sail being turned under and double sewed with the coconut fibre, which material is also used to bend the sail to the spars. A sail made in this fashion is verv stroncr and will stand a ercat strain. It is about twice as heav}- as an ordinary mat, and little heavier than canvas, and if wet becomes dangerous to use if suspended from the mast. The Micronesians, in a rain storm, prefer to lower the sail and roll it up in an envelope of pandanus or banana leaves which they generally carry for that purpose. The Hawaiian sail was made in strips, but that of Tahiti seems to have been composed of several large square mats sewn together, and could not have been a very strong combination. In all cases the work of weaving devolved on the women, while the men attended to the sewing and shaping of the sail." "' Narrow strips of matting like those made for sails were woven by the Samoan women for the game of Lajoga (pron. lafonga), in which it served as a sort of alley. These mats were about 7 in. wide and 17 ft. long. Two of these mats and a complete apparatus for playing this interesting game of skill are in the Bishop Museum, the latter the gift of Lieutenant Edw. E. Goodhue, U. S. N. Satchels of Pandanus. — A form of pandanus weaving from Micronesia and Fiji must now be noticed, of which tliere are examples in most large museums. Four from the Gilbert group are shown in Plate II, Nos. 7096, 7095, 7094 and 3349, and there it may be seen that they are formed like the mats previously described from Ebon, of broad leaves which are split in j^laces to receive the small black strips which serve both as ornament and binding. I do not know the exact use of these satchels, and the carr3'ing capacity of those '•'Occasional Papers, vol. i, no. 2, p. 26. 48 Mat and Basket Wra^'ing. from the Gilbert Ids. is very small. They come to tlie Museum labelled "Basket for books," and if this is their only use their origin must be recent, for it is not many decades that these islanders have had any books to put into them. On the other hand those of similar shape from Fiji are certainly ancient, some specimens in the old Marine Museum at Salem, Massachusetts (now Peabody Academy of Science) dating from the early years of the last century when the cannibals of that group had no books. While all the satchels in this Museum from the Gilbert group are made of broad leaves sewed together and not interwoven, the embroidery of black or red strips Fig. 51. PANDANUS S.^TCHEL, FIJI. being inserted by splitting the leaves as needed and after the basket is put together, those from Fiji are genuine mat work with broad strips, the embroidery being by the same method in both cases. In the Fijian satchel. No. S198 (Fig. 51 ), the small strips inserted obliterate the irregularities left by the alternate raised and depressed squares in the weave of the satchel and leave a uniform surface. This is especially noticeable in the band along the upper edge. In the Fijian satchel the bottom and edges are continuous and not sewed as are the bottoms of those from the Gilbert Ids. There is a fine series of the Fijian satchels in the Salem Museum, better, I believe, than else- where. The dimensions of the.se satchels in the Bishop Museum vary considerably and are as follows : Mats for Clothi)ig. 49 8349. 8.5X7 in.; coarse; upper border of finer mat work. Gilbert Ids. 7094. 6.7X6 in.; a continuons band sewed at bottom, and bound at top; red and black strips. Gilbert Ids. 7095. 12.5X6.5 in.; a double thick folio sewed only at bottom; has two bands of embroider}'. Gilbert Ids. 7096. 11.2X9-2 in.; continuous band, sewed at bottom; two bands embroidered. G. I. S198. 34X14 in.; woven with inch strips; inside plain; outside mostly covered with embroidery. Fijian. Fig. 52. MARSHALL ISLANDS MAT. The Fijian specimens at Salem vary in length from 32 to 17 inches. lu st^'le of decoration they are well represented by the specimen in this Museum, already figured. Mats for Clothing. — Although the Hawaiians certainly used mats for cloth- ing, not onlj' as wraps, but as waist cloth (malo), the principal development of clothing made of mat work, that is woven of grass or leaves instead of finer fibre, was on the Marshall Islands. The beautiful work of these islanders is familiar in all museums, and much has been written about their manufadlure and use. I do not here intend to treat these mats from the point of the designer, for the various bands shown in the figures here given have names and appropriate uses. For all these I must refer the student Memoirs B. P. B. Museum, Vol. II, No. 3217 :i2L'5 3227 :12:!2 322(1 3224 323^ 50 Mat and Basket Weaz'iiig. to the most complete account known to me, that of Prof. Dr. Augustin Kramer;"^ but for the technic of the mat mak- ing I shall take the liberty of translating from this in- teresting paper. As the or- thography- of the names of the mats and their parts seems wholly nnsettled, even German writers not agreeing among them- selves, we may pass over the native names and their etymology. The illustra- tions given here of these mats are all from the Bishop Museum collecT;ion, except Plate V, which is made from a large mat long used as a table cover by the writer, who has thus had an oppor- tunity to test the durability of these admirable mats. Nearly all here figured differ more or less from those fig- ured b\- Dr. Kramer, and many of them are of con- siderable age. I translate freely from Dr. Kramer, omitting much of the philological matter as, however interesting and valuable, foreign to our present purpose. "Among the produc- •* FIG- 53- GKUUP OF MAKSIIALI^ ISLANDS MATS. tions of the Marshall Islanders doubtless the mats used for clothing take the first place The.se were made freehand without apparatus or loom, only a long pointed "Die Ornamentik der KleidmaUcii und der Tatauieruiig auf den Marshallinseln iiebst technologischen, philolo- gischeti und ethnologischen Notizen. .A.rchiv fiir Anthropologie, Neue Folge, Band II, Heft i. Braunschweig, 1904. •iMliii Technic of Marshall Islands Mats. 51 weaving needle, called oca^ generally from the long wing bone of the albatross, serves to lift lip the strands which are woven on a lapboard of breadfruit wood one to two feet long and six to twelve inches wide, called digcuat. \ A similar one is figured in the se(5lion on ]\Iat Sails, p. 46.] For the embroidery a needle of human, dog or fish bone; for trimming the leaves a pinna shell, djahor. All these things are kept in a basket woven from coco leaves called bodjo. For the mat the leaves of Pandanus odoi-alissi- iinis are used. For the fine mats these are plucked young and roasted over a fire ; for sails or coarse mats the older leaves are seledled and left to dry in the sun, and then beaten with a hard mallet, draggeiiiia^ of Tn'daciia shell [see No. 7832, Fig. 54]; the beaten leaves are rolled into bundles [like those shown in Fig. 38] and laid aside for occasion. "When the weaving begins, leaves are taken four inches wide and three to seven feet long, and slit into strands which have various names according to their width, as ajclai\ three fingers wide, ajomcii thumb-breadth, d/clcrik still thinner (3/16-5/16), and nia the finest (1/16-3/32). The red fibres are from a creeper, Ficaria raiiticiiloidcs^ FIG. 54. MAi.i.KTs FUR HKATiNG PANDANUS i.KAVKs. callcd adad. The bark of the thinner shoots is stripped off, scraped and freed from the epidermis and dried ; the bast then assumes a brown or reddish-yellow hue. The black fibres here, as on other islands, are from the bast of Hibiscus dyed with soot or mud. According to Finsch the black dye is obtained from the fruit of the mangrove. These are grated and boiled in marine shells or coconut shells and the bast is soaked in the decoAion until the desired tint is obtained. For a red color the same strips are passed through a dye made of the root bark of a certain tree {Morinda citrifolia ?) with the addition of lime." After describing the sleeping mats, which were arranged much as in Hawaii, the coarser Aveave at the bottom and the finer at the top of a pile often four to six inches thick, the bottom ones being sometimes of coco leaf, Dr. Kramer describes very fully the fine clothing mats, of which class we have illustrations before us. (Figs. 52-3.) It is unnecessary to go into the analysis of the ornament with the Dodor, but his ex- 52 Mat and Basket Weaving. planation of the way of wearing these rather uncouth garments, which are fast giving place to the foreign cloth loose gowns, ma)- well be noticed. There are two distinct ways of wearing them : either two mats, one in front, the other behind, are worn as aprons and fastened around the waist with a coil of the interesting coco and pandanus cord already described, a coil that is often sixty feet long (and Finsch reports one fifty metres long), or a single mat is passed between the legs and the corners brought over a similar girdle forming a sort of malo, an arrangement that would seem to pre- clude any rapid running by the wearer. The thickness of the cord girdle increased with the rank of the wearer. There were also in these mats (called //') patterns suited for men, others for women, some for commoners, others for chiefs, all of which Dr. Kramer fully describes. These mats were often given as presents, and in the early sixties some of the Marshall Islanders made a large one suitable for a bed cover which they sent as a present to Kameha- meha V in recognition of the good done for them by the missionaries from his kingdom of Hawaii. On the vo3^age from Micronesia to Honolulu rats destroyed a corner of the elaborate border, and the partial ruin was purchased b}- the writer and is now in the Peabod}' Museum at Cambridge, Mass. When the use of mats for gar- ments ceases, as it will very soon, it is questionable whether the}- will still be woven. For table covers they are well suited, as well as for other purposes of domestic orna- ment or use, and it would be a pity that the}- should be relegated to the class of lost arts. It cannot be denied that mat weaving here, as elsewhere in the Pacific, is rapidly declining, owing largely to the unwise method of instrudlion that has been in vogue throughout the Pacific where the methods of Anglo-Saxon education have been forced upon peoples generally unable to assimilate such intelledlual food. If the white men had simply endeavored to make better specimens of the various races, saving what was good in their work, gently eradicating the heathen tendencies, it seems possible that many of the useful industries of the Pacific might have been saved from the list of abandoned arts. The pandanus and makaloa might still be fashioned into fabrics which would bring revenue to the makers and credit to the islands of the Great Ocean. Hawaiian I^auhala Mats.— Enough has already been described of the meth- ods of other Pacific islanders perhaps to render superfluous any minute description of the ways and means of the old Hawaiians in lauhala mat making. Doubtless it was the oldest form of mat making on these islands, and fragments are found in the most ancient burial caves quite like the mat work of the present day. No such work as that just described from Micronesia was made here, perhaps because the use of kapa for garments rendered their use for clothing unnecessary; but it must not be supposed that the old Hawaiians did not make fine pandanus mats, for there are suffi- Legendary Origin of Haiuaiian Mats. 53 cient witnesses to the fact in tlie cases of the Bishop Mnsenm, where most beautifull}- regular and finel}^ woven mats are in considerable number. As to the legendary origin of Hawaiian mats I can give no information. Not only have the modern natives preserved no remembrance of the songs or legends of any such theme, but unfortunately the present writer negledled to take down from the lips of the old women who, in the middle of the last century, were singing such songs as they plaited the mats in the Puna caves, songs which undoubtedly related to their earl}' predecessors in the mat making work ; he also neglec^ted to ask what were the tutelar gods of the craft, and today the remaining natives are unable or unwilling to give the ungathered information. We only know that the Hawaiians had among "the forty thousand and four hundred thousand gods" one or more especiallj' worshipped by the pious artisans in basketry of former days. We also know that they had songs and legends of earl}' mat or basket makers who were famous at their craft, and were per- haps apotheosized by their successors for their skill. The teaching of a new religion seems to have driven from their thoughts many things worth remembering while by no means strangling the superstitions of the neophytes. Their love for the recital of w^/r, which was quite oriental and kept alive in their memories the doings of their forefathers, their genealogies, and the rules of life, has largely disappeared from this group, and aimless political discussion or modern tales of a very different nature have usurped its place. Leaving then the folklore and poetr}' of the basket and mat makers untold, we must present their work, which is so like that of all the other Polynesian groups that it hardly deserves an extended paragraph. The hala leaves were gathered, they were dried either in the sun or in the cover of the house, were scraped, trimmed and beaten, then rolled and stored until enough and more had accumulated for the intended mat. The shell trimmer was that used everywhere, but the mallet was not like those figured from the Marshall Islands (Fig.54 ), but either a plain round club (ho/ioa)^ or more commonl}- an old kapa beater {ie /&?^/'«), of which the engraved sides had become smooth from long use. These four-sided clubs with rounded handles were common all over the group, as they were necessary iitensils in every family, and they are to be found in most ethnological museums. I have asked a skilful maker of lauhala mats, still resident in the once famous mat making district of Puna, to make for me the first few inches of a common mat, and the result is shown in Fig. 55. To show her good will as well as skill she has embroidered her own name and address on the obverse of the figure, and my name on the reverse ; but in spite of that fanc}- the specimen shows perfectly well the be- ginning of the weaving. No vestige of a loom, not even pegs to hold the row of leaves in place is needed, but a sufficient number of hala leaves is taken to cover the width of 54^ Mat and Basket Weaving. mat desired, and these are strongl}- sewn together as shown in the figure. The split- ting into strands of the sele(fled width is then done and the plaiting begins. Each strand passes diagonall}' to the edge of the growing mat and then turns and continues its journey at right angles to its former course imtil it reaches the other side, or is continued^by a succeeding strand which is neatly inserted under the overlapping trans- verse strand. The skill with which the weaver keeps the long strands from hopeless entanglement and puts each in its destined place is surprising to the uninitiated. To ^' j>-.-f-:*. ■"''">; Fir,. 55. THE BECIXN-INCr OF A MAT OF I,AUHAI,A. the success of the mat the weaver's fingers must put the right tension on each strand or the mat will be bellied or warped, and will not lie flat or be redlangular when finished. An inexperienced or careless mat maker is alwaj-s known by the irregular mat. While the whole leaved edging is sometimes left on the finished mat, it is usuall}' cut off when the distal end has been bound in. The size of these mats was unlimited, except by the use to which they were destined. Often they extended quite across the house, and when intended for covering canoes might be eighty to a hundred feet long. Long and narrow ones were used for a table at an ahaaina or feast ; a moderate sized one in the Bishop Museum, made for this purpose, measures 3 ft. in width and is 28.7 ft. long (No. 25S3). It was not un- common to weave in coarser mesh or strand immense mats to spread on the ground for I Neiv Hcbridcan Dresses. 55 seating a large company, and I have seen in old native churches mats many j-ards square covering the whole floor, there being no pews or seats of other nature. From their size and weight they were seldom removed, and often became unsanitar^•, accord- ing to modern ideas. The bed mats have been several times referred to, but the\' ma}- be more fullj' described here as a series of mats woven, in the best houses, to fit an allotted place, and arranged in accordance with their fineness from the coarsest, which rested on the gravel floor of the house, to the fine mat on top that showed the wealth or taste of the owner. To keep the mats in place (and as the whole family slept on the same bed, and some of the bedfellows might be uneasy from overfeeding, this was no simple matter) mats of the hikiec were sewed together along one edge, and this edge generally raised by the interposition of strips of the same matting. Matting of the lauhala in coarse weave, one inch strands or larger, were in con- stant use to cover propert\- from the sun or sudden showers, to spread nuts or herbs iipon while drj-ing in the sun, and to wrestle on in an indigenous form of that maul}- exercise, where the contestants clasped hands and, without touching any other part of the body, endeavored each to push the other off the mat. In the earl^- days of the American Mission on these islands the simple homes of the missionaries were generally carpeted, if at all, with lavihala mats woven to fit the room, and examples are still ex- tant of mats of great beauty given b}' early converts to their respe(5led teachers. New Hebridean Dresses. — On several of the islands of the New Hebrides, a group using the loom, and famous for fibre weaving, are found dresses of finely cut pandanus, so closely resembling grass work, that until the material was considerably magnified the author was inclined to class them with the makaloa mats of the Hawaiians. As will be seen in Fig. 56, they are aprons of no generous size (some are hardly two inches wide), but the peculiarity' of their openwork weave recalls the Tongan mats of the opposite side of the Pacific. These aprons are worn by the women by means of some sort of belt, and on some of the islands are stained a magenta red which does not add to their beauty in the eyes of a foreigner, and almost conceals the openwork pat- terns. Some of those from Malekula have been washed until the pandanus fibre is broken and roughened. The following list of women's dresses from Malekula, Oba and Ambrym shows the difference in size: 8143. 3.2 ft. X 10 in. — 16 strands to inch. Malekula, twilled weave, no coloring; Fig. 56, No. 3. 8144. 3 ft.X9 in. — 15. Malekula, plain weave; Fig. 56, No. i. 8145. 3 ft.Xii in. — 14. Malekula, twilled; Fig. 56, No. 2. 56 Mat a)/d Basket IVcavnie^. 8447. 3 ft. long X 9 in. wide — 20 strands to inch. Oba, with various openwork de- signs at ends, and stamped with magenta stain. 844S. 3.2 ft. longXg-S in. wide — 20. Oba, like last, short fringe on edges, long on ends. 8449. 1.5 ft. long X 3 in. wide — 26. Ambrym, broad red stripes, openw-ork ends, with long ( 10 in.) fringe. 8459- 3.7 ft. longXi.7 in. wide — 21. Ambrym, all red, two sewed together in the middle, ending in four braided tails and long fringes. Fig. 56. WOMEN'S DRESSES FROM MALF:KULA, N. H. 6627. 3.5 ft. longXio.5 in. wide — 15. Oba, openwork ends, stamped purple all over^ fringes. 6628. 3.6 ft. longXii.5 in. wide — 16. Oba, same treatment as last. Fijian Coffin. — This remarkable specimen of basket work is 22.5 in. long, 9 in. wide in the middle, 5.5 in. at the ends, and 9 in. deep. The bottom is wholly of breadfruit wood, and the cover is lined with the same material stitched with pandanus strands to the upper edge. The interior is lined with a plain broad weave of pandanus extending above the external rim by a foundation of two splints bound together by pandanus strips; this forms a hold for the cover. The illustration (Fig. 57) shows Basket from Fiji. 57 what is really the distinftion of the whole work, the beautiful design covering the ex- terior. Whether this cofifin-shaped basket was really intended for the use its name would imply may be doubtful ; certainlj^ in this connedtion it is unimportant ; the re- markably' effedlive design of the basket work is all in all. The technic is simple, a vertical series of small black strips of uncertain origin form the more conspicuous trellis, over and under which pass portions of the close band of light brown grass-like ?IG. 57. FIJIAN COFFIN-SHAPED BASKKT. strips forming the design. Between two bands of zigzag is a much broader band of diaper work ; the tipper rim, which serves also as support for the cover, is a braid of black rattan strips. The design of the cover is not so successful. The vertical por- tion repeats the zigzag band of the basket proper, but the top is covered with a longi- tudinal band of diaper with two zigzag narrower bands on either side, with partings of subsidiary bands of a dark brown weave. Whatever the purpose of this basket we may be sure it was intended to hold something very precious, and I must doubt, in view of the light esteem in which infants were held in ancient times throughout the Pacific, that it was intended to hold the decaying remains of any baby, even the offspring of the highest chiefess. 58 Mat and Basket Wcaviiijr. List of Pandanus Mats in the Bishop Museum. 2580. 7.2 ft.X4.7 ft- — ^6 strands to inch. Very fine and flat weave. Hawaiian. 2581. 34.3 ft. X 10.5 ft. Hawaiian. 2582. 10 ft. X 7-2 ft. Hawaiian. 2583. 25.7 ft.X3 ft. Dining mat. Hawaiian. 2586. 6 ft. X 4-7 ft. With diagonal stripes. Hawaiian. 2591. 18.2 ft.X8 ft. — 16-12 to in. Very flat weave. Hawaiian. 2598. II. 5 ft. X7 ft. — 7-8. Hawaiian. 2599. 10.7 ft.XS ft. — 6. Hawaiian. 2602. 8.3 ft.XS ft. — 7. Hinano = young leaves, Pnna. Hawaiian. 2603. 9 ft. X 7-7 ft. — 3. Hawaiian. 2780. 10 ft.X6.5 ft. — 4. Checkered. Rotuma. 2781. 10.5 ft.X6.5 ft. — 4. Diagonal brown stripes. Rotuma. 2782. 10.5 ft.X8.5 ft. Hawaiian. 2783. 10 ft.X9 ft. — 4. Hawaiian. 2784. 11.3 ft.XS ft. Old. Hawaiian. 27S5. 10 ft.X9 ft. — 7. Hawaiian. 2786. 12.2 ft. X 7.5 ft. — 6-7. Flat weave. Hawaiian. 2787. 6 ft.X4 ft. — 5. Brown in patterns. Hawaiian. 2785. 5.5 ft.X4.7 ^^- — 5~6. Diagonal brown stripes. Hawaiian. 2789. 9 ft. X 6.3 ft. — 12. Hawaiian. 2791. 8.5 ft.Xs ft. — 15. From Queen Emma colleAion. Hawaiian. 3239. 5.6 ft. X 3-8 ft. — 6. Black and white checks and twills. Nine. 3240. 6.2 ft.X5-i ft. — 4.5. Black and red zigzags. Nine. 3498. 9.5 ft. X 5-2 ft. Leaves of two colors. Gilbert Ids. 3499. 4.8 ft.X4.2 ft. Gilbert Ids. 3500. 6 ft. X 4-5 ft. Two colors with long suture. Gilbert Ids. 3501. II. 7 ft.X7 in. — 9. Fringed belt, black stripe. Samoan Ids. 3505. 6 ft.X5-7 ft. — 13. Fringed. Hawaiian. 3506. 8 ft.X5 ft. Hawaiian. 6310. 14.5 ft. X 13.5 ft. — 2. Embroidered with red. Tongareva. 63 1 1. 8.3 ft. X 6.5 ft. Strips of leaf sharply cut. Tahiti. 6629. 10 ft.XS ft. — 6. Mrs. Bishop's colleAion. Hawaiian. 6630. 6 ft. X 3-6 ft. — 6. — Mrs. Bishop's colledlion. Red border. Nine. 6631. 8 ft.X7 ft. — 10. Mrs. Bishop's colle(ftion. Single leaf. Hawaiian. 6632. 6.2 ft.X5.9 ft. — 10. Mrs. Bishop's collection. Embroidered black and brown. Marshall Ids. 6655. 5.8 ft.X4.5 ft. — 16. Irregular serrate edge, fringed leaves on ends. Samoa. 6657. 7.4 ft.X4-6 ft. — 12. Double black transverse stripes. Gilbert Ids. 6659. 4 ft.X2 ft. — 10. Fringe on all sides, embroidered with wool. Tonga. 6660. 5.2 ft.X4 ft.— 6. Nine. List of Paitdamis Mais. 6661. 4.5 ft. X 2.6 ft. — 5 strands to iuch. Friuged ou surface also. 6662. 5.7 ft.Xi.2 ft. — 14. Twilled belt, open edging on ends. Santa Cruz. 7718. 1 1.5 ft.X9.5 ft. Hawaiian. 7755. 2.2 ft. X 1-8 ft. — 4. Samples made in 1834. Hawaiian. 7962. i3ft.Xioft. Hawaiian. S061. II. I ft. X9 ft.— 5. Marshall Ids. 8444. 6.5 ft. X 2.5 ft. — 9. Fringed with feathers. Tonga, N. H. 8445. 10.5 ft.X22 in. — 9. Fringed on long edge. Tonga, N. H. 7729. 9.8 ft.X7.5 ft. Hawaiian. 59 These 3215 3217 3218 3219 3220 3221 3222 3223 3224 3225 3226 3227 3228 3229 3230 3231 3232 3233 3234 3235 3236 3237 323S 6656 6663 6664 7563 Pandanus Mats from the Marshall Islands. re all embroidered around Ihe border with hibiscus fibre, black and red-brown; the smaller sizes are used as aprons. 2.2 ft.X2.i ft. — II strands to incli. With hau trimmings. Maloelab, 1839. 6 ft.X6 ft.— II. 6 ft.X5.5 ft. Narrow border. 6 ft.X6 ft. — II. Remarkably wide border, mostl}- black. 5.5 ft.X5-5 ft. — II. Wide border, mostly black. 6 ft.X5-5 ft. — II- Narrow border. 5 ft.X6 ft.— 12. 6 ft.X6 ft. 3 ft.X3 ft. — II. Wide border, mostly red-brown. 3 ft.X3 ft. — 9.5. Wide border. Fig. 53. 2.7 ft.X2.7 ft. — 10. Wide border. Majuro. Fig. 53. 2.7 ft. X 2.7 ft. — II. Narrow border not extending to the edge. Fig. 53. 2.7 ft. X 2.6 ft. — 9. Wide border. 2.9 ft.X2.9 ft. 2.7 ft.X2.7 ft. 2.7 ft. X 2.7 ft. — 10. Narrow border, mostlj- black. Fig. 53. 2.7 ft. X 2.5 ft. — 10. Narrow border. Fig. 53. 2.5 ft. X 2.5 ft. — 10. Narrow border. 2.7 ft.X2.2 ft. — II. 2.2 ft. X 2.2 ft. — 10. Fig. 53. Only one with diagonal weave. 2 ft.X2 ft. 2 ft.X2 ft. 5.3 ft. X 5-3 ft. Medium border. 6 ft.X5.9 ft.— 12. 12.7 in.X7.7 in. — 12. Satchel, embroidered like the mat border. II in.X8.7 in. — 12. Satchel, embroidered like the mat border. 3 ft.X3 ft. — 9. Mat with wide border. 6o Mat and Basket Weaznng. Baskets of leie. — We are now to consider one of the two most distindlive Hawaiian works in basketry, and one which, for solidity and durability of workman- ship, I believe to be unsurpassed in this Pacific region, nor do my limited studies in the field of basketry embrace any baskets of any region which may justly be accorded a higher rank than pertains to the baskets made from ieie b}' the Hawaiians. Unlike Fig. 58. freycinetia leaves. Fig. 59. FREYCINETIA ROOTI.ETS. the pandanus, its near relative, the Freycinetia does not offer in its leaves its most useful product, although these are similar in shape but smaller in size than those of the pandanus. It is a slim, scandent plant, either making dense jungle with its partly recumbent stems, or climbing high among the trees where its tufts of green leaves enclosing in the flowering season the bright red flowering leaflets, among which the three flower heads rise, always attract the eye of an observant traveler. From the slender stems perpend long rootlets of almost uniform diameter which cling to the Hawaiian Tcie Baskets. 6i trees or swing freely in the air. Fig-. 59. Slight and flimsy as these rootlets appear the}' are the all-important product of the plant. Hanau ka ieie hihi i ka nahck-. Born is the tangled ieie in the forest. is a common saying of the old Hawaiians, for the plant is found in abundance in the forests, especially the more elevated ones, all over the group. It is also found on other Fig. 60. BA.SKET OF IEIE. SEE PL.4TE VII FOR COVER. Pacific islands, the name ie, or ieie, extending through Polynesia. The Maori ikiike, or kiekie {^Freycinetia baiiksii^ is the same word. In the days of idol worship the ieie was used for decorating the idols, and at the present da}- the word is applied to a person decorated with lei or wreaths. For general rustic decoration it is also very fit, as it does not soon fade nor drop its leaves. Ornamental it will always be, but its usefulness seems to have departed from these Hawaiian Islands, for no longer are the baskets made that attraAed, alone among the 62 Mat and Basket JVeai'ing. basketrj- of the Pacific, the wonder of early visitors. There may still be a very few specimens stored in the closets of some old residents, but I know of none in any museum except the two complete with one coverless specimen in this Museum, and I shall therefore both describe and figure these most carefully. There is little to say about the preparation of the rootlets : they were sometimes split in halves and sometimes used whole. The structure is rather peculiar and serves to distinguish this material from the manj- vines of the Convolvulus family that have a somewhat similar appearance. In the centre are five or six tubes which give elas- ticity' to the strips ; the ferns having in seAion a harder tissue of horseshoe shape. In illustrating the finest basket of this material that the Museum possesses we find diiificulty in distinguishing the two colors which were once very distinct, since age has so browned the light color of the natural stem, and so faded the deep purplish black of the contrasting portion, that while the eye with difficulty makes the distinc- tion, the photographic plate has failed to do so, except in the middle of the second band from the bottom. Fig. 60 shows the body of the basket, and Plate VII the cover, the basket belonging to what we may for convenience call Class I. The full descrip- tion is rather dull reading to an^- but an expert in basketr^', but is worth giving, there are so few specimens left, and time must destroj- these. No. 7651. Diameter 26 in., height 17 in. Specimen in good condition, except that the cover is cracked all along the upper rim, and several of the side cover loops are gone. Starting from the rim, which is of course reversing the order of manufac- ture, this rim is flat, consisting of two rods carefully clothed with flat strip, every fifth and sixth one going down under two horizontal twisted rounds in which the strips pass over two of the upright strips. Then a band of dark and light strips, originally light brown and black, 16 rounds of two-ply over uprights of 3 rods each; 2 rounds twisted, followed by 12 plain brown; 2 twist, variegated band of 17 rounds; 2 black twists and I brown, 16 brown and 2 twists; sides then turn in to form bottom with a variegated batid of 16 rounds, i brown, i black and i brown twist; 11 variegated rounds, 7 brown rounds, 3 brown twists; 15 brown rounds, 2 twists; 10 brown rounds followed by 7 twists to centre of start. There are 19 braided sennit loops to attach cover. Cover is formed with the same kind of rim, 2 twists, 2 black, 7 brown rounds, i twist; 19 varie- gated, 2 twist, turning edge; iS brown rounds, 2 twist; 25 variegated rounds, 12 twists; 8 brown and 2 black twists; 12 twists to centre. This basket, supposed to have come from the island of Hawaii, was many years in the cabinet of the A. B. C. F. Missions in Boston. No. 65S9. This basket is plain brown without color decoration, 17 in. in diame- ter and 10 in. high. The rim is similar to the last, but has three bars as a foundation instead of two. The succession from the rim is as follows : 2 twists, 10 rounds ; Hawaiian Icic Baskets. 63 6589 FIG. 61. JCDCH OF BASKET NO. 6589. 2 twists, 19 rounds; 2 twists, 17 rounds; 2 twists, 17 rounds on the bilge; 2 twists, iS rounds; 2 twists, 12 rounds; 7 twists to centre. The cover rim has one rod and the binding strip attaches it to the two succeeding twists, by the sixth and seventh round some go over both twists, others over only one. Then come 12 rounds, 2 twists on the edge of cover; 12 rounds, 2 twists; 14 rounds, 2 twists; 11 rounds, 2 twists; 6 rounds, 4 twists to centre. The label which came with this speci- men reads as follows: "This was originally the property of Mrs. Coan, of Hilo, Hawaii, but it has been in use in the family of Mrs. S. N. Castle, of Honolulu, for over fifty-five years, and is given b}- her to Mr. Brigham for the Bishop Museum." No. 6942. Diameter at bilge iS in., at rim 12 in., height (without cover) 9.5 in. The rim is composed of five small rods bound and attached in a way similar to those in the two preceeding specimens, but with less regularity. The sequence is 2 twists, 10 rounds; 2 twists, 20 rounds; 2 twists, 15 rounds; 2 twists, 15 rounds on the bilge; 2 twists, 13 rounds; 2 twists, 15 rounds; 3 twists, 11 rounds; 8 twists to centre; no cover extant. The outside has been painted all over a lead color, but the inside is fresh and unstained. Specimen from the curious old stone house at Kailua, Hawaii, built by Reverend Asa Thurston (arrived 1820), and was used by the Thiirston family there and elsewhere for .sevent}'- five years. These three are the only surviving specimens belonging to this class of ieie baskets that are known to the writer. The length of time these have been in use is remarkable, and shows the great strength and durability of the material. The three are of the same general pattern, and might have been made by the same hand; the perfection of the work seems to show that there must have been many made before such skill could have been acquired. The next class comprises the fine baskets woven over wooden umeke or gourd calabashes or bottles, and several very fine specimens are in the Bishop Museum. Exaftly how these were made I cannot say, and the technic must be gathered from the specimens, for in the early daj-s of m^- explorations on this group I never saw an}- such baskets, and have not been able to learn much about them from the present gen- eration of natives. It seems to have been the only basket figured and noticed by the early explorers, and Freycinet makes a curious mistake in the material used. It seems worth while to quote his account : 64 Mat and Basket Weaving. "On fait avec la feuille du ti, quelques ouvrages dans le gout de ceux de nos vanniers, mais peut-etre mieux finis; les uus sont d'une seule teinte; les autres, tels que les casques, les eventails, les panniers, le clissage des calebasses oii Ton serre les effets precieux, &c., sont nuances de couleurs diverses. Ces ouvrages, confies aussi a la main delicate des femmes, offrent des formes et dimensions tres-variees.'' What the French Com- 38»o ssss mander took for the leaves of the ti (ki, Cordylinc tcrmiualis) was afterward named for him Frcyci- netia, and the figures given in Freycinet's report leave no doulat that the l:)askets we are now con_ sidering were the ones referred to in the quotation. As he saj'S, these hinai poepoe were used as receptacles for their greatest treas- ures, and we know that when the breakage of the interior unieke or calabash deprives these baskets of their waterproof character and they fall to baser uses they will still serve for many years as fish bas- kets, or for other purposes. Demi- jons and carboys, hard receptacles covered with basket work by more civilized people, are only fit for destrudlion when the inner part is damaged ! Figs. 62, 65 and Plate IX give fair rep- resentations of the best of these hinai poepoe, and the following list of those in the Museum will explain the illustrations : No. 3890 (Plate IX) measures, as figured, 21.5 in. high with cover; greatest diameter 9.2 in. The inside wooden (kou) umeke measures 13.2 in. in depth (internal), and 5.7 in. at the mouth. The cover is of the same material and has 6 in. and 6.7 in. internal measurement. The outer basketry comes to within three-quarters of an inch of the top of the umeke, and is finished ofi^ by a braid which does not project beyond the general surface; a net of olona is attached below the fourth round. The cover has a rim like the baskets of the first class, but smaller, and in both the weaving is close and varied only by reversing. "Preycinet, Voyage autour du moude, 1817-20, II, p. 613. Pig. 62. HINAI I'lii.i'Di. Hawaiian Hinai Poepoe. 65 No. 1405 (Fig. 65) measures 28 in. high and 13.5 in. at the greatest diameter. The interior is a gourd 23 in. deep and 5.2 in. at neck. The strands are three-ply in cover of which the gourd is lost. No. 3889 (PI. IX) has a wooden umeke as well shown in Fig. 63; there is no cover; height 12.5 in., while the umeke measures inside 10.5 in., with a diameter of 6.5 in. at neck. The beautiful and close weave at the neck is shown in the figure. No. 1403 (PI. IX) is 25 in. high and 16 in. in diameter : cover empty: both baskets neatly fin- ished at ends by braid. No. 1406 has no cover; is 14.5 in. high, 13.5 in. in diameter, and con- tains a gourd; closely woven around the rim. No. 1550 (Fig. 65) is remarkabl}' long and originally covered a wooden umeke which was broken many years ago, and for more than a score of years the basket has been used as a receptacle for fish. It is 36 in. long, — the longest known. The base shows an unusual FIG. 63. MOUTH OF HINAI POEPOE NO. 3889. beginning f Or thcse bas- kets ( Fig. 64). The variations in the bands are shown in PI. VI and, in part, in Fig. 65. No. 3890 (PI. IX) is woven around wooden umeke, in body and cover; height 2 2 in., diameter 9 in. Best in collection. No. 1409 (PI. VIII) is one of the flat calabash baskets, 12 in. high and 18 in. in diameter. As this is the only one of this flat form and has no cover, it is difficult to say what that part consisted of, but it was probabl}' a flat basket work disk, and the basket was used for the preservation of some choice feather cape or lei. No. 7705 has a diameter of 16 in. and is 27.5 in. high. It is woven on a gourd, but the cover is now empty and the whole is contained in a koko or net used for such purposes. Memoirs B. P. B. Museum. Vol. II, No. i.— 5. 66 Mat and Basket Wcaz'iiip-. No. 3844 has lost its internal unieke or gourd. The weaving is much like that in Class I, within which it would be included but for its shape. Height 24 in., diame- ter 14 in. It has heavy cord loops to secure the cover. No. 1404 has a height of 16.5 in., and a diameter of 9.5 in.; gourds in basket and in cover; the general finish very good, the weave being in a continuous spiral from beginning to the rim. There are six black bands on the basket and five on the cover; the net for the cover is of olona (PI. IX). In all this class of ba.skets the work is far above the average of Hawaiian, or indeed Pacific, basket making; the weave is exceedingly close and regular, and the form is peculiar to this group, but most closely resembling certain Hawaiian containers made for holding fish lines, but of wood and gourd without any basketry. We now come to another class where the workmanship is far inferior, although the gen- eral form still reminds us of the fine hinai poepoe. The material, ieie, is the same, but the strips are either used unsplit or carelessly combined. In .some an attempt to imitate the older and better work seems present; in others the product is a coarser basket without any enclosed container. The basket becomes suitable for holding fish or any coarse or dirty objec^ts without having to wait until the inner gourd or umeke is broken. The list is not a long one. No. 1407 is a very modern attempt to imitate the fine work of the old Hawaiians, and, apparently has never been used. Two gourds are emploj'ed for foundation, and the basketry is of the coarsest two-ply weave over single rods of the same material ; it has much the same appearance as the common demijohn wicker work. Height, with cover, 19 in., diameter 13 in.; net of coco fibre cord. The ieie strips have not been halved; they have only a small portion removed, and the strip is of course less pliable. No. 4050 is a modern basket finished mitch as the last, but independent of gourd or umeke, and with a flat top. Height, with cover, 12 in., diameter 10.5 in. Fig. 66. No. 1408 is of the same general form as the preceding, but is of greater age. It has apparently been woven without any interior vessel, and with Nos. 4050 and 3S45 might be put in Class III for convenience of classification. The workmanship is by no means so fine as in Class I, but is better than in No. 4050. Height, without cover, 13 in., diameter 11 in. The rims of both basket and cover are made exac^tly like that figured in No. 6589, and the latter is secured by a single coco fibre cord ex- 15 50. FIG. 64. BOTTOJI OF BASKin' NO. I55O. Haivaiiau Hinai Pocpoe. 67 t. V >rf-v j^^^MiY'^'" l.i.-iO jfjG, 5j, HAWAIIAN HINAI I'OEPOE. H"" tending from its centre to the encircling net of the basket, this net being made partly of olona, partly of coco fibre. Central specimen in Fig. 66. No. 3845 is like the last specimen, as may be seen in Fig. 66, but is much smaller, and in place of the usual net to fasten the cover in place, has two loops of cord knotted through the sides, which serve also for handles. 68 Mat and Basket Wea-ving. No. 3887 has no cover and was made as an independent basket ; the rim is neatly braided and the rounds are three-ply. The rods, added as the diameter increases, are not always spliced to those already in place, but are interpolated like a V, making two rods for each insertion. Height 17 in., diameter 14 in. Has been mended in places with a "darn" of coco fibre cord. I 1 Fig. 66. group of flat-topped Hawaiian baskets. No. 3843 is similar to the last, but of coarser workmanship; the rounds are still three-ph', and the rims both of basket and cover are of a number of strips intertwined around the bent ends of the rods, as shown in Fig. 67. Height 17 in., diameter 12 in. Not onU^ umeke and correspond- ing gourds were covered with basketry by the old Hawaiians, but also the huewai or gourd water bottles, prob- ably in all cases for additional strength; but with the huewai it also helped to keep the contents cooler. When I first saw the following specimen I was in- clined to think the natives in modern times had copied the wicker-covered glass bottles used as perfume bottles or pocket flasks by the foreigners, but I am assured by old Hawaiians that their people made such water bottles in ancient times. No. 5350 is a huewai or water bottle of gourd covered with a closely-woven FIG. 67. 3 S +3. RIM OF BASKET NO. 3843. Hazvaiiaii Fish Baskets. 69 three-ply spiral band, alternately brown and black. There are two braided ears on opposite sides of the spring of the neck, to which are still attached the remains of a square braid cord ; the cap or capsnle is of the same work. We come now to an entirely different kind of basket, both in shape and in pnrpose, althongh there is, it is true, some likeness in shape to the beautiful baskets of Class I. Its shape and weave is sufficiently shown in the figures to dispense with an}- particular description. A people depending, as did the Hawaiians, on fish for food would naturally have developed their basketry in that line, and we shall briefly touch upon the com- mon forms, leaving a fuller description of their adlual use for the chapter on Hawaiian Fisheries. Man}- of the baskets that have been described have FIG. 68. HAWAIIAN FISH BASKiiT, NO. 3842. bcen uscd for fish when no longer fit for their original purpose, but the specimen here figured was made for fish and not for anything else. No. 3842 is a fish basket with a height of 17 in., with cover, and a diameter of 19 in. The weave is three-ply on the basket, and two-pl}- on the cover, and there are no handles nor fastenings for the cover. See Figs. 68 and 69. ^ These baskets were common and seem better suited to their purpose than those of woven pandanus leaves already described, and thej- are both stronger and more durable, 70 Mat and Basket Weaving. but, perhaps on account of their use, they are not often found in colleAions, as the old ones become very dirty and fishy. Hawaiian fish traps were sometimes made of ieie, sometimes of other vines, but genearlly of iwaiwa or fern stem, and liave been described under that division. The curious shrimp baskets deserve notice here, and Plate IV ma}- be consulted for their general appearance and constru(R:ion. As that plate shows, they var}- considerably in size, and in recent times the vend- ors of curiosities have introduced immense ones to serve for hall decora- tions or umbrella holders. The scale on the plate is sufficient guide to the size of those figured. Nos. 7681 and 7682 are of a more common size. It should be added that these baskets are used to catch the shrimps, and not to store them. Idols and Helmets. — Another extensive use of ieie strips and rods is fonnd in the mauufadlure of images of the gods, and helmets. Of the for- ^i^- ^9- enlarged portion of no. 3842. mer, huge simulacra were constru(?ted, none of which are now known to exist, although in the earl}' da3's of the American Mission (1820-1835) sone large specimens remained at Kailua, Hawaii, and it is quite possible others were concealed in caves which may Hawaiian Jl'ickcr Hehiiels. 71 still preserve them. The late King Kalakana described to me such images he claimed to have seen in a cave in the cliff above Kealakekua Bay. There are still in existence, however, a num- ber of smaller images, notably those of the war god Kukailimoku, some of which have been figured and described in the first volume of these Memoirs. These consisted of a head and neck of ieie basket work, often ver}' neatl}- made, which was covered with a closeh' fitting net of olona fibre, to which were attached the red, 3'ellow and black feathers which distinguished the god. The helmets of the ancient chiefs are even more widel3' distributed through the principal mu- seums of Europe and America. Some of these are made in a comparatively rough manner, as the wicker work was to be covered with feathers, while others were finished specimens of basket work. Illustrations of both classes have been given in the Memoir on Ancient Hawaiian Feather Work,"" and Figs. 70 and 71 maj' here repeat examples of them for convenience. The first. Fig. 70, which is in the Australian Museum, Sydney, N. S. W., was once covered with feathers, and is made firmly of a braided strudfure ingeniousl}^ adapted to the irregular shape of the helmet. The other. Fig. 71, was beautifull}- woven in the manner of the best baskets and was worn without the feathery decoration. Others, ap- parentl}' intended to remain featherless, are in Berlin, London, Paris and Cambridge, Mass. I have not examined all the sub- structures of the feather helmets figured {loc. cit?j^ but as all in this Museum are of the same material I infer that it was generally used for these images and helmets. WICKER WORK HKI.MET. <% FIG. 71. HEI.MET OF IEIE. Fern Stems. — Although among the Hawaiians fern stems were not in common use, they served for small and ornamental "Memoirs B. P. Bishop Museum, Vol. I. No. i. 72 Mat and Basket Weaving. baskets, as at the present day. Three ferns bore the general name iivaiwa;'' Aspleniuni Adiantwn-nignim, Pteris decipiens and Adiantum capillus Veneris; all of these having smooth dark-colored stipes varying in color from dark brown throngh pnrple to black. A material not very flexible (except when soaked), bnt glossy and durable. In the days when Honoluln was a ren- de/A-ons for whalers in the Pacific a fash- ion was introduced among the Hawaii- ans that must have borrowed from meth- ods used long before, so that in describing it the mere curiosity of the result yields wholly to the sug- gestion of a manu- facture of fern-stem basketry, of which no other remains are extant. The hats for women, still made in fancy form, do not suggest any ancient origin. Captains of whalers almost invariably had on board their vessels at least one silk hat, not always of the latest block, which they were accustomed to wear ashore on public or important occasions, and these absurd and uncomfortable head-dresses, which fashion imposed upon the white man, roused great admiration in the breast of his dusk}' imitator. Wives or sweethearts soon fashioned the respedlable imitation (it was in the days of the early missionaries), which is shown in Fig. 72. Fern stems and horse hair are the components, and it was certainly better fitted to withstand tropical rains than its silken prototype. The particular specimen figured, now in the Bishop Museum, was "Andrews givts the (Iciinition "He mea ulu e like me ke palaa, ua ulana pa ia me ka papale mauu." FIG. 72. HAWAIIAN FERN-STEM HAT. Fern Stem Baskets. 73 given to the Rev. Rufus Anderson, D.D., Secretary of the American Board of Commis- sioners for Foreign Missions at the time of his visit to these Islands in 1863. About the same time I saw in the streets of Honolulu a powerfully built native clad onl3' in such a hat and his scanty nialo. To the present day the Hawaiians show a respect for this form of hat, although now select- ing the imported va- riety. The specimen illustrated weighs 10.5 ounces. In the M u s e u m collection is No. 8090, shown in Fig. ']2,^ a graceful modern bas- ket made for sale rather than for use. It is Japanese, but was brought by two old natives, known to me for many years as dealers in mats and baskets, with the assurance that it was made by Hawaiians. On their statement the figure was in- serted and remains as a warning to col- ledlors of basketry. On the other FIG. 73. MODERN JAPANESE FERN-STEM BASKET. ffrouDS to the south- ward more use was made of fern stems, especially for fish traps and baskets, as they are not readilj^ rotted in water. In New Zealand are made fine eel traps {Hinaki) and baskets in which to keep fish alive i^JVhakarapa)^ both shown in Fig. 74. These are made of vieiigemenge fern {^Lygodiiini articulattini) a twining fern with stems of great length. The genus is fovmd as far away as New England. The stipe of the Henthcru fern ( Todca {Lcptoptcris) superbd) is said'" to be used in the manufa<5lure of combs, ^°A. Hamilton, Maori Art, p. 302. 74 Mat a)id Basket Weaviiig. Hcni, but the stipita of this splendid fern are very short, seldom exceeding three inches in length, and, however fit for the teeth of combs, not suited for weaving baskets. Whnkarap.i. Hinaki. I JflG. 74. WHAKARAPA AND HINAKI, NEW ZEALAND FISH BASKETS. The illustration shows that as specimens of basketry both the Hinaki (compare Hawaiian Hinai basket) and Whakarapa are well and strongly made. The handle of the latter is of a form unusual in this region and resembles the bail of a pail. In the Hinaki one long handle extends nearly the length of one side, strengthened in the middle ; while two smaller ones are at the lesser end and 90° from the main handle; the cover is hinged and has a loop to fasten it, and the entrance for the eels is at the larger end. Fern Stem Fish Traps. 75 No. 6954 is a very large fish trap made of the same material, measuring 48 iu. in length, and 24 in. in diameter. It has three loops on each side near the smaller end. When wet it may be folded flat, although very stiff and firm when dr}-. The exquisite braid work of the Solomon Islanders has generally been described as of fern, and has even been attributed to a species of Gleichcnia^ but we have in the Museum specimens of a grass from which this attractive work is made, so we must defer farther notice of this to succeeding pages, where the grass manufadlures take their turu. Fig. 75. HAWAIIAN FERN-STEM FISH TRAI'S. So far as is known the old Hawaiians made nothing of fern stems of such great size as the Maori hinaki ; indeed they had no fern stems to compare with the Lygodiuvi of New Zealand, but the}- made similar traps of the much shorter Hawaiian ferns. Two of these are shown in the figure, and I am informed that such traps were much used in ancient times; and while sometimes made of other material, as at the present day, the fern was much preferred, and considered more durable. It is quite possible that the dark color of the fern gives some advantage to the fisherman. Hawaiian Sandals =Na Kamaa Maole. — Necessity must, in very primi- tive times, have given birth to the invention of sandals in a coimtry where volcanoes often spread lava streams of most extreme roughness across the paths until for miles a surface is formed that not even the hoofs of a horse can stand. Few who have not dis- sedled the foot of a man accustomed to go barefoot can have any idea of the exceeding 76 Mat and Basket JVeavhtg. toiighness of siich a man's sole; but as little can one who has not walked over it appre- ciate the sharp roughness of a stream of Hawaiian aa^ the roughest form of lava. The toughest soled native needs some artificial prote(?lion for his feet. In climbing Manna Loa, the largest of the Hawaiian volcanic mountains, in 1S64, the author wore out three sets of rawhide sandals used as prote(5lion to stout walking shoes, and only occa- sional beds of aa were encountered. No one could travel around the island of Hawaii, near the coast, without meeting man}' miles of this rough road. Where a permanent path was needed smooth beach pebbles were laid for footway, but on occasional journeys one often found no such provision for his feet, and he was forced to weave or braid some mat-like strudlure for sandals. Any tough fibre at hand was pressed into service; pandanus {/au/iala), dracsena (/c?'/ Xv'), hau bark (/// //a/t), banana (///// maia), poaaha or waoke, all answered sufficientlj' for the temporarj^ need. Plate in shows examples of all these. In strudlure there were two general wa3-s of plaiting the hastily constructed basketr}- : one, the simplest, consisted of a stout cord of any suitable material formed into a loop for the toes and over and between the paral- lel sides formed by the cords the leaves were braided or entwined, the tougher stems be- ing left for the under surface, as shown in the first example on Plate III. The loose ends of the looped cord served to bind the " ' - • • sandals to the ankle. The third specimen shown in the same plate is a little more complicated, there being four instead of two cords about which the weft is twined. Fig. 76 shows this more clearl}', the loops work in opposite direcftions and when pulled draw the substance of the sandal together. When waoke or maia was used a very comfortable and serviceable foot support resulted : such are used even now by the white man as well as by the Hawaiian when walking much on the rough coral reefs. Of the sandals in the Bishop Museum at this writing the following list will fairly show the comparative frecjuency of the material used : 4535. Sandals of lauhala, dr_v but without special preparation. 4537. Sandals of lauhala. 4538. Sandals of poaaha or partly beaten waoke, rectangular in shape. 4539. Sandals of poaaha, thick and tough. 4540. Sandals of banana leaf stem. These are always rectangular. 4541. Sandals of banana leaf stem with ili hau cords. 4542. Sandals of malina (American aloe). 4543. Sandals of hau bark ^ ili hau {Paritium tiliaceuni). Hawaiian Makaloa Mats. 77 4544. Sandals of lai ki [Coi dyliiic Icnniualis). 4545. Sandals of lai ki for a child. 4546. Sandals of lai ki, stont and coarse. 4547. Sandals of lai ki, well braided. 4548. Sandals of lai ki, from the Queen Emma collection. 4549. Sandals of malina or aloe. 5107. Sandals of poaaha or partly beaten waoke (yBroiissoiictia papyrijcra) . 8986. Sandals of lai ki. 8987. Sandals of lai ki. 8988. Sandals of lai ki. 8989. Sandals of lai ki. 8990. Sandals of hau bark = ili hau. No other covering for the feet than these sandals were known to the Hawaiians before the advent of foreigners. Although they had the pig I do not know that they ever used pigskin for making sandals, not even after they had seen the pigskin-soled grass slippers of the Chinese, who were earlj' immigrants, Vancouver finding, only fourteen 3'ears after the death of Cook, many of these orientals in the country. Niihau or Makaloa Mats. — In describing the Hawaiian baskets of ieie mention was made of another notable manufadlure of these islands, one which is now fast passing to the region of lost arts where the ieie baskets have gone. A few old women still make the mats from the sedge which grows commonly enough along shores and in brackish marshes. It was chiefly on the little island of Niihau that the maka- loa mats were made, although the sedge {Cypcnis Icpvigattis) grows on Oahu and other islands of the group, and it is claimed that those with colored figures {^Na moena paivehe) were not made elsewhere. Exact information is hard to obtain on such mat- ters, but certainly Niihau, populous enough in olden time, was the chief fac^torj' and gave name to the fabric. This was doubtless the mat noticed with admiration b}- early voj^agers, of which mention has been already made. The preparation of the sedge was simple, but only while the plant was 3'oung could the finest mats be woven, so that the time of working on these was limited to a few months each year, and the people had not learned the method of procuring a suc- cession of crops. The weaving of the mat was begun, as shown clearly in Plate X, and the leaves added as needed. In comparing this figure with that of the pandanus mat. Fig. 54, the great difference in the length of the staple will be noted. As with the hala, the leaves dried over a fire were whiter, and in the case of the makaloa these were called olala; it was the favorite kind of leaf used, except perhaps for the large 78 Mat and Basket Weaving-. floor coverings. The red figures are formed from the lower portion of the stem, which is naturally colored ; but while this color is fairly durable, it turns a dirty brown on long exposure to the light. The fresh color, both of the olala and the red figures, is well shown on Plate X\T. These figures are an embroidery and do not show on the rever.se. The patterns used are very simple geometric ones that do not present the variety shown in the decorated Hawaiian kapa, but the nature of the texture kept these within narrow bounds. Even on kapa, where the surface admits of any form, we seldom find ^¥\^> '■■-'W'? pi^p^ V*- a^gs^ Fig. 77. N'liHAr makai.oa mat. the attempt to represent natural obje6ls; almost always it was ornamentation com- posed of triangles, rhombs, squares and lines. All these patterns had names, and I have been at some pains to preserve these wliich will be found in the illustrations (Figs. 77-Si, and Pis. XI-XIII, and XVI). Not only did the weaver use these exter- nal decorations, but in the very weave she used her fanc}- in a way seldom found in the hala mats, and some mats are a series of stripes, each of a different weave. These fancy stripes sometimes alternate with those of a plain weave, and on these latter were generally embroidered the red figures. One mat in this colle6lion. No. 10,072, recently made, shows not only nearly all the usiial figures biit is woven in nearly all the known styles, including fancy twills. The names of the patterns are as follows: Mat Designs. 79 Keekee. Bent, zigzag; a favorite form, remindiug one of the Egyptian hierogl3^ph for water. Fig. 78. Olowahia. A saw; this is also a favorite design. Fig. 79. Kuhann. Breath of Ku (the god of war). Fig. 79. (Second and fourth band.) Humnuniki. Squares joined. Fig. 80. (Central band.) Papanla. Red row. Fig. 80. (Second band from bottom.) Pnakala. Rough, like the leaf of the Argemoiic. Fig. 80. (Bottom band.) Hale = house. Fig. 81. Nene = geese fl3ang. Fig. 81. Papa konane = checkers. In some of the oldest mats the keekee and olowahia alone appear; in others the weaver seems to have exhausted her list of designs, as shown in Fig. 77 and Pis. XII, XIII. A combination of triangles was constantly used, either a large triangle made up of a series of smaller ones (Fig. 83B),or a row where the apex of one touched the centre of the base of the succeeding (Fig. 83G), or two triangles were arranged with the apex of each in juxtaposition (Figs. 80, 830): the last was greatly used as a central dotting. The combination of triangles and squares, of which the distindlive name is not known ( that is, they are not found in my note book of fort}' years ago) may be seen in Fig. 83, where the numbers there given are used in the descriptive list of the mats given below. After the islanders had learned to read, under the instru'ffM>^4Mft>^fi^^'
  • >- FIG. 83. MAT DESIGNS. 2576. II. 9 ft.X7-5 ft. — 10 to inch. Keekee and olowahia in short stripes. 2577. 10.5 ft.X7.5 ft. — II to inch. 2578. 8.5 ft.X6.5 ft. 2579. II ft.X7.5 ft. 2593. 10 ft.X7 ft- 6878. 9.3 ft. X 6.7 ft. Keekee and olowahia stripes. 10,072. 9.2 ft.X6.7 ft. 6633. 8.5 ft.X6.5 ft. — 12 to inch. Triangles, squares, etc. 6634. 9.2 ft.X6 ft. — 9 to inch. Stripes and raised bands. 84 Mat atid Basket Weaving. Coarser mats were made from akaakai {^Scirpus laciistris) of which abundance is found in the marshes around Honolulu. These were used as mats for temporary purposes, or sometimes for the lower layers of the hikiee bed, but they were not durable and the surface was soon destroyed. The following are examples from the Museum colleAion. FIG. 84. GRASS MAT. MoENA Akaakai in the Bishop Museum. 2584. 5ft.X2.5ft. 2585- 6ft.X6ft. 2594- 10 ft.X8 ft. 2595- 8.7 ft.X7 ft. 2596. 20 ft.X 10 ft. 2597- 9-5 ft-X7 ft. 6635. 11.6 ft.X9-3 ft. — 7 to inch. Heavy floor mat. Grass Baskets are not much in evidence as an Hawaiian product, but in the Bishop Museum is a very pretty basket from Maloelab (Calvert's Id.) in the Marshall group. No. 3347 in PI. I. This is said to have been produced in 1849 and has been somewhat of a puzzle as to its constituents. It is small, height 4.7 in., diameter 3.2 in., and the bottom is a close ordinary weave of a dark brown grass which also is IVearn'ug of Grass. 85 gathered into vertical rods and horizontal coils, these two elements forming a trellis not interwoven but the vertical rods are all outside and the horizontal rings all inside; over this trellis runs a sewing strip in four colors, straw, chocolate, white and yellow These various colored strips seem to be split straw, but I have not definite knowledge of them or of the dyes used. The ornamental bands are shown in the plate, the dark is always chocolate; the interior of the upper band is half straws and half M'hite, of the lower band, j-ellow. The vertical rods are closed in at the rim by a braid which con- ceals the ends. The interior is rough where the ends show : there is no handle nor cover. Solomon Islands Grass "Work. — Although authors have stated that the beautiful braids of black, red and yellow material that the Solomon Islanders are so fond of using to adorn their choice spears, clubs and, in ciirious company, their combs, was from the stem of a species of fern i^Glei- chenia)^'' it now seems settled that it is a grass, species unknown, that is used. We have in the Bishop Museum a parcel (No. 8287) of the dyed grass used in this work brought from Bougainville by oiir colleftor. The native name is latilc^ and it seems to have a semi-sacred nature for it is used to form amulets, and perhaps hence its appro- priate use in choice spears and clubs. It is said to be dyed with lime and the root of a certain tree {^Moriiida f). The dyed grass averages 27 inches long, and the leaves on stem split feadil}^ showing a considerable rough fibre beneath the smooth surface. Although the clubs and spears show a great deal of fine work, I am inclined to give the palm to a comb in this Museum, No. 1941, shown in Fig. 85. This is made of the usual palm splints (not whalebone as mentioned by some authors), and of a form well known in colle6lions. This particular specimen was colleAed on Bougainville fifty years ago and is 7.5 in. long. The design is alike on both sides, biit in the narrow portion of the handle the colors yellow and black are reversed on the side not figured. The two bars of the lower portion are red embroidered with yellow. The braid at the base and around the upper and middle portion is brown. The illustration unfortunately can give no adequate idea of the color and the beautifully exact and appropriate design. Another specimen (No. 6329) is smaller (6 in.), and although made on a plan similar to the preceding shows signs of deterioration in the work- manship: it was colleAed in 1903. "Ling Roth, Spears and other articles from the Solomon Ids. Archives Intern. d'Ethnologie, xi, 154. Pig. 85. COMB OV PALM WOOD AND GRASS BRAID. 86 Mat and Basket Weaving. One of the fine spears in the Solomons colleAion has twenty-four bands of woven red and yellow covering, separated by plain circumferential bands and exhibiting per- haps six or seven different designs. An implement called a "chief's wand or sceptre" has ten of such woven bands. The weaving on the clubs of lenticular sedlion is very exact and tense or it would soon get loose on the tapering bodj', unless, as sometimes seems to be the case, it is cemented on. Spears with carved head inlaid with pearl shell often have a narrow band of this finely woven covering. Bambu Work. — We may now return to Ha- waiian matters. Modern fans are often made of split bambu which affords thin laminae separating much as does the paper birch bark of our north- ern regions. These strips, of varying width and length according to the use intended, are very beautiful and the fans made from them (which fig. S6. baskets from northern Australia. are perhaps much" less beautiful than the bambu strips of which they are constructed) are shown in Plate XV, where Nos. 19, 22-24, in the lower part of the plate, are of this material. The brilliant white is often contrasted with the oiiter skin of the black banana. A golden brown fern stem is also used in these fans for color effedls. The banana is shown in No. 19, and the fern in No. 7. Sugar Cane. — An uncommon but very beautiful material for braids used in hat making is found in the sugar cane. This grass was found on these islands by the early voyagers, but so far as I am aware its use in basketry is modern. The strips are very glossy and become, by age or exposure to the sun, a golden brown. Cane leaves have long been used for thatching the native houses, but are less easily worked and much less durable than the common pili grass. Baskffs frojit Australia. 87 ii,,in>w\iimi«Mti>>wa>.iitf*<«w"",T;jjw mlj.' .1...' I aM^liU^i I I ' N-ii.^|>W»i*>VM7ii|iilri#lV'';A"*'t..,,, . Australian Baskets. — All tlirougli thi.s account of Hawaiian basket and mat weaving illustrations have been drawn from other Polynesian sources, nor has that been the limit, for Micronesian and Papuan sources have also been freely taxed in order that perchance the geographical origin of certain forms or methods might be indicated : and for this it seems as important to show the work of tribes within the Pacific area, even when that work appears widely divergent from that forming the basis of this treatise. If we had a fairly complete colledlion of Pacific basketry, such as this Museum is striving to gather, much might be gleaned of the traces of ancient intercourse, if not of common origin, of the tribes whose descendants are now verging toward extinAion on the islands of the Pacific. This would be truer of baskets than of mats, for the latter have always been deemed of greater value as property, and have always been favorite obje^ls of barter or of plunder. Reference has already been made to the care be- stowed by the old Samoans on their choice mats and the great length of time during which they have been preserved. The baskets seem to have been made under a less favoring star, and however useful, however ornamental they may have been, they certainly have not stood so high in the estimation of their makers or owners as have the mats. Perhaps enough of the human has clung to them from the busy fingers of their makers to impart to the sense- less intertwined and knotted strips the usual human lot, where the quietly usefiil people are, when dead, soon forgotten, while the brazen warriors or the astute politicians are preserved in marble or bronze or aere perennms in the pages of history. The coiled form of basket, so common among the Amerinds, is also found spo- radically in the Pacific regions, as at Fiji, New Britain, New Guinea, and here in Australia. All the Australian coiled baskets that have come to my notice are knotted coiled, that is, the thread that unites the adjoining circles of the coil are knotted be- tween the coils, as shown plainly in Fig; 87, or perhaps better in the specimen in this Museum given in Fig. 88. In all the foundation of the coil is some small fibre or grass, to which I am unable to give a name, and the connedling thread has much the appearance of rattan, and is perhaps from some vine allied to the Calanms. The best FIG. 87. COILED BASKET IN THE .\USTR.\I,IAN MUSEUM. 88 Mat and Basket Weaving. ~2tC li^.n.ima j4«QSfc. description of these and other Australian baskets is found in R. Brough Smith's Abo- rigines of Victoria/^ and in the account of certain decorated baskets by R. Etheridge in the Archives internationales d'Ethnographie.''' In the specimen in this Museiim (Fig. 88) the shape varies a little from that shown in Fig. 87, which is more like that figured by Smith. Here the diameter is 10 in., height 6 in., with a handle of Eucalyptus fibre ( Eucalyptus obliqita ) net- work. The native name in Gipps- land is Minni-gnal-ok. This old form is now scarce, as the native women make them with many modifications to sell to whites. This basket is strong and elastic but not finely made. A peculiar form of coiled basket is shown in Fig. 89. This was formerly common among natives of South Australia but now rarely seen. Eyre says that in one part of South Australia this basket is tailed Pool-la-da-noo-Ko.'^ Two are in this Museum; the one figured, No. 1916, is 8 in. in diameter, and is decorated by spiral stitching; the other is of the same diameter but has a neck of three coils, and is also decorated by red stitching. In both the handle is formed by a continuation of the coil. As to the material. Smith says that both Poa australis and Xerotes longijolia were used for basket ^"^- ^^- coiled basket from orppsLAND. making. In Fig. 89 is also shown a common form of basket of which two are in this -'R. Brough Smith, The Aborigines of Victoria. Melbourne, 1878, vol. i, p. 345. -••Archives internationales d'Ethnographie, xii, p. i. On the Ornamentation of some North Australian "Dilly Baskets." A study in Australian Aboriginal Decorative Art, by R. Etheridge, Junr. , Curator Australian Museum, Sydney. Mr. Etheridge has also described the manufacture of baskets in the Macleay Memorial, vol. Lirn. fo. N. S. Wales, 1893, pp. 247 and 249. "Aborigines of Victoria, p. 345. 1 Atistraliaii Dilly Baskets. 89 Museum. In the one figured, No. 7430, from Queensland, the weave is quite open, and the basket at the mouth measures 8.5 in., is 12 in. high, and 14 in. between the points. The handle, attached to one side onl}', is of white strip long enough to go over the shoulder. The other speci- men is larger, measuring 10 in. at the mouth, 14 in. high, 16 in. from point to point. In both the handle is securely fastened to three of the regular ribs and to an additional rib inserted for strengthening this at- tachment and extending only half around the basket. The choicest of all the Australian baskets are those known as "Dilly Baskets" and used to carry the small tools and treasures of the owner. Three are figured in Fig. 86 and another in Fig. 90. Seldom of large size, they sometimes are very small; in one figured by Etheridge {loc. cit.^ p- 5) the length is only 6.7 in., and the mouth only i in. in diameter. This little bas- ket was decorated by first coating the entire surface with Indian red pigment and then with lighter red, white and yellow, the orna. mentation was completed. The Australian Museum possesses this gem of Australian, basketry. Of the three dilly baskets shown in Fig. 86, the first. No. 8741, is 9.5 in. long, and 5.5 in. in diameter. The strudlure is loose but strong, of a grass-like fibre PIG. 89. AfSTRALIAN BASKETS. 9° Mat and Basket Weaving. and this particular variety is often much larger, as shown in a specimen in this Museum, No. 8740, unfinished, which measures 18 in. long by 14 in. in diameter. The second specimen, No. 8750, measures 13 in. in length by 5 in. at the neck, and 7 in. through the collar. It is decorated with white lines on a dark red ground. No. 8756, the third in the figure, measures 12 in. long, with a diameter of 5 in.; the original decoration of red zigzags is nearly obliterated. These three dilly baskets were colledled by Mr. Harry Stockdale, of Sydney, in the Alligator River District near Port Darwin, and are FIG. 90. IIKCOKATKII DII.lv BASKKT. FIG. 91. PLAIN DILI.Y BASKET, AUSTRALIAN MUSEUM. now in this Museum. From the same district, also from Mr. Stockdale's colledlion, No. 8755 comes to this Museum, and it is a choice specimen of the most elaborate decora- tion of the Australian basket makers. Unfortunately the method of coloring does not prove lasting, and the pigment rubs off much in the manner of ordinary whitewash ; still in this specimen, as may be seen in the illustration, Fig. 90, enoiigh remains to show that the skill of the decorator was above the ordinary. The decoration does not extend quite around the basket and is divided into five zones, the upper one composed of one horizontal and a number of vertical white lines; below this seven vertical lines, rather shorter than the last, flanked by two rosettes and what seems to be a hunting scene. Then comes a corroboree in which the male figures are quite distinct; then a band of confused figures, Decorated Dilly Baskets. 91 '■ * f ' and the curved base of the basket lias a row of white squares, then a zigzag line below which are other triangles and dots. The basket is closely woven and of good form. It does not seem to have been figured by Mr. Eth- eridge, whose familiaritj^ with all forms of Austral- ian decoration might have explained more clearly the designs. A more complete description of this import- ant basket is as follows : />^ * ' No. 8755. Length 14 in., diameter 5 in., rounded bot- tom. Upper rim sewed with stout twisted cord : surface divided into seven zones by double raised bands ; painted dull red, on which ground are black and white designs which extend over more than one- third of the circumference, and there are indications of other marks on the side worn . next the bodj' now unde- cipherable. Upper zone nar- row and plain ; second has 14 vertical white stripes and rosettes at each end of the series ; third has 7 similar stripes, but shorter, and with rosettes at each end ; fourth has what seem to be hiero- glyphics(?); fifth has a cor- roboree with 5 male figures ; sixth has confused marks, unintelligible ; .seventh has 7 rectangles forming a zigzagr with a figure in the midst FIG. 92. KETE OF H.\R.\K.\KE FIBRE. rcsembling a three - barred gate, triangular spots within lower angles of band which extends two-thirds of the circumference ; 5 square spots above the band and the same number below, this zone ending in 5 concentric rings. Inside of basket unpainted ; carried by five cords attached in two places on one side of the rim. A plain dilly basket in the Australian Museum is shown in Fig. 91 by the kind- ness of Mr. Etheridge. It has a rather flatter base than the others, and the rim is slightly detached from the filling. UHKNiCE fAUAHl BJSHiJF MU^hUM 92 Mat and Basket Weavmg. New 2/ealand Kete. — The so-called "flax" of New Zealand {P/ioniiiiivi tenax Forst., native Harakakc) is one of the most important commercial fibres of the Pacific region. This liliaceous plant, of which the observant old Maori distinguished fift}' or sixty varieties, and the botanists note two species, grows wild over a vast extent of marshy land in New Zealand, and the traveler through the North Island sees the plant with its tall flower stems on everj^ side. Indeed on landing at Auckland one sees bale upon bale of the attractive looking fibre awaiting shipment. It is unnecessary here to go into the commercial preparation of the fibre for several full accounts of this are accessible, ^"^ and we need only present some of the producT:s made from this raw material in the line of our present studj-. Kapa or bark cloth made of felted fibre was useless in the cool and wet climate of New Zealand, and the Pol3mesian immigrants must soon have de- vised a way to utilize a fibre as beautiful as it is tenacious and durable. It is often stated in books on New Zealand that no implement or machine can clean the fibre with the perfedlion of that very primi- tive tool a Maori thumbnail, and while this is tnie in a sense it must be acknowledged that modern machinery certainly turns out a very fine product. It would be very interesting to go more fully into an examination of the weaving of the wonderful cloaks, of which extraordinary examples are in this ^^^- 93- maori satchels. and other museums, but this has been well done by Mr. Hamilton, now DireAor of the Wellington Museum {loc. cit.^ p. 271), and his illustrations show the perfedlion to which the native manufadlure attained. There was no loom, but the pegs {tn ni / ii rn-parawai) which held the web for the webster during weaving were often grotesquely carved. I have, however, thought best to give some examples of the kete, kits or satchels, in this weav- ing as they illustrate the work as well perhaps as the more elaborate cloaks. Fig. 92 shows two of these, the i:pper one, No. 5S19, is of bleached fibre mingled with a portion dyed yellow (with a Coprosmaf). The weave is very simple and effective, as can readily be seen in the illustration. In some remarkably fine and costly ones that I found in the Taranaki district the weaver had introduced the beautiful cone of '"' Elsdon Best, The Art of Whare pora Maori flax weaving: Trans. N. Z. Institute, xxxii. A. Hamilton, Maori Art, quotes largely from Mr. Best's essay. See also the Reports of the New Zealand Flax Commission. Sir James Hector, Phormium tenax, 1892. Maori Mat Work. 93 Mt. Egmont in a triangle as easily recognized as the innumerable representations of Fujiyama in Japanese decoration. In the lower example in the same illustration, No. 1590, the white and glistening fringe admirably sets off the black body of the kete, through which are woven patterns of the Maori tribal tatauing, and these patterns are not less interesting for their coincidence with many of those embroidered into their makaloa mats by the Hawaiian women as shown in Fig. 83. It may simply mean that a triangle is an easy figure for the technic of the mat, but the resemblance is there between the ancient decorative forms of the two Polynesian families now at diagonally oppo- site parts of the Pacific. Many pages might be devoted to a descrip- tion of the various waj's in which the long flat leaf of the Phormium is used in garments waterproof and orna- mental, of which this Museum has a fine col- le6lion, but these are Fig. 94. MAORI BAST SATCHEL. , . • n i • i not strictl}' either mat or basket work, and we must pass to the use of the leaf in basketry where it takes the place of the pandanus of warmer climates. Fig. 93 shows several of these and tliej- are still made of every size and variously decorated. The larger ones in the illustration, Nos. 7585-86, are woven inside out from the bottom and then turned, leaving a rough seam within. No. 7588 is, I think, made of Nikan palm. The two small satchels are woven without the bottom seam. The handles of all except the bottom one are of harakake fibre; of that the handle is of twisted leaf. The leaves of the so-called cabbage-palm (a Cordylme) are also used in making these satchels and for other similar purposes, but they are not so flexible as the harakake leaves. It is not only the leaves but the flower stems that are useful in Maori mat work, for the panels between the sculptured slabs in the carved houses of the Maori are made of these cylindrical sticks combined in great variety.'^ This is due either to alterna- tions of color in the sticks, or depends on the arrangement of the strips that bind these together. New Zealand is often described as a group possessing a wonderful ■'Hamilton, loc. cit., p. 86, PL xiii. These panels were called tukiitnku. 94 Mat and Basket Weaving. variety of woods both useful and ornamental, and I am tempted to give a single example of what might be regarded as a refined "splint" basket. Fig. 94 shows No. 6562 which is simply woven of the bast of antetaranga {Pwielea airnaria). It is light and per- haps flimsy, but none the less artistic and attradlive even as a mere ornament. Fibre Mats. — Passing again from the Hawaiian group, where mats of fibre were not made, to Samoa, a group so closely allied to the Hawaiian iu langiiage, cus- toms and physical form, we find the fibre of the bast of the fau (Samoany^?;/, Hawaiian liau^hibisais) iised for fine mats which were greatly valued. The Hawaiians knew the hau and used its fibre for many textile purposes, but did not reduce it to its con- FIG. 95. IE SINA, SAMOA. stituent fibres, a process well known to their southern brethren. Far to the west of these Polynesian groups the people of Micronesia, as we shall see presently, made great use of the hibiscus bast divided into fine strips, although not made into threads as in Samoa. On the latter groiip the ie sina were woven by hand, without implements, and while sometimes made of the unbleached bast strips, as No. 2193, in this Museum, they were usually of fine thread made by pounding the bast and then bleaching the fibre until it is as white as well cleaned banana fibre. In No. 2186 the finished mat is dyed with ochre or some other dye producing a red-brown color. Not only were these mats of comparatively fine weave, but their substance was more than doubled by a nap or pile put on after the mat was woven by passing a parcel of the fibres with a full turn about a mesh of the mat at suitable intervals, and these can be pulled out only by loosening the loop formed about the mesh ; pulling on the ends only tightens the hold on the mat. The length of this nap was variable, but in Loom-woven Mats. gc No. 3570 it is seven inches, and the resultant mat looks like a sort of vegetable fur, as may be seen in Fig. 95, and like fur would make fairly warm and comfortable gar- ments. For presents these mats were greatlj- valued, but their use in the curious "Tokens of Virginity" custom adds especial interest to the ethnologist, and it is not iincommon to find them stained with human blood. Of those in this Museum the fol- lowing list shows size and fineness : — Samoan Ie Sina in the Bishop Museum. 2193 21S5 3570 3571 3572 21S6 4 ft.X3 ft. — 9 to inch. Unbleached, unheckled bast. 5 ft.X3 ft. — 17 to inch. White, fine fibre. 6.1 ft.X4.2 ft. — II to inch. White, fine fibre. 5.7 ft. X 3-7 ft. — 9 to inch. White, fine fibre, coarser weave. 6.7 ft. X 3-5 ft. — 13-18 to inch. White, fine fibre, repaired. 4.7 ft.X3 ft. — 10 to inch. D^-ed, fine fibre. The weave is so loose that it is easy to increase the width at the expense of the length, so that the measurements are approximate onl}'. I/OOm-WOven Mats. — From the Gilbert Islands on the east to Guam on the west we find a rude but efficient loom for weaving fine mats, while throughout Poly- nesia no looms are used, and deft fingers must plait together the strands of whatever material to form a mat. With leaf strips or grass stems this is not very difficult, but with fine flexible threads the difficult}^ is increased, and the finer the thread the greater need of some mechanical assistance which the loom bars, however simple, and the shuttle afford. Two types of weaving apparatus are found in this region, one which is a loom in all its features, from which can readily be traced the more complicated forms of the modern textile art, and a specimen from Ruk in the Caroline Islands is shown in Fig. 96; the other, which, so far as I am aware, is confined to the island of Kusaie, is a very different contrivance, and although I have two specimens in the Museum before me, I cannot understand fully its working, nor have I been able to gather from those who have visited Kusaie anj^ intelligible facts regarding its employment, and of this t3'pe Fig. 98 presents an example. Under ordinary circumstances one should be able, when he has the tools used and the finished product, as in the present case, to connect the two, but in this I have failed, and the Masters Finsch, Kubar\' and Parkinson, who have explored that region with ethnological skill, have failed to help me. The latter gives a full and interesting account of the first loom,'^ and from his account I shall take the liberty of quoting all that may explain more fully the loom before us and its work. It is interesting to find that on this island of Ontong Java, which was named by Tasman in ^'Nachtrage zur Ethnographie der Ongtong-Java-Inseln, B. Parkinson: Archives Interna. d'Ethnographie, xi, p. 207. In a note on p. 242, a figure is given to show the horizontal position in which this loom is used. 96 Mat and Basket Weaving. 1643, although previously seen by Mendaua, the inhabitants appear to be of Polynesian origin (and the figure of the weaver referred to in the note is strongly Polynesian), and the men are the websters, while in Poljmesia proper the mat weaving is the women's work. Here then is a Polynesian family who have adopted the loom of their neighbors without alteration or improvement, and the question not u n n a t urall}' arises how it is that none of the other tribes did the same thing, if they all entered the Pacific through the western gate ? In all these sim- ple looms the warp- beams are two sticks of equal and suitable length called o. In the simplest form in this Museum these ^'^- ^'- ^°°" ''^''^' c^^o^^^n isi^ands. are lengths of the light and smooth stem of some large palm leaf, but in the better one figured they are of heavy wood cut with some care and knobbed at the ends (Fig. 96, A). The widest of these seldom exceed three feet, and those of Santa Cruz are hardly a third of that length. Around these pass the longitudinal fibres or warp; these are continuous and slide on the beam. The length of the warp determines the length of the mat, and the number of warp threads its width. The beams armed with the warp are stretched in a horizontal position, one being held by cords passed around a post or Alicroiicsiaii Looiii. 07 tree, and the other by a belt or sliug (h) which passes behind the back of the weaver, who is seated on the ground, so that he can easily regnlate the tension on the warp. There was no yarn beam and cloth beam as in American and European looms of higher development ; the whole fabric, warp and completed mat, was free to move over both beams like an endless apron. A similar arrangement is found in the Zufii loom."' The slings are called kiPii. Next the weaver are two bambu sticks, about an inch wide, called api (1;), thrust through the warp, so as to separate this into an upper and under layer. This would be suiificient in the simplest loom, but usually a rude heddle or harness, na^ (the sticks next the upper beam in the figure), consisting of a thin stick with a continuous cord looped along its length, and through these loops every other thread is passed before the loom is set ujj. In more complicated looms there are often man 3^ of these heddles through which varying portions of the warp pass and which are lifted at suitable intervals by machinery. In our present loom this motion is by hand, and it operates to "form a shed" by raising alternate threads from their neighbors far enough to allow the shuttle, si^ ia (G), on which the filling or woof, ago s/''/(7, is wound, to be thrust between the two layers from right or left. When the shuttle has passed, the third process comes into plaj-, — the thread left in the shed (lay) is beaten up with a lathe or batten (d). Often the shed is formed, as in the loom from Ontong Java, by a sword-shaped strip of wood, /aga, which the weaver holds in his right hand and there- with separates the laj-ers of the warp so that the shuttle can easily slip through, and also serves to press the last thread of the woof close to the preceding one. This cycle constantly repeated completes the weaving. Of course the threads that were lifted for the first passage of the shuttle with the woof ai'e depressed for the next passage, or, what amounts to the same thing, the others are lifted by being passed through another harness. When the shuttle is emptied of filling another is substituted, and the junction of threads is made by a neat knot j^recisely as in the more complicated loom of modern fadlories. That my readers may have the full benefit of the German description, I give here the important portion of Mr. Parkinson's account: — Die eiiizelneu Theile des Webeapparats auf Ontong-Java (olgeu in nachsteheuder Anordnuiig. Die Kettenfadeii {Hau) sind iiiii zvvei rutide Holzer geschluugen, welche etvvas latiger sind als die Breite der herzustelleudeti Matte. Diese beiden Holzer werden O genannt. Das ein wird niittelst einer Schlinge oder eines Baudes, an beide Enden des O verbunden, an einen Pfosten befestigt ; das andere Holz hat ein ebensolches Band, welches der auf dem Bodeu sitzende Weber um die Taille legt, wodurch er es in seiner Macht hat die Kettenfaden straff anzuziehen ; diese Schlingen werden KiCti genannt. Dem Weber zimachst sind zwei etwa 2-3 cM. breite Bambusstabchen {Api) Avixc\\. die Kettenfaden geschoben, so dass sie diese in eine obere und in eine untere lyage trenuen. Dan folgt ein diiuner Stock {Ka'o) an dem, durch eine fortlaufende Fadenschlinge {U'a), die untere Kettenlage befestigt ist, so dass, wenn das Stockcheu gehoben wird, die untere Kettenlage fiber die ='Otis T. Mason. A Primitive Frame for Weaving Narrow Fabrics. U. S. Nat. Mus., 1899, p. 492. Memoirs B. P. B. Museum. Vol. II, No. i. — 7. 98 Mat and Basket Weaving. obere gehoben wird, und man danii das Scliiffcheii (S/'i'a) mit dem Einschlagfaden (Oao sf'/a) von rechts oder links liindurch stecken kaun. Hinter dem A'a'o folgt ein zienilicli dicker Holzstab, manchmal auch ein dickes Stiick Bambusrohr {Porogu) welches die Kette trennt und mit dem Ka'o zusammen dazu dieut, die Kettenschichten abwechselnd zu heben nnd zu senkeu. Nach dem Porogu folgen abermals Zwei schmale Api wei zu Anfang. Zu dem Webeapparat gehort ferner noch ein schwertartiges Instrument {Laga ) welches der Weber in der rechten Hand halt und damit die Ketten- faden trennt so dass er das Schiffchen bequem durchschieben kann, ferner auch um damit durch- geschobenen Einschlagfaden fest an die vorhergehenden anzudriicken. Die Herstellung der Kette erfolgt nun folgendenmassen. Die einzelnen Theile des Apparates werden tlieils in den Fussboden der Hiitte fest eingesteckt, theils von Gehilfen in Position gehalten. Zunachst schlagt man in den Fussboden die zwei O fest ein, etwas weiter aus einander, als die Halfte der Lauge der herzustellenden Matte betragt ; manchmal uimmt man statt der O auch zwei dickere Stocke und ersetzt sie spater durch die O. Von unten anfangend legt man nun den Faden um diese Stabe, scliiebt aber zugleich die iibrigen Theile des Apparats mit hinein, namlich die vier --//>/ (je zwei), den Kao und den Porogu: der Kao wird gewohnlich auch in den Fussboden eingeschlageu, die iibrigen Theile von Gehilfen festgehelten. Der Faden wird nun so umgelegt dass er abwechselnd iiber oder unter den,-7/i/und dem PorogiiYAuit. wodurch die Kette in zwei Lagengetrennterhalteubleibt. Hochst sinnreich ist nun die Auorduung wodurch bewirkt wird, das man die untere Ketten- lage abwechselnd iiber die obere heben und darunter senken kann. Dies wird bewirkt durch den dicken Stab Porogu und das Stabchen Art '0 mit den Schleifen U'a. AUe Kettenfaden laufen unter dem Stab A'a'o fort, jeder Zweite Kettenfaden wird durch eine lose Schlinge an den A'a'o befestigt, so dass der Weber es in seiner Macht hat durch Heben des A'a'o die untere Kettenschicht iiber die obere zu heben. Der Weber setzt sich auf den Fussboden und spannt die Kettenfaden wie oben beschrieben an. Ihm zunachst liegen die beiden schmalen Latten Api d\& an beiden Enden durch einen Faden mit einander verbunden sind. Er fasst nun mit der eiuen Hand, den Stab A'a'o und hebt denselben, wodurch die untere Kettenlage iiber die obere empor gehoben wird, nun steckt er das breite, diinne und sehr glatte Schwert, Laga, eurch die entstandene obere und entere Schicht und dreht dasselbe um, so dass die Kanten nach oben und unten stelien, nun schiebt er das Schiffchen mit dem Einschlagfaden durch, legt das Schwert flach und schlagt den Faden leise, an worauf er das Schwert herauszieht. Jetzt schiebt er den Porogu etwas von sich ab, ebenso den Ka'o wodurch er bewirkt class die friihere obere Kettenlage die untere wird ; das Schwert wird wieder durchgesteckt, ebenso das Schiffchen, und der Einschlagfaden angetrieben, Porogu und A'a'o schiebt der Weber nun an sich herau, hebt den A'a'o so dass die Kettenlage wieder nach oben kommt und auf diese Weise geht es nun fort bis die Matte fertig ist. Ein geschickter Weber kann in drei bis vier Arbeitstagen, jeden von Sonnenaufgaug bis Mittag gerechuet, eine, Weibermatte fertig stelien. By arranging the warp threads of different colors longitudinal bands were formed, and by varying the color of the filling on the shuttle transverse stripes were produced; the former were more common. Twills or other fancy patterns could be introduced by increasing the number of harnesses or by raising, instead of alternate threads, two, three or more adjacent ones. This loom gave opportunity for decorative weaves that were earl}- discovered, and some of the results on the mats of the Carolines used as women's dresses are shown in Fig. 97. Both hibiscus and banana fibre were used in this mat weaving, as maj- be seen in the list of mats in this Museum given below. There are specimens of the same pattern in hibiscus from Guam and the Gilbert Ids., and if we did not know they were made in both places it would be easy to imagine their transport as merchandise for I Caroline Islands Mais. 99 FIG. 97. MATS OF BANANA FIBRE, CAROI,INE ISLANDS. lOO Mat and Basket Weaving. barter from one group to the other. Of the finer banana mats we have specimens from Ruk in the Carolines, Santa Cruz in the New Hebrides, and Nine. The Santa Cruz mats are remarkably well made and tastefully decorated, and are sometimes narrow and long, with pleats and tufts for additional ornament. The beauty and durability of these banana woven mats are so marked that it is surprising that a trade to foreign countries has not been developed. I do not know what the condition of the native manufaAure may be at present: perhaps, like so many good things of Pacific art, thej- have passed b}' to give place to the cheap calico that civilization has brought upon the islands. I The Tol. — Besides the loom just mentioned, the people of Kiisaie in the Caro- line Islands have another contrivance, by no means an orthodox loom, but still a "webstuhl," as our German friends might call it. To be- gin with, the writer must again confess he has never seen the instrument used, nor has got any intelligible information from any one who has, and in saying this he would disclaim any intention to speak disre- spectfully of the little machine or of the intelligence of those who would have enlightened i'ig. 98. framk for toi, weaving. him, if possible: it is simply in apology for not handing on in these pages some in- telligible explanation of its working habits. The machine, as will be seen from Fig. 98, is a biped with a long straight body, on top of which are inserted loosely certain pegs, around which are wound threads of banana fibre colored to suit the work in hand. Other spools of this fibre are at hand, if we can dignify a mere length of bambu, around which the thread is wound, with the name of spool. Then there is a frame which may possibly serve as the harness in an ordinary loom; there are shuttles of good form, and battens of considerable weight to drive home the woof or filling, and there are clam shells to serve as scissors. The threads, whether for the warp or woof, are of well cleaned banana fibre dyed in various colors, and in lengths of about five feet. To obtain the continuous length of thread these are neatly tied together b}' a double and almost invisible knot. I have given the machinery, and I am forced to give the result without the inter- mediate processes. There are two of these "looms" in this Museum, differing in size and ornamentation, but each provided with the same attachments, and it is a subject Rnk Loo7n-ivo7/X'c;=navel, beginning, /i/t/nu=^root, 01 /'^>c;///(^//7/=starting place. The term poalia has been ap- plied to this part, and many of the piko when completed would be large enough to act as a poaha. The latter was a ring of rope or bound pandanus leaves, placed on the ground, on which the rounded bottom of the umeke would rest. The second or main part was often referred to as koko, but there was a technical word, Iianai, to represent it ;" the word opii^=he\\y ^ any swelling surface, was also used for this part. The third, called kakai ox aUIu\ was a cord interlooped. Fig. 113, with or knotted, Figs. 114 and 115, to the outer edge of the lianai in two series. The names kakai and alihi seem to have been used as frequenth' for the suspending cords of the koko, but since the name alihi is also used for the head and foot ropes of the fishing nets, it might be better to retain the name kakai for the koko. Each series was bound in the middle (ultimately the top) by a single smooth winding, and .sometimes by half hitches, to make a/>« (handle), and into these pu was the end of the aiiaiiio^ bearing stick, thrust to carry the load, Fig. 153. The auamo, or miniaka, is a stick of hard heavj' wood, generally kaiiila {Alphitonia cxcclsa)^ about six feet long and borne across the shoulder. Fig. 109; the ends drop a little below the middle and are either notched or neatl}- carved to hold the kakai. Fig. 153. For prote(5lion of food against animals, the koko was suspended fi'om a wooden hook, kilon/' attached to the ridgepole of the hut, or from a crossbar shaped like a canoe and notched around the edges; this implement was called olco/e^'' or luika, and was placed on the top of a pole set in the ground. The work of making koko was done b}' one of the ka/iii or body servants of the alii. Frequentl}' a kahii well skilled in such arts was in the service of the king, who, to show favor to his friends, would place the skill of the kahu at their disposal. Another source of manufa6lure was the common people, who brought great numbers of koko to the alii in payment of taxes, and after the best had been selec^ted by the chief and his friends, the poor ones might return to the commoner. Since this essay is intended to place on record this part of the natives' art now forgotten, and the usefulness of which has entirely cea.sed, and also to give a catalogue ^Lorrin Aiulrcws, A Dictionary of the Hawaiian Language. Honolnlvi, 1S65. ^'Eflge-Partington and Heape : Eth. .\lbuni of Pacific Ids., Vol. Ill, PI. XV, No. I. ■"Ibid., Nos. 2 and 3. Koko Cotiiahiing Unicke. 113 a. Koko puupuu, Haiiai E. b. Koko pualu, Hanai A. (". Koko puupuu, Hanai D. d. Koko puupuu, Hanai D. Memoirs B. P. B. Museum. Vol. II. No. i. — 8. a. Koko puupuii, Hanai E. b. Koko puupuu, Hanai D. c. Koko puupuu, Hanai G. d. Koko puupuu, Hanai K. 114 Hawaiian Nets and Netting. of what is now available in this Museum to students of ethnology, greater attention has perhaps been given to detail than a general description would call for. It has been necessary to make a liberal use of the native names, which at the present day are liable to be inaccurately applied. And again it has been found that some- times in different islands of the group one name may be emplo3-ed for different articles, and dissimilar names for the same thing. Piko, — The piko is a ring of cord at the bot- tom of the koko, attached to which, by loops, knots or half hitches, is a row or circle of loops on which the hanai is begun. The habitual position of natives when doing all such work was sitting on the ground, with one leg over the other and the upper foot pro- jeAing slightly. From the big toe of this foot the cord was stretched, and the diagrams of the tech- nique herein have been drawn as if in the same posi- tion, unless specially mentioned. Of piko, there were observed fourteen forms in specimens obtainable, and in the following descrip- "i"- "'"■ "uhwai. g(hiki) water p.otti.k. tions of the modes of operation the arrow heads point towards the shuttle or ball of cord. Piko A (Figs, no and III). — A piece of cord is first knotted into a ring with the loose ends closely severed. In the first figure, at the point b the shuttle is passed over and under the ring, over the free end of the cord ^?, under and over the ring and back through the loop along a. At c a simple knot is tied on a and the half hitches at b are repeated at b' leaving a loop at d. The loops (from 6 to 12 ) are formed in this manner and completed by the return cord c making a knot with a at f. The work from this point is part of the hanai. FIG. 1117. OLOWAI. IIOUKI) WATER BOTTLE FOIt (.'ANOE. Various Piko. 115 This is Piko A in its simplest form, and in the second figure a variation is shown where the shuttle cord, after tj'ing at b, makes a different knot with a at c\ the details of which are shown at c' . All the knots at c are tied similarly to c\ and at c" the order of tj'ing is merely reversed. The return cord c knots with a at f as shown. Piko B (Fig. 112) is as in Piko A as far as c where the shuttle cord is knotted siniplj^ thus leaving a loop, through which the loop d is passed. Then the shuttle is run twice round the loop r, and back through the two loops thus made. This knot is the same as a fisherman's knot with a double turn. Piko B differs from the former in that an additional loop d is added to each knotted loop c. Piko C (Fig. 114). — A simpler form than the following piko. In this the cord is twice looped, and the free end /; passed under the shuttle cord a and around a and c. The loop d is left and the free end knotted around the base of the loop, thus fastening the ring. The shuttle is then sent around the ring aty^ when the process of the first knot at c is repeated reversed. Piko D (Fig. 116). — A double ring being made, the free end is passed over the shuttle cord, under the other ring cord and back over the shuttle cord, follow- ing which a half hitcb is made around the ring and the knot at b completed, Diag. i. Then, Diag. 2, three loops are made through the ring at c/, the shuttle return- ing at e and being sent around the loops and cord a three times and through the loops thus formed. The knot at //, Diag. 3, is thus made, and the cord/ carried to the next point ^^'^ on the ring, where the last described knot is repeated. Piko E (Figs. 117 and iiS). — In the samples examined, a double ring was made, with the free end «, Diag. i. Fig. 117, passing under the shuttle cord d and over the middle cord b. Then, Diag. 2, a is looped, passed through the ring at b and knotted simply to itself at c\ looped again and a fisherman's knot with two turns run around the part of the ring at b. This process results in the knot at //, Fig. 118. Then the shuttle cord d^ after tying at c'\ continues, a.s> f /] the work of the piko, which is complete when / joins the free end at / and begins the hanai asy. This piko in principle is ii6 Hawaiiati Nets and Nrtling. the same as the previous two, but the compression of the ring at the four points gives a ver^• different effect. The five foregoing piko were all made with the shuttle and attached to netted hanai. They were, too, more often used in the koko pualu than in the koko puupuu. Piko F (Figs. 119 to 122) was found to be the most common in the koko puu- puu, and this and the following styles were made from cord on a ball, not a shuttle, and were attached to the knitted hanai. Four methods of forming the ring in this piko are ilhistrated in Fig. 119. Diagram i shows the end cord a looped and tied with the ball cord, forming b. Then the ball cord is looped at g^ passed under /; at c and through the loop g at d ; or, a half hitch is formed and slipped over /;, following which the ball is passed ai'ound /; at i\ the loop / passed through c d and the ball carried round b to the next point. In Diagram 2 the cord is bent at b and the ball cord carried round the end cord a 2X c and the loop / made as before. In the next diagram, a I'l^-- '"'•'■ native with alamo or imoarixg stick. slip knot is tied to the end cord a at c/, whence the loops /j etc., are formed as usual. The loops of the piko in these three diagrams continue along b until the piko is of the requisite size, when the end cord a is passed through the end of the loop /;, knotted or twisted and concealed, as also in Diagram 4, in the first knot of the hanai. In Dia- gram 4 a double ring is made and the ball cord wound once around the end cord a at /;, and the loop / proceeded with as before; the cord a is left free to enlarge or decrease the ring during the work. The appearance of Piko F is shown, obverse side in Fig. 120, and reverse side in Fig. 121. In this piko and some of those following, the end cord a after being attached to b is sometimes wound around the bases of the outer loops / once or twice for ornamental effect. Fig. 122 shows Piko F with the end cord run twice around the loops. PoDiis of Piko. 117 PiKO G (Fig. 123) is shown with a quadnxpled cord forming the ring. After doxibling the cord twice, the ball cord is tied with a simple knot aronnd c and a at ^, and leaving a loop aty^ two half hitches are rnn aronnd or slipped over b and d as shown at g\ h and /. In the speci- men iignred the ring was joined by passing b and d throngh c and tying a to the latter. Then the loops b and d together were used for the first knot of the hanai. Piko H (Fig. 124). — After be- ing looped at b and tied at c, two half hitches are taken around the base of b leaving a loop at d. Then the ball cord is passed behind and around b ( seey" ) and the loop^^'- slipped through the loop thus made. At //, the last loop of the piko, an additional half hitch is slipped over b after the loop has been made. FIG. nu. PIKO A. HANAI ( Piko J (Figs. 125 and 126) is simpler than F and one step further on than H. In the specimen taken as an example and figured. Fig. 125, the piko could have been as easily made with one piece of cord, bj- run- ning )' on as a^ as with two. A de- tached cord is bent double at z. Then the ball cord is passed around X and y and over the free end «■, to begin the work, and back under x and )'. Following this a loop is made (seert'), twisted round and slipped through the loop b thus left, and the work continued with the ball cord c. In the example figured, the ring was joined by passing .r through z and tying with/, when x and r were concealed in the first hanai knot. Fig. 126 has the end cord wound once around the base of the loops as referred to in the paragraph on Piko F. rili. 111. riKO A. HANAI r. ii8 Hawaiian Nets and Netting. PiKO K (Fig. 127). — This is similar in constru(5lion to the last, but with the outer loops far apart, and the spaces on the ring between wound with the ball cord. In Diagram i the cord has been doubled, leaving a loop at b. Then the ball cord is looped at c and at «', the latter being passed around a and through r, and drawn taut. The ball cord is tightly wound around the bend of b for several turns and two loops formed, the second being passed under b and through the first. Diagram 2 shows the method of fastening the ring, which being done, the ball cord c and end cord a are com- bined to make the first knot of the hanai. This piko con- sists of from fifteen to thirty loops, and is generally very large in diameter. Piko L (Fig. 128).— A triple ring is made by doub- ling the end cord at b over the ball cord a and running the former twice around, Diag. i. Then the end cord c is wound round b once and the cords of the ring for about one-quarter its circumference, and is passed back and forth through the ring to form the loops o o, Diag. 2 ; after the last turn e this cord is wound four times around the bases of 00^ Diag. 3, passed through the windings w and around the cords of the ring at /and the winding on the ring continued at^'. It is probable that the piko in the specimen examined was made with a shuttle or a small ball of cord in addition to the large ball, for the amount of this end cord is considerable. Then the ball cord a is looped three times, the beuds being doubled over /; (making six loops at//) and bound with four half hitches by a to make the knot at /'. The base of one of the loops of / is pulled out at /, the ball passed through the bight and / closed by drawing on //. The end cord // returning from the circuit is then bound over k with half hitches and concealed in the hanai which the ball cord a commences with one of the loops of/. FH:. 11-' IKO It. IIANAT II. I Forms oj Hauai. 119 FIG. n:). HANAI D. FIG. 114, PIKO C. HANAI A. FIG. llo. HANAI I20 Hawaiian Nets and Netting. PiKO Vl (Fig. 129). — With the end cord a the slip knot a he is made. Then a loop d of the ball cord is slipped through c and the ball passed once around the base of c at c leaving a loop at/ Through /another loop^' is passed, the ball cord return- Flli. llr.. I'IKO D. ing aty', with which h is made similarly to d. The appearance of the piko completed is given from y' to X', and strongl}^ suggests crochet. Piko N (Figs. 130 and 131). — Now, if in Piko M, the cord c were wound around c twice or more, it would represent Piko N. To commence this piko, a slip knot is made, h c d, and the ball cord /wound round the end cord a at f from two to ten times as wanted; then a loop^ of the ball cord is pushed throxigh the windings, re- Piko and Ha iiai. 121 tvirning aty. This knot can be and probabl}' was made in a simpler wa}'. With the ball cordy looped through _^ at //, the loop / is made leaving a large slack at w and ;// then this slack is tightly wound around g and / the required number of times, the balance of the slack being taken in by drawing on /. Reference should be made to the knot in Hanai D in Fig. 137. Piko O (Figs. 132 and 133 ). — In Diagram i, after doubling the cord at^ make the loops /;, ^/ and y^ with the ball cord, leaving enoixgh slack. Then lay the ball cord e along the bends of the loops and wind the slack c a few times around the whole. Then loop the cord a g at h and continue winding with c. When the slack of c is FIG. 119. PIKO F. taken up by ^, the knot .i", Diag. 2, will serve to begin the piko, and the details of the regular knot are shown in the same diagram. X being completed, with the ball cord e pass the loop b' through d leaving slack at //', and over // laj* / and the loop d\ leaving another slack at c . Then c is bound around all the cords on a level with d^ and as the binding proceeds, the slack // is divided into two loops, //' //', Diag. 3, and held in place by c . Sometimes the ball cord c is not laid along d\ but is carried for- ward outside F. In such cases this piko differs from Piko N only in the additional loops // //' projecting from the side. Fig. 132 was drawn with the details in an upright position. Fig. 133 shows this piko, which in the specimen photographed was small. Hanai. — The hanai begins where the last loop of the piko was made. Hanai A (Figs. 114 and 118). — This is the simple fisherman's knot, known by natives as /v?, //w//', etc., as shown by g g in the figures. 122 Hawaiian Nets and Netting. Hanai B (Fig. 112) is the fisherman's knot with one or more extra windings around the engaged loop, as at g and g . In some of the koko puupuu the knots of both Hanai A and B are used as in Fig. 134. Fig. 135 is a specimen of koko puupuu with Hanai B. Hanai C (Figs, no, in and 136). — This is the square or reef knot, makili^ gg\ Fig. 1 10. Sometimes in beginning a large hanai, additional loops //, Fig. in, were run on to the loops of the piko for the purpose of enlarging the periphery of the hanai. Fig. 136 shows the speci- men from which Piko D, Fie. 116, was illustrated. This was a particularly well made and finished netting used as an cl;c. On completion of the piko the shuttle cord X' was looped at ;/ (without tying to the free end c?) and knotted to the piko loop p and so carried round the circuit to /. The cord / was then looped and the rows of the hanai com- pleted, the work travelling bous- trophedon. To join the ends of the rows, the shuttle cord ni is brought back from the outer edge of the hanai by knotting from side to side until tied with a to /• at /'. The three foregoing styles represent the netted hanai, which have the ends of the rows joined in the manner just described, or else by the free end a netted from side to side towards the outer edge. Hanai D (Figs. 137, 113 and 133). — The principle of this, the basal knot of all the knitted koko, has already- been partially illustrated in Piko M, N and O, but since it is so generally used, fuller details of its technique in the hanai are given in Fig. 137. The loops XX are those of the piko, andy^is the completed knot of Hanai D, known to the natives as pan. In Diagram i, the ball cord a is looped b through the next piko loop, and, Diagrams 2 and 3, the slack c being left, the loop d is bent and placed under b. Then c is wound round b^ d and the bight of .r and the slack taken Fill. ll'd. riKll F. oliVIJUSE. Forms oj Haiiai. 123 lip by d and c. In knitting the succeeding row of the hanai, the nearest loops of adja- cent pun of the first row are placed together and bound by the outer knots. Sometimes when the mesh of the hanai is intended to be large, two loops of the piko are combined in one pun. And occasionally, but only with the closelj- knitted koko of coir, the hanai commences with but few pun near the piko. Fig. 133, and the number is doubled by using each loop of the pun of the inner row as a base for that in the outer. In the figure the num- y ji ^ A ber of pun in the second, third and sixth rows have been suc- cessively doubled. The work of the knitted koko was direAed either con- tinually to the right, or to the left, or had the alternating rows in each direAion ; there seems to have been no set rule, but one circuit was always completed before the next began, in which respect it differed from that in the netting. The knitted knot can at once be detedled by the cord hanging from and connect- ing the bases of each pun (see a in Figs. 108 and 137). When one row was finished, the ball cord b^ //, Fig. 108, was carried up with the loop of the last knot to begin the row above. Hanai E (Figs. 138, 104 a^ and 105 a). — This is the same as Hanai D, except that the outer loops are interlaced before the ends are bound together with the pun. Hanai F (Figs. 139 and 115). — Before completing the pun, the ball cord c, Fig. 139, is looped _^- through the slack r, which is taken in as usual. Then round the bends of ^- the ball cord // is very tightly wound, finishing at /', Diagram 2. Then the bight of ^'' is placed on that of .r and the loop / on the ball cord run through .r and^, after which the pun is repeated. A koko of this st3'le, uuicpie in the colledlion, is illustrated in Fig. 115 in its suspended position. FIG. 121. I'IKO F. UEVEUSE. 124 Hazuatian Nets and Netting. Hanai G (Figs. 140, 141, 131 and 105 c). — Tliis hanai, in addition to the pun of Hanai D, has from one to three knobs construdled like the pnu, attached to the base. Fig. 105 c, shows this hanai in inverted position with one knob attached, and Fig. 131 the .same stj'le with the three knobs, in which case the name of the koko was koko punpiin liuilnii. Following the completion of the pun. Fig. 140, the ball cord e is looped.^'- aronnd the connedling cord a and the piui repeated (^, Diag. 3) the loops ,^ and j being drawn flush with the wonnd cord. This is followed by a similar pun, / and ;■, on the bends of -V. Diagram 4 shows the under side of the huihui. In Fig. 141 the pun (7, p and /' are all made around x and a. It has been observed that if the mesh of the koko is small, the rows of knots alternate with the styles of Hanai D and G. Hanai H (Figs. 142 to 144). — This hanai has from one to three pun looped around the puu of Hanai D. The method is similar to that in Figs. 140 and 141, but the loop j is allowed to project slightly from the puu o. In Fig. 142, the puu p is attached to the base of o and the loop ;/ allowed to protrude. Then ;/ is passed through 7, slipped over b and r;;g;!^ 1- y '^UBI "-y -^C^i^^^^^^y manshlp of which is not \^[^^^t^t^ff^-'* -S. :^f -it^ ^^^ill^MM Wk recognized as Hawaiian by any native who has seen it. It was in the late Hawaiian Government Museum, and came to the Bishop Museum as a Hawaiian koko with the rest of the collection, but the curator of the for- mer institution can give no information, and conse- quently none of its history is available. However, an old native claims to have seen some such nets in the time of Kamehameha V (1863-1872), when they were considered foreign. The material has the appearance of jute, dyed an indistinct blue-green on the surface, but there are sure signs that the cord was not made by machinery-. It differs from the Hawaiian koko in having been made from a number of cords, and these cords diverge from a piko filled in with matted string suggestive of the plaiting in straw hats (<•?, Fig. 149). Another point of difference is the four separated handles in which the cords of the hanai are incorpora- ted, each handle (/>) being a square braid of twelve strands. The technique. Fig. 150, is the same as observed in the simpler portions of macrame work, and also the netting or basket of fern stems around certain Japanese flower vases. Another bag. No. 4455, is FIG. 126. PIKO J. 128 Haivaiiaii Ncls and Netting. made in the shape of a koko with the hauai of black wool netted as in Fig. 150, and a kakai of heavy cord similar to that in No. 4454. The koko were divided into two classes b}' the Hawaiians: the koko puupuu, the property of the chiefly class, (?///, and the koko pnalu, nsed by the niakaaiitaiia or Fig. 127. piKO K. Fig. 128. riKO I.. plebeians. Mechanically there was no difference between the koko pualu and some of the koko puupuu, but those of the chiefs were always recognized as being better made, more elaborate and of superior cord. Koko Pualu. — The koko pualu was a plain netted bag, made of coir, liau or ahuawa. Figs. 104 b, 114. The term pualu was little used to designate this class. The material most generally employed in making the koko pualu was coir, and a coir koko I Koko Pimpiiu. 129 of the class pualu would be referred to merely as koko. To oue made of aliuawa the jx, name koko ahuawa was applied, while the chiefs' koko were never known other- wise than koko puupuu. The knots used in the strucflure of the koko pualu were the ka, and more rarely the makili. The mesh was always large and plain. A variation of this form with a smaller mesh was employed to carry such articles as sweet potatoes, taro, etc., without the aid of an umeke. It was made with a piko and hanai as in the koko pualu, but instead of the kakai as shown in Fig. 108, a cord was run through the outer selvage of the bag, thus serving to close the open- ing and act as a handle. This implement was known as an eke or laiilau. Koko Puupuu. — The koko puupuu FIG. 129. PIKO M. -was reserved for the sole use or service of the alii, who were exceedingly jealous of their dignity, promptly punishing by death any presumption on the part of a plebeian in making personal use of such a koko. One reason for which a makaainaua might carry the koko puupuu for his own needs, was protection against rob- bery by the kahu of some alii for his master. A commoner carrying food in a koko pualu was always liable to have it appropriated b}' any one of higher rank than himself, since the koko pualu indicated that it was only the belonging of a makaainana; but the subterfuge of utilizing a koko puupuu, implying that the carrier was in the service of a chief, would protect the '"•^- '3°- ''''^° ^• property unless the man were found out, when the consequences to himself would be Memoirs B. P. B. Museum. Vol. II, No. i.— g. I30 Hazuaiian Nets and Ncttins:. very disastrous. This koko took its name from aipiiiipiiii^^* — the steward of a person of high rank, deputed to carr}^ and care for the alii's food. The modern native gives the meaning of puupuu" in this conne<5lion as a complicated or thick knot, of which this koko is generally formed, but since some of the koko puupuu are of simple netting. Fig. 134, though of very fine mesh, it would seem that the derivation of the name came through the word aipuupuu, the explanation of which was contribu- ted by a very old native fisherman. The material in the koko puupuu is mainl}' waoke, but coir, olona, and in later days horsehair, were also used. Frequentl}' the piko and lower part of hanai are of coir, and the upper part and kakai of waoke, rarely olona, Fig. 104 c and c/, and 105 b^ known 3.spaitk/i. Occasionally two or more materials, generally coir and waoke alternate in rows, when the koko puupuu is termed onionio or paukukii. It is quite possible that horsehair suc- ceeded human hair, which was iised very extensivel}- in the neat braid of the niho palaoa and for decorating fan handles. The knots in the hanai were either knitted or netted, taking for the application of the word knitted, a slip knot, or one that when formed may be undone by drawing on the last end of the cord. The knitted knot is shown in detail in Fig. 137 and following. It might be here mentioned that enquiry among the older generation of living natives elicited almost no information concerning the manufacflure and use of koko, particularly the koko puupuu, as these articles have been out of use for many years. One old man explained that he had seen his grandfather make them, but that he himself had had no use for them so did not take the trouble to learn. However, he could make good fish nets ! There are a few natives in Honolulu who make koko for sale to tour- ^* Aipuupuu = callous-necked. These men (class) were always recognized by a large callus on the shoulder, caused by the auamo, and were very proud of the mark denoting their office. "Puu, puupuu, pu and pupu, in Hawaiian are very closely allied. Puupuu and puu mean, in short, "Any round protuberance belonging to a larger substance." (See Andrews' Hawaiian Dictionary.) Fig. 131. PIKO N, H.\NAI G. Antiqiti/y of Koko. 131 ists, but thej' unfortunatel}^ did not acquire the art by inheritance, merely having picked it up by unravelling some old specimen. It was particularly desired to learn the exact process of the native knitting and the names of such styles of hanai as E, F, H-N, and failing native sources, a thorough search among old voyages was carried out, with poor results. Only two works mention the existence of the koko. Freycinet wrote of the nettings for gourd bowls i^" "Apres le repas, on pose un de ces plats sur I'ouverture de la calebasse qui contient la poe, et le tout est surmonte d'un couvercle, qui n'est aussi qu'un morceau de cale- basse; le tout est enveloppe et assujetti dans un filet a larges mailles, qui sert a le suspendre." Fig. 132. PIKO o. Wilkes gave several illustrations of the methods of carr3ang, and remarked:^' " . . . . they [calabashes] are surround- ed by a net made of fine twine or sennit of the cocoanut." It might appear that the knitted knot, the pun, was a conception entirely ^''^- '^^- ""^^ °' "•^''•^' °- Hawaiian, for none of the other members of the Polynesian race seem to have pos- sessed such a knowledge. However, some doubt has occurred to the writer as to whether it was even native. The koko noticed in the narratives just mentioned, as a reference to the volumes will show, were koko pualu and of course netted, while some ■'"'Freycinet. Voyage autour du monde. Historique, Tome II, pp. 582 ; Atlas, pi. 86. ■"U. S. Expl. Exped. Narrative, Vol. IV, p. 96, plate opposite p. 55, and other illustrations. 132 Hcnvaiian Nets and Netting. of the koko pmipiui, Figs. 115, 147 and 148, were surely too remarkable to have escaped observation. All the older natives conversed with and enquired of claimed that the pun was very ancient — long antedated the ad- vent of Captain Cook. Still, too great reliability cannot be placed on these claims when the history of the Eskimo netting needle mentioned among the tools is considered. A comparison of the puu with the knot commonly known as the hangman's knot will show that the former is but a slight modification of the latter. Foreign sailors have closely associated with the natives since 1778, and the sailor with his knowledge of knots on the one hand and the native on the other eager to learn the foreigner's ways would make it a simple matter to intro- duce a new method into their work. A koko puu- puu of waoke in the Museum, No. 9050, and another of olona, found in the possession of a gentleman in Honolulu, were begun in the native way with Piko F, but had the hanai made of simple knitting. Fig. 151, and finished with a row of the puu. It is known that the natives were taught plain knitting by the missionaries, who came here in 1820, and proved Fig. 134. HAN.\i .\+B. Fig. 135. HANAI B. J^arioHs Hanai. 133 FIG. 136. HANAI C, PIKO D. FIG. 137. HANAI D. 134 Haivaiiati Nets and Netting. apt pupils. It ma^' be that the natives owe the knowledge of the pun to Juan Gaetano's Spaniards of about 1550, as with the shapes of their remarkable feather helmets. However, no matter what the origin, even if the knot were of foreign introduction, the natives had seen its adaptabilit}' to their work, and bj' their adoption of it into the manufaAure of their articles, they have surel}^ given it a domicile sufficientlj' Hawaiian. There are many at- tractive patterns in the koko puupuu, to which no native in these days is able to attach any significance or name. One old native, after being questioned in vain, remarked disgustedly: "The haole ( foreigners ) want all the time to put a number or a name on everj^- thing, but these to the natives were just koko." It was thoiight by the writer that the different grades of alii (which were multitudi- nous) might be entitled to distinct styles of koko, and without doubt, the greater the chief, the better was the koko. Certain chiefs were entitled to a particular form of tabu (or kapu) and in their presence, or when their food was carried by, the common people were required to fall on their faces, sit or kneel, according to the order of the tabu. It seemed quite feasible that certain of the koko might designate the rank of the owner by the pattern and so make it known to the people, but Dr. Alexander, a recognized authority of matters Hawaiian, stated that for the purpose of proclaiming the degree of chiefship, a crier was sent before the O-y-tJt^. FIG. I- Various Haiiai. 135 FIG. 139. HANAI F. FIG. 140. HANAI G. FIG. 141. HANAI G. 136 Hawaiia)! Nets and Ncitijig. FIG. 142. HANAI H. KIG. 143. HANAI H. VlO. 144. HANAI H+J. /// Old Tuucs. 137 aipuiipuu. An old fisherman claimed that by the creak of the koko on the unieke as the aipuupnu walked, the people conld tell when the koko of a chief were approaching. Tyerman and Bennet referred to this custom:''^ "So stately, too, was the royal etiquette, during his reign [Kamehameha I] that whoever happened to meet the king's calabash of water, as it was brought from the spring to the house, was required to un- robe, and lie down upon the earth, till the bearer of the vessel had gone b^-." Fig. 146. HANAI K. In Andrews' Hawaiian Didlionar}', a definition of the word viaololia occurs as: "The ancient name of the strings or net for a calabash, equivalent to the modern word koko." Natives were questioned about any knowledge of maoloha in these days, and one old fisherman said that koko makalii or koko maoloha existed before the time of Kamehameha the First, but were now no longer; that the name was now applied to the koko carried by the man in the moon. David Malo, in his account "Moolelo ■■-James Montgomery ; Journal of Tyerman and Bennet, Vol. II, p. 69. Boston, 1832. 138 Hawaiian Nets and Netting. Hawaii," ■" refers to the koko of Maoloha as being connec^ted with the ceremouies of the Makahiki festival. Following the end of the services, he says : "A net [koko] with large meshes was then made, which, being lifted b}' fonr men supporting it at fonr corners, was filled with all kinds of food, snch as taro, pota- toes, breadfrnit, bananas, cocoannts, and pork, after which the priests stood forth to pra}-. When the kahnna [priest] in his prayer littered the word hapai (lift) the men lifted the net and shook it back and forth, to make the food drop throngh the meshes, snch being the pnrpose of the ceremony. This was called the net of Maoloha. If the food did not drop from the net, the kahuna declared there would be a famine in the land ; but if it all fell out he predicated that the season would be fruitful." Dr. Emerson adds the following interesting note to his translation: — ''''Koko a Maoloha^ the net of Maoloha. The ex- pression is used Ke koko a Maoloha i ka lani. Tradition says that the first appear- ance of the Koko of Maoloha was in time of famine, when Waia was king on Hawaii. " Hawaiian Antiquities (Moolelo Hawaii) by David Malo, Translated from the Hawaiian by Dr. N. B. Emerson. Honolulu, 1903. p. 197. FIG. 147. H.\NAI I,. Koko a Makalii. 139 lu view of the famine that distressed the land, Waia, who was a kupjia^ possessed of siiperhnman powers, let down from heaven a net whose four corners pointed to the North, South, East and West, and which was filled with all sorts of food, animal and vegetable. This done he shook the net and the food was scattered over the laud for the benefit of the starving people." Also :^^ "Apropos of the net of Maololia^ at the time the net was filled with food and shaken, the following responsive service, called ka pule koko, the prayer of the net, was celebrated. The uet is lifted and the kahuna opens the service sa3'ing : — E uliuli kai, e Uli ke akua e ! E uli kai hakoko ! Koko lani e Uli ! Uli lau ka ai a ke akita. Pilio lani koko ; e lu — ! " Then the people respond : — 6. E 111 ka ai a ke akua ! 7. E lu ka lani ! 8. He kau ai keia. 9. E lu ka honua ! 10. He kau ai keia. 11. Ola ka aina ! 12. Ola ia Kane, 13. Kane ke akua ola. 14. Ola ia Kanaloa ! 15. Ke akua kupueu. 16. Ola na kanaka ! 17. Kane i ka wai ola, e ola! 18. Ola ke alii Makahiki ! 19. Aniania, ua noa. Kahuna: Noa ia wai ? People: Noa ia Kane. Oh deep-blue sea. Oh gfod Uli ! Oh blue of the wild, tossing sea ! Net of heaven, oh Uli. Green are the leaves of God's harvest fields. The net fills the heavens — Shake it ! Shake down the god's food ! Scatter it oh heaven ! A season of plenty this. Earth yield up thy plenty ! This is a season of food. Life to the land ! Life from Kane, Kane the god of life. Life from Kanaloa ! The wonder-working god. Life to the people ! Hail Kane of the water of life ! Hail Life to the king of the Makahiki ! Amama. It is free. Free through whom ? Free through Kane. " Then the kahunas stand up holding their hands aloft, and the people exclaim: 'Ua noa. Ua noa. Ua noa.' At the same time holding up the left hand, and at the utterance of each sentence, striking with the right hand under the left arm-pit. "When the kahuna utters the words 'iT hC — in the 5th line — those who are lift- ing the net shake it and make its contents fall to the ground." A legend called the "Koko a Makalii" was narrated to the writer by a Hawaiian repository for such and is here repeated in brief: "Kane, the highest god in the Hawaiian pantheon, had conne<5lion with a woman on earth and became the father of Makalii. During a time of severe drought, the people made prayers and offerings to Kane for rain, which were disregarded by the deity. Then Makalii, grieving for the "Ibid, p. 204-6. I 140 Haivaiian Nets and Netting. fm f ^ ' ^ ■- ^ ■ tf >^, '.< > ,' :■ /• ^ <:. ? starving people, declared that lie would go to his father and demand rain. Provided with koko of exceeding fine mesh" he approached the Polynesian Jove, who welcomed his son and filled the koko with w a t e r. W h e n Makalii returned to earth the water dropped through the meshes of the koko and fell as rain." It is probable that the two ac- counts refer to the same thing, as the pule koko recorded by Dr. Emerson would fit the Koko a Makalii, if anj-- tliing, better than the Koko a Mao- lolia. As far as known, no such net as that used in the ceremonies has ever been pre- served. There may have been a service at the end of the makahiki festival somewhat similar to that described by Malo in commemoration of Makalii's dar- ing and beneficial act, but Malo's account is so nearly parallel with that of Peter's vision'*'' that the two must have been confused. At the time of Malo's conversion to '^ Maka = mesh, and Hi = small. '"'Acts, 10; 11, 12. 1 FIG. 148. HANAI M. List of Koko. 141 the Christian religion he was considered exceptionally well versed in native traditions, but following the change he became an ardent but narrow Christian, unable to distinguish between good or bad in the native lore, and condemning the whole as evil. While his mind over- flowed with biblical accounts, he wrote his "Moolelo Hawaii," so that it would be quite natural for a bible story to occasionally creep into his relations. In the following lists of koko in the Museum, the measurements are given in inches. The length men- tioned is only that of the hanai and piko combined, the kakai being about as long again. The pu of the kakai is no unless specified otherwise. A comparison of the size of the knot or pun with the mesh will give an idea as to the closeness of some of the netting, and it is quite probable that those koko with the very fine mesh were used as eke, as such are generally accompanied with a piko, the ring of which is very small. Koko Pualu. 4423. Ahuawa; Piko A, Hanai C; length 21, mesh 2.2; kakai, pu lino; heavy cord. 4424. Coir; Piko A, Hanai C; length 19, mesh 3.4; rope. 4425. Ahuawa; Piko D (small), Hanai C; length 22, mesh 2.9; kakai, pu lino; heavy cord. Piko E, Hanai A; length 30, mesh 9; kakai, pu lino; rope. Piko A (small, diam. i), Hanai C; length 16, mesh 4.S; heavy cord. ?IG. 149 4426. 4427. 4428. 4429. 4430- 4431- 4432. 4433- 4434- 4435- 4436- 4437- 4438- 4439- 4440. 4441. 4442. Coir: Coir Coir Coir Coir Coir Coir Coir Coir Coir Coir Coir Coir Han Coir Coir Coir Piko A, Hanai C Piko E, Hanai A Piko E, Hanai A Piko E, Hanai A Piko E, Hanai A Piko E, Hanai A Piko D, Hanai A Piko D, Hanai A Piko D, Hanai A Piko E, Hanai A Piko E, Hanai A Piko C, Hanai A Piko D, Hanai A Piko D, Hanai A Piko C, Hanai A length 30, mesh 8; heavy cord. length 14, mesh 5; kakai, pu lino. length 17, mesh 5.5; kakai, pu lino. length 18, mesh 5.5; kakai, pu lino. length 14.5, mesh 5.5; kakai, pu lino. length 19, mesh 6.6. length 12, mesh 3.1. length 12.5, mesh 3.3. length 13, mesh 3.2. length 27, mesh 9.5; no kakai; heavy cord. length 18, mesh 5; kakai without pu. length 16, mesh 7.3. length 18.5, mesh 5.7. length 18, mesh 5.2. length 13, mesh 4.3; heavy cord. Fig. 114. 142 Hawaiian Nets and Netting. 4448. Oloiia; Piko C, Hanai A; length 20.5, mesh 7.5. 4449. Waoke; Piko A, Hanai A; length 12, mesh 5; kakai, pu lino. 4450. Waoke; Piko C, Hanai A; length 7.5, mesh 1.6. 4452. Cotton cord; Piko D, Hanai A; length 6.5, mesh 4. 4453. Waoke; Piko D, Hanai C; length 12.5, mesh 3.5. 4463. Coir; Piko B, Hanai A; length 12, mesh 1.2; no kakai. 6S59. Coir; Piko C, Hanai A; length 17, mesh 5.6. 9017. Han; Piko E, Hanai A; length 8, mesh 4.5. 9018. Coir; Piko A, Hanai C; length 15, mesh 4.5. 9019. Coir; Piko C, Hanai C; length 23, mesh 8. 9020. Coir; Piko A, Hanai C; length 29, mesh 7.5; heavy cord. FIG. 150. Fig. 151. H.\NAI OF PL.\IN KNITTING. 9021 9022. 9023, 9024. 9025, 9026. 9027, 9028. 9029. 9049. L 196. L 197' L 198. L 394. L395' L396. Coir; Piko D, Hanai C Coir; Piko E, Hanai A Coir; Piko E, Hanai A Coir; Piko D, Hanai A Coir; Piko D, Hanai A Coir; Piko E, Hanai A Coir; Piko D, Hanai A Coir; Piko E, Hanai A Coir; Piko D, Hanai A Coir; Piko E, Hanai A length 14, mesh 4.5. length 30, mesh 10.2-11.8; heavy cord. length 22, mesh 9.2. length 17.5, mesh 4.5. length 17, mesh 3.8. length 16.5, mesh 6.5; kakai, pn lino. length 14, mesh 3. length 15.5, mesh 6.2; kakai, pu lino. length 14, mesh 4.3. length 20, mesh 6; heavy cord. Hau; Piko E, Hanai A; length 11, mesh 6. Coir; Piko D, Hanai C; length 23, mesh 12. Coir; Piko E, Hanai A; length 16, mesli 11.5; kakai, pu lino. Coir; Piko E, Hanai A; length 10.5, mesh 6.7. Coir; Piko D, Hanai C; length 13, mesh 13. Coir; Piko E, Hanai A; length 14.5, mesh 9. List of Koko. 143 3070. 4353- 4354- Koko Puupuu. Ouionio, coir and horsehair; Piko N, Hanai D; length 16.5, mesh 3.2; pun I inch long. 3071. Onionio, coir and horsehair; Piko N, Hanai D; length 16, mesh 3, pun i. 4346. Waoke; Piko J roughly made, Ha- nai G with 2 puu; length 17.5, mesh 3.2, pun 1.2. 4347. Pauku, coir and waoke; Piko F, Hanai D; length 13, mesh 2.3, pun .5; no kakai. 434S. Pauku, coir and oloua ; Piko D (small, .6 diam.), Hanai B+D; length 24 ( i6-[-S ), mesh no kakai. 4349. Waoke; Piko N (27 in circumfer- ence), Hanai H with 3 puu; length 16.5, mesh 6, puu 1.8. 4350. Waoke; Piko F, Fig. 122, Hanai D; length 13, mesh 1.8, puu .8. 4351. Waoke; Piko N, Hanai G with 3 puu. Fig. 131; length 15, mesh 4.5, puu .7. 4352. Waoke; Piko N, Hanai G with 3 puu; length 12, mesh 3.5, puu .5; no kakai. Waoke; Piko N (24 in circ), Hanai G with 3 puu; length 12, mesh 2.1, pun .6. Waoke; Piko N (large), Hanai G with 3 puu; length 11.5, mesh 1.5, puu .6. :FIG. 152. SAMPLES OF KOKO. 144 Hawaiian Nets and Netting. 4355- Waoke; Piko H, Hauai G with i puu; length i6, mesh 2, pun .6. Fig. 105 c; koko inverted. 4356. Olond; Piko N, Hanai H-|-J, Fig. 144; length 15.5, mesh 5.7, puu .6. Olona; Piko N, Hanai M; length 14.2. Fig. 148. 4357- 4358. 4359- 4360. 4361. Olona; Piko F, Hanai L; length 15, mesh .8, puu .4. Fig. 147. Pauku, coir and waoke; Piko C, Hanai B; length 15, mesh .6, knot .4; no kakai. Pauku, coir and waoke; Piko D (.6diam.), Hanai B; length 21.5, mesh .4, knot .2. Pauku, coir and waoke; Piko O, Hanai D; length 9, mesh .8, puu .7; damaged, no kakai. 4362. Pauku, coir and waoke; Piko O, Hanai D; length 12, mesh .5-I-1.5, puu .5-)-. 6. 4363. Olona; Piko N (27 in circumference), Hanai K; length 21, mesh 2.8. Fig. 105 d; koko inverted. 4364. Pauku, coir and waoke; Piko F, Hanai D + H, Fig. 143; length 14, mesh 2.5, puu .6; no kakai. 4365. Pauku, coir and waoke; Piko A, Hanai A-|-B; length 13, mesh .3-I-1; kakai with i pu uo bound together at the two ends. Fig. 134. 4366. Pauku, coir and waoke; PikoO (i diam., Fig. 133), Hanai D; length 20, mesh .7-2.3, puu .6-1.8; no kakai. 4367. Pauku, coir and waoke; Piko F (iS in circumference), Hanai D; length 14, mesh 1-1.4, puu .6-.S; no kakai. FIG. 153. END OF AUAMO SHOWING PU UO. 4368 4369 4370 4371 4372 4373 4374 4375 4376 4377 4378 Pauku, coir and waoke; Pauku, coir and waoke; Pauku, coir and waoke; Pauku, coir and waoke; ; Piko O, Hanai D; length 17, mesh .6-2.4, puu •5"~i4- ; Piko K, Hanai D; length 11.5, mesh 2.5, puu .6. ; Piko N, Hanai D; length 21.5, mesh 3.3, puu .4-.8. ; Piko F, Hanai D; length 15.5, mesh 2.5-3.2, puu .6-.8. Pauku, coir and waoke; Piko F, Hanai D; length 13.5, mesh 3, puu .9-1.4; no kakai. Pauku, coir and waoke; Piko N, Hanai D; length 20.5, mesh 2.3, puu .6. Pauku, coir and waoke; Piko N, Hanai D; length 25, mesh 3.4, puu .7. Pauku, coir and waoke; Piko N, Hanai D; length 12.2; mesh 2.4, puu .7. Pauku, coir and waoke; Piko F, Hanai D; length 11, mesh 2.1, puu .4; no kakai. Pauku, coir and waoke; Piko N, Hanai D; length 15, mesh 3.1, puii .7; no kakai. Pauku, coir and waoke; Piko J, Hanai D; length 15, mesh 2.9, puu .8; no kakai. List of Koko. 145 4379. Paukn, coir and waoke; Piko J (21 in circumference), Hanai D; length 13, mesh 4, pun .8. 4380. Paukvi, coir and waoke; Piko F, Hanai D; length 17, mesh 2.4, pun .8; no kakai. 4381. Pauku, coir and waoke; Piko N, Hanai D; length 18, mesh 3, puu .7; no kakai. 4382. Pauku, coir and waoke; Piko N, Hanai D; length 13, mesh 2.8, puu .7. 4383. Pauku, coir and waoke; Piko N, Hanai D; length 12, mesh 2.7, puu .7. 4384. Paiiku, coir and waoke; Piko N, Hanai D; length 15.5, mesh 3.2, puu .8. 4385. Onionio, coir and waoke; Piko F, Hanai D; length 14, mesh 2.1, puu .8. 4386. Olona; Piko F, Hanai D; length 22, mesh 5-6.5, puu .9-1.1. 4387. Waoke; Piko F, Hanai D; length 20, mesh 1.8-2.5, P^^^^^ •6- 4388. Waoke; Piko M, Hanai D; length 9.6, mesh 1.3, puu .4. 43S9. Olona; Piko G, Hanai D; length 11, mesh 2.3, puu .8. 4390. Waoke; Piko F, Hanai D; length 14.5, mesh 2, puu .5. 4391. Waoke; Piko F, Hanai D; length 10.5, mesh 2, puu .6. 4392. Waoke; Piko F (i in diam.), Hanai D; length 15.5, mesh 2, puu .9; no kakai. 4393. Waoke; Piko N, Hanai D; length 10.5, mesh 1.8, puu .8. 4394. Olona; Piko G, Hanai D; length 17, mesh 1.6, puu 7. 4395. Waoke; Piko K, Hanai D; length 10, mesh 1.7, puu .5. 4397. Olona; Piko F, Hanai D; length 13, mesh 3.1, puu 1.3. 4398. Waoke; Piko F (1.2 diam.), Hanai D; length 10.5, mesh 2.3, puu .9. 4399. Waoke; Piko F, Hanai D; length 10, mesh 1.9, puu .9; no kakai. 4400. Waoke; Piko H (.6 in diam.), Hanai D; length 15.5, mesh 3.5, puu .7; no kakai. 4401. Waoke; Piko F, Hanai D; length 14.5, mesh 4.6, puu .6-. 8. 4402. Waoke; Piko A (.8 in diam.), Hanai D; length 14.5, mesh 3, puu .6. 4403. Olona hanai and waoke kakai; Piko F, Hanai D; length 14, mesh 2.8, puu i.i. 4404. Olona; Piko F, Hanai D; length 8.5, mesh 2.7, puu .4. 4405. Olona; Piko G, Hanai D; length 12.5, mesh 2.6, puu .8. 4406. Waoke; Piko K, Hanai D; length 12, mesh 2.1, puu .7. 4407. Waoke; Piko K, Hanai D; length 11, mesh 2.1, puu .7. 4408. Olona; Piko F, Hanai D; length 23, mesh 8, puu .8. 4409. Waoke; Piko F, Hanai D; length 17, mesh 3.7-4, puu 1.5-2.2 long and thin; no kakai. 4410. Onionio, coir and waoke; Piko H, Hanai D; length 12, mesh i, puxi .7; no kakai. 441 1. Onionio, coir and waoke; Piko N, Hanai D; length 17.5, mesh 3, puu .6-.7; no kakai. 4413. Pauku, coir and waoke; Piko N, Hanai D; length 9, mesh 2.8, puu .5. 4414. Onionio, coir and waoke; Piko N, Hanai D; length 13.5, mesh 2.8, puu .7. 4415. Pauku, coir and waoke; Piko N, Hanai D; length 14, mesh 3, puu .9. 4416. Onionio,coir, waoke and horsehair; Piko N, Hanai D; length 15.5, mesh 2.7, puu. 8. 4417. Pauku, coir and waoke; Piko N, Hanai D; length 15.5, mesh 2.9, puu .8. 4418. Onionio, coir, waoke and horsehair; Piko F, Hanai D; length 11, mesh 2.1, puu .6; no kakai. 4419. Waoke; Piko F, Hanai D; length 9.2, mesh 2.3, puu .7. Memoirs B. P. B. Museum. Vol. II, No. i.— lo. 146 Haiuaiian Nets and A^cttii/g. 4420 4421 4444 Oiiioiiio, coir and waoke; Piko E, Hanai B; length 14.5, mesh 3. Pauku, coir and waoke; Piko A, Hanai B; length 10.5, mesh 2.3. Pankn, coir and waoke; Piko D (?), Hanai B; length 17, mesh 4.3, knot .5; modern looking. 4445. Pauku, coir and waoke; Piko A, Hanai A; length 19; mesh 3. 4446. Pauku, coir and waoke; Piko D, Hanai B; length 16.5; mesh 5.5. 4447. Pauku, coir and waoke; Piko B, Hanai B; length 19, mesh 4.7, knot .4. 4456. Waoke; Piko F, Hanai E; length 21, mesh 3.5-5, puu .5. Fig. 105 a. 4457. Pauku, coir and waoke; Piko N, Hanai D; length 12, mesh 2.7, puu .8. 445S. Pauku, coir and waoke; Piko N, Hanai D; length 17, mesh 2.5, puu .7. 4459. Waoke; Piko N, Hanai D; length 10, mesh 2.4, puu .5. 4460. Pauku, coir and waoke; Piko N, Hanai D; length 13, mesh 2.5, puu .7. 4461. Waoke; Piko J (Fig. 126), Hanai D; length 10, mesli 2.4, puu .7. 4462. Pauku, coir and waoke; Piko F, Hanai D; length 9.5, mesh 2, puu .5. 4464. Waoke; Piko F, Hanai D; length 12, mesh 1.4-2.3, puu .9. F'igs. 120, 121. 4465. Waoke; Piko F, Hanai D; length 13.5, mesh 1.7, puu .9. 5324. Coir; Piko O, Hanai D; length 8.4, mesh 1.1-1.3, puu .6-1; fragment, lower two-thirds onl}-. 6857. Waoke; Piko F, Hanai E; length 20, mesh 4.2, puu .4. Fig. 104 a. 6S58. Waoke; Piko F, Hanai D; length 8.8, mesh 2.2, puu .3. 7703. Waoke; Piko L, Hanai F; length 15, mesh 3, puu .6. Fig. 115. 7706. Pauku, coir and waoke; Piko F (.6 diam.), Hanai B; length 16, mesh .8. 7707. Waoke; Piko F, Hanai D; length 15, mesh 1.7, puu .6; no kakai. 770S. Pauku, coir and waoke; Piko D (.4 diara.), Hanai B; length 21, mesh .8; no kakai. 7748. Pauku, coir and waoke; Piko F ( i diam.), Hanai D; length 15, mesh 2.3, puu .6. 7966. Waoke; Piko N, Hanai D; length 15, mesh 2.5, puu .8. 9013. Waoke; Piko K, Hanai D; length 10.5, mesh 2, puu .8. 9014. Pauku, coir and waoke; Piko N, Hanai D; length 15, mesh 2, puu i. 9015. Onionio, coir, waoke and horsehair; Piko N, Hanai D; length 17, mesh 3, puu i. 9016. Waoke; Piko N, Hanai G with 3 puu; length 13, mesh 3, puu .4. 9050. Waoke; Piko F, hanai of plain knitting; length 6.9, mesh .3. -J- ( Pauku, coir and waoke; Piko O, Hanai D; length 10.5, mesh .5-1.5, puu .4-1.1. "" ( Olona; Piko F (large), hanai of plain knitting (Fig. 151); length 12.5, mesh .4. L 194. Pauku, coir and waoke; Piko D (.3 diam.), Hanai B; length 15, mesh .5, knot .3; kakai in four parts. E 195. Paukix, coir and waoke; Piko N, Hanai D; length 17, mesh 3, puu i. L 397. Coir; Piko F, Hauai D; length 9, mesh 2.5, puu .8; fragment, no kakai. KoKo Eke. 4422. Cotton cord; Piko (?), hanai, foreign knot, probably Chinese; length 20, mesh 4.2. 4451. Waoke; Piko D, Hauai C; length 12.7, mesh 3.5; kakai, a double draw string. Fig. 136. 9051. Waoke; Piko D, Hanai C; length 11.8, mesh 5; kakai, draw string. 4454- 4455- Alia Hcnvelc. Miscellaneous. Foreign koko; length 37.5, mesh 2.2. Figs. 149 and 150. Foreign koko; black woolen braid; knot, Fig. 150; length 20, mesh 2.6. 147 Aha. — Of the aha^^ or cords snrronnding and permanently fastened to gonrd water bottles (hnewai, olowai, etc.) the most common was that known as haiuele — sometimes referred to as koko liawele — and shown in Fig. 106 enclosing a hnewai. There is in the colleftion a great nnniber of drinking and other gonrd vessels withont cordings which wonld have been carried in koko punpnn or pnaln. The hnewai was a drinking gonrd of large body and narrow neck for general rise. The variety of forms of gonrd vessels was very great, and was generally the resnlt of manip- nlation when the frnit was green. The word hawele means a tj-ing or binding on in which the joinings or loops are never knotted, and the koko or aha hawele may be so recognized. Details of teclmiqne of the aha hawele in Fig. 106 will be seen in Fig. 154, where, after a loop, /; ni c, is made ronnd the neck of the bottle and tied at c?, the cording continues in the direction indicated. There is a slight error, for which the writer is re- sponsible, in the drawing, for the cord f^ instead of encircling b and e, shonld pass nnder and over FIG. 154. DKTAiL 01^ .4 HA HAWEi,E. ^, ^nd d and under itself. A bail is then made by carrj-ing^ over the top of the bottle, over ;;/, under //, over n and back around a and d. When there are four or five thicknesses of cord in the bail, it is bound at several points with half hitches by the end of .^, the latter being finally fastened to the wooden or shell stopper. The material used was coir, spun or braided — seldom olona or hau. A similar lashing, but more complex, was noticed on a gourd water bottle. No. 193 1, from New Caledonia. A very neat fastening of the aha hawele has been made around a hnewai pueo (Fig. 155) which is a water gourd shaped like an hour glass. In the figure, when the aha hawele, as' shown in Figs. 106 and 154, was attached to the lower bulb, to its upper cords another binding was added for the purpose- of enclosing the other part. However, in a great number of the pueo with hawele the lower bulb is alone corded ■"Aha should be applied only to (i) coir cord, (2) cord of human liair, (3) strings made from intestines. (See Andrews' Dictionary.) Other cords, such as olona, should be termed a/w. MS Hawaiian Nets and Ncltitig. and the bail fastened to the waist. A plain and chaste style of aha hawele was found on an old broken hucwai in the colleftion Fig. 156, the cord of which was more neatly braided than on any of the other specimens. Another form, rarely used, is shown in Fig. 157. A form of water gourd used in the canoes was called olowai, Fig. 107. The writer has not been able to learn any specific name for the aha on this gourd, and since only two (Nos. 3877 and 38S0) out of the five specimens in the colledlion have aha exadlly similar, the probabilities are that each individual followed his own taste in cording. In the aha figured, Fig. 107, the work was done hy making two half hitches on the bottom of the gourd, and on the rings thus made four loops of single cord were formed large enough to reach the middle. Tlie cord was then brought to the neck of the bottle and fastened b}' two half hitches. A set of four loops was attached to these rings, at the same time passing in turn through the bights of the previous set, thus drawing the lashings tightly around the gourd. The cord was then made to follow the cords of each loop and an eye was formed by tying at each bight. The double ring at the middle was added last. For the' bail, suspended over the side of the bottle, a separate cord was used. Another aha, on ^^"^^ 'ss- hiewai puho with aha hawele. specimen No. 3995, is similar to that in Fig. 107, except for the double ring in the middle. Of the other specimens of corded olowai, one. No. 3879, is enclosed in an aha hawele, and the other. No. 3881, in a large-meshed netting. Ipii Ic^i^ ipn ho/o/iolona, poJio alio and ipu alio are among the names used to desig- nate a utensil, consisting of two pieces, for containing fish hooks and lines. There are two general forms of this article: one with the lower and smaller part of wood covered with a larger gourd. Figs. 158 and 159, and the other of gourd with the lower larger than the upper part. Fig. 160. There is some confusion now as to the correct names and uses of the different styles, but the best information seems to be that the former, called ipu le'i or ipu holoholona, was for the purpose of holding bait in addition to hooks and lines, and the latter, polio alio, for the fishing tools alone. Some of the IpH Lei. 149 polio alio were composed of bowl-shaped gourds, and others of long narrow gourds, covered with half a small gourd or coconut. Similar cordings enclosed both st3-les, and the most finished in appearance is tliat around the ipu le'i in Fig. 158. This cording, called koko, or koko ipu le'i, is a combination of the aha liawele firmly fastened around the wooden part, to the upper cords of which a netting is attached to en- close the cover. The netting is closed around the cover by a draw string, which also serves to carr}- the utensil. In a few specimens the aha liawele has been dispensed with, the netting being fastened through holes bored in the upper edge of the lower part. Fig. 159. This last method was also used on the Iiiiiai pocpoc, which have been treated by Dr. Brigham in the previous portion of this memoir. An- other and sim- pler method of attaching cords to these articles is shown in Fig. 160; here the ^'''- '56- F'g. 157. HURWAI WITH AHA HAWELE. edge of the lower part has been pierced at two opposite points, from which two cords were led upwards through holes in the cover. Before passing from the gourd cordings, it might be in order to mention the several means of securing handles to gourds in vogue among the natives. The sim- plest form noticed is a coir or hau cord around the neck of the huewai, when the mouth of the gourd bulges. Fig. i6ia\ some of the cords were roughly' made and tied, but in others the cord has been braided and then attached by a double half hitch. A few of the huewai pueo have also been treated in this manner, Fig. 161b. When the huewai was not pueo or bulged at the mouth, then a handle was sometimes attached by making a hole at the base of the neck and through it drawing and knotting both ends of a loop of coir or hau cord or braid. Fig. i6\d, or, by boring the edge of the mouth at two I50 Hawaiian Nets and Netting. points opposite, through Avhich the ends of a haii or olona line were passed and knotted. Fig. i6i c. In some of the bowl-shaped gourds, eight holes were pierced in pairs and each end of two cords passed through two of the holes and knotted, Fig. 162. Handles, gen- erally more than two, were so attached to many of the hinai poepoe, and icie and pandanus baskets. FIG. 1 58. IPU LEI. FIG. 159. IPU l.EI. List of Gourd Vessels, with Cordinos Attached. 1093. Huewai; aha hawele, coir. 1094. Huewai; aha hawele, coir. 1099. Huewai, aha hawele, coir. 1 100. Huewai; aha hawele, coir. iioi. Huewai; aha hawele, coir, Fig. 156. 1 102. Hviewai; aha hawele, cotton cord. 1 104. Huewai; aha hawele, coir. 1 107. Huewai; aha hawele, coir. 1 1 12. Huewai pueo; aha hawele, coir. Fig- 155- _ 1 1 13. Huewai pueo; aha hawele, coir. Huewai pueo; aha hawele, coir. 1114 1121 1122 1124 1128 1129 Huewai Huewai Huewai Huewai Huew-ai aha hawele, coir, aha hawele, coir, aha hawele, coir, aha hawele, coir, aha hawele, coir, Fig. 106. List of Corded Ipii. 151 1 133. Huewai; aha liawele, oloua. 1 137. Huewai; alia hawele, coir. 1 26 1. Huewai pueo; aha hawele, coir. 3934. Huewai; aha hawele, hau, Fig. 157. 7667. Huewai; oloua. Fig. 161^/. 766S. Huewai; coir, Fig. 161^7. 3877. Olowai; coir. 3879. Olowai; aha hawele, coir. a b c d Fig. 161. HUEWAI WITH CORD H.\NDI.ES. PIG. 160. rOHO AHO WITH CORD. 3935 Huewai; 393S Huewai; 3940 Huewai; 3996 Huewai 7750 Huewai; 1097 Huewai; 1 108 Huewai; 1 109 Huewai; mi Huewai 3942 Huewai; 3944 Huewai; aha hawele, coir. aha hawele, hau. aha hawele, coir. pueo; aha hawele, coir. aha hawele, hau. coir, Fig. 161 a. coir. Fig. 161 rt'. hau. Fig. 161 c. pueo; coir. Fig. 161 /;. hau. Fig. 161 a. coir, Fig. 161 c/. Fig. 162. GOURD UMEKE WITH CORD. 3880. Olowai; coir, Fig. 107. 3881. Olowai; coir, netted. 3995- Olowai; coir. 3951- Umeke; oloua. Fig. 162 3952. Umeke; oloua. 152 Hawaiian Nets and Netting. Ipu Le'i. 3850. Oloiia and coir. 3S54 Coir. 3858. Olona 3851. Hau, coir and oloiia. 3S55 Wooden part onl3-, 3859- Olona 3S52. Olond, cords of lia- ] jierced. 3860. Hau. wele doubled. 3856 Coir, Fig. 158. 3994- Coir. 3853- Olona. 3857 Coir, Fig. 159. PoHO Aho. 3861. Olona. 3873 Olona. 3953- Olona 3869. Olona. 3874 Olona and coir. 5027. Coir. 3870. Olona. 3875 Olona, Fig. 160. 6373- Coir. 3871. Olona. 3876 Olona. 3872. Olona. 3927 Coir. Nets, Upena. — As the second division of net work, fish nets and similar fabrics for different uses come up for consideration. The material most preferred and used, except where noted below, was olona spun into cord varying greatlv in thickness. The method of beginning a net is interesting, and has been shown to the writer by fishermen on Oahu and Molokai. The native, having filled his shuttle from the ball of twine without severing the line, takes another cord. Fig. 163 aa^ the ends of which he ties together. Sitting on the ground with feet far apart, he inserts the first toes into the ring and stretches it tightly-. The ball is passed under and over a a three times towards the right, and the fri(?lion on the cords is sufficient to keep the line c taut while closing the knots. Then the shuttle is passed around the gauge, over a «, under ^, and c is drawn down by the shuttle cord to the gauge where the knot at h is made. To begin the second and alternate knots the shuttle would of course pass under a a and over c. Cord for the continuance of the loops is drawn from the ball by loosening the windings around a a. When the knot (called ka, umii, etc.) slips properh' into place, as at b and c/, the worker ejaculates kakiokolie with satisfa(R;ion, but if it misses the loop, as at f, the word omauokole is uttered with disgusted tone. Sometimes, as the work proceeds, two loops are by mistake enclosed by one knot, the name for which is maiiac. In this division come the nac^ or netting which constitutes the groundwork of feather garments,'*'* netted inalo^ bird net and the regular fish netting. Nae. — Nae is a netting of ver}^ fine mesh, varying in the Museum feather garments from .05 to .27 inches. The twine is not as fine as the size of the mesh might siiggest, averaging .025 inch in diameter, so that in some of the specimens the *'E. P. B. Museum; Memoirs, vol. i, nos. i and 5. Nac oj Feather Garments. 153 netting is as close as loose burlap.'''' As stated by Dr. Brigliani, in Feather Work, the bases of the cloaks are composed of several pieces of nae, each being cut to iit. The pieces are not al\va3-s of the same mesh, that of those in the cloak of Kiwalao, for instance, varying from .1 to .25 inch. However, the mesh of the majorit}- varies but slightl}' in each individual. The closeness of the mesh in some cloaks and capes has suggested research for a suitable shuttle to do the work, but be3'ond the niao, alread}' described, and needle of kauila wood, no implement has been found. In a piece of nae, specimen No. 2S40' ", prepared for the addition of feathers, the mesh increases from .05 at the top to .08 inch at the bottom and is too fine to allow even an ordinarj^ pinhead to j^ass through readily. The specimen is 37.5 inches wide and 11.5 inches long, the edges of which have been trimmed at the knot without leaving the usual netting selvage. There are from four to ten row^s of netting continu- ousl}' in both diredlions, while of course in netting; the rows usually alternate. This nae was made without a shuttle with lengths of twine drawn through the loops, and for such purposes the natives dipped the ends of the twine into the highly saccharine juice of the ki root, which, dr3'ing rapidly, made the fibre stiff enough to be threaded through the holes. It is not the intention of the writer to give the impression that all nae have a number of rows of netting continuously running in the same direcftion, for in most of the speci- mens the rows alternate regularl}'. The nae with a mesh of .25 inch could be and prob- ably was made with a shuttle such as that shown in the middle of Fig. loi, or a niao. In the colle6lion is a piece of unfinished netting of rather fine mesh, the histor}- of which is unknown. The upper part is 41 inches wide with a mesh of .25 inch in- creasing to .44 inch in the lower, and it has, as far as completed, the shape of a feather cloak. The first half has been made with four shuttles, one following the other, and possibly by four people, and the unfinished half with three shuttles as the three loose strings on the bottom indicate. The specimen is Hawaiian and is probabl}- the result of an incompleted later day attempt to prepare nae for a feather cloak. Sometimes in ■"Ibid, pi. ix, upper figure, left side. 5° Ibid. FIG. 163 CUMMENCEMENT OF A NET. 154 Hawaiian Nets and Netting. making very broad fish nets, or when there was need for rapid niaunfadlure, two or three men would net along the same side, following each other. The cord, of the first man to finish, was knotted to the ends of the succeeding rows when completed and was the first to begin the next series. Netted Malo. — Of netted malo there are two specimens in the Museum, each with a mesh of about .25 inch. One, No. 2842, is a plain piece of netting 15.3 ft. long and 7 in. wide.*' The malo, or loin cloth, of olona netting was alwaj'S an alii's gar- ment and was worn hy him on canoe voyages. The other specimen,'" No. 6921, 12.2 ft. long and 5.7 in. wide, came to the Museum from the Provisional Government of Hawaii after the revolution in 1893, as a relic of royalt}-, with the royal feather robes. It has been the nae of a feather malo, and in sewing on the tufts of feathers, the meshes have been so compressed as to give the fabric the semblance of cloth. The stitches are in good order still, but few shafts and no feathers remain. There are not left enough fragments of feathers to even show the general color, but red, yellow and black stumps were found, all on the front end. Near this end are the remains of black feathers on the outline of a diamond, and a little nearer the middle, thread bindings indicating the former presence of a feather cross shaped like St. Andrew's. To the ends of the malo are attached rows of human molars, and to the sides near the cross a single molar. Looped to the sides, every two inches for the entire length, is a cord enclosed in a fine cjdindrical netting, by which feathers have been fastened in the form of a lei. Upena Manu. — The Museum possesses but one specimen of bird net,*' No. 1 38. This is a diamond-shaped net, mesh 5 inches, made with the same sized twine as in nae. Its total length when stretched is 16.25 ^^^t, and in width it increases from 1 8 meshes at the beginning to 40 meshes at the middle, thence diminishing to 13 meshes at the end. At each end the meshes are bound together. The increase of the number of meshes in a row is accomplished by running on an additional loop (as at^, Fig. Ill) at regular intervals, while the decrease is effedled hy the use of the mauae knot, called viakakiikai. A cord of about .1 inch in thickness is run along both sides, but there are no sticks attached, nor any place for attachment. The net ^'Ibid, pi. ix, midilk- of upper figure. '-Two old Iliiwaiiaii ladies who had been continually at court durini; the reigns of several monarchs were shown this malo, and both independently affirmed that it was the malo of I_,iloa. The story of Liloa and his son Umi has been told too many times to repeat here. Liloa reigned on Hawaii at the latter part of the lifteenth century. It is possible that olonA fibre may outlast the intervening period, and, as the malo constituted the most important part of Umi's family credentials, the succeeding ruling chiefs would have preserved it with reverential care. There are several faint stains on the netting, but whether Ijlood stains or not it is impossible to tell. "15. P. B. Museum Memoirs, vol. i, no. i, p. 13. Fish Nets. 155 lias been taiiued and, where broken, mended witli tanned and untanned twine. The method of use is unknown, but judging from the large mesh and fine twine it was used for entangling. Natives have told the writer of two methods of bird-catching with nets, one of which was to lie concealed on the hills in wait for the birds returning from the sea at nightfall, and as thej- approached heav}^ with food and flying low, to sud- denly appear and raise the net : the startled bird darted upward and became enmeshed. The other method was used for plover when resting on rocks inside the coral reef. The hunter chose a dark night and waded at high tide to one of the uncovered rocks. The net was wetted and silently shaken over the sleeping birds, which, waking and ex- pedling rain, crowded closer together and slept again. The shower was repeated several times and the hunter, then sure of his quarry, threw the net over the group and held down the edges close to the rock with his hands and feet. When the birds were ex- hausted with their struggling, the net was gathered together with the booty inside. Fish Nets. — Upena, or fish nets and methods of using them have been described with considerable detail in Mrs. Beckley's" concise and complete article on Hawaiian fisheries, and Mr. Cobb's" comprehensive work on the commercial fisheries of these islands, in which is incorporated that of Mrs. Beckley. It will therefore not be neces- sary to do more than mention the nets of onl}- Hawaiian origin, and describe and list such as are in the Museum. The netting tools and the sizes of the mesh have been mentioned above. The material used was olona, with rare exceptions. There is an upena ahuulu in the collec- tion. No. 764, with uniisually thick closely spun cord, which on examination proved to be waoke. In this example — a small hand net — it is probable that some cord prepared for koko was substituted during a scarcit}- of olona. The other materials were hau, coir and ahuawa. For the shark net, upena mano, hau bark was stripped from the tree and while fresh was roughly twisted into thick rope. Hau and ahuawa were used for the rough turtle net — upena honu. Coir netting is said to have been used to lie in the water, as a prote(?ttion against sharks, a few feet outside an olona net in which cap- tured fish were awaiting landing. Fishermen declare that sharks dislike the roughness of an article manufaftured of coir. Another fisherman stated that the Hawaiian nets were not of coir, and that the use of such material was due to the Gilbert Islanders, of whom there was a large colony here for a generation up to a few years ago. Taking into consideration the abundance of olona in this place and its flexibility, and the small size and scarcity- of coconuts in these islands, it would seem natural that of the two the natives would turn to olona as the staple. A number of the Museum nets have been "Emma Mctcalf Beckley. Hawaiian Fisheries and Methods of Fishing. Honolulu, 1883. ''John N. Cobb. Commercial Fisheries of the Hawaiian Islands, U. S. Fish Commission Report for igoi. Wasliinj^ton, 1902. U might be mentioned that in Mr. Cobb's excellent report are a number of sketches of fishing implements in this Museum, the acknowledgment of which has undoubtedly been overlooked by that gentleman. 156 Hcnuaiian Nets and Netting. tanned, tlie process merely- requiring the steeping of the net and fresh bark of the knkiii {Alcni-ites triloba) in water for several days. Fish nets are freqnentl}' composed of two or more pieces of netting run togethei with a cord. Five methods of making this joining were noticed, the simplest being with a cord, called aea^ drawn through the meshes of the two edges in turn. The second was by placing the meshes of the two edges exadlly opposite and with a single cord t3'ing them together in order, using a single knot. The third was to laj- the edges together with the meshes alternating and then running the cord from one side to the other with a single knot at ever}' mesh; this at first glance has the appearance of continuous net- ting. The fourth method of joining was to net on one edge and run the shuttle in turn through the loops of the other edge. The fifth was by joining the two edges by the usual form of netting. A number of the smaller nets, particularl}- those of small meshes seem to have been cut down or made of pieces taken from larger nets ; in such cases the netted knot forming the edge seems to have stood the usual wear and tear. When a net has been torn or broken, more generalh' with the fine meshed nets, the mending was sometimes effected with an aea, or a piece was cut from an old net, laid over the hole and sewed on bv the same method. Large nets are sometimes fastened together bj' a rope called kukai. The name alihi was applied to the head and foot ropes, the full names being respec^tivel}- alihi pikoi and alihi pohak/i, or more recently alihi kepau. The pikoi was a float of wood, hau preferably, but failing that, kukui. Wiliwili {Eiytkrina mo)iospern/a), the lightest wood in the islands, was not considered sufficiently durable. In recent times two other woods have become available for the pikoi, and since their introdu6lion oleander {Neriiim oleander) and the castor oil tree {Riciniis con/nnniis) have been added to the list. The shape of the pikoi varied. In some, a large branch was cut into seAions and the pieces sjjlit, and from these the pikoi was made (a block about 4 inches long, 3 inches wide and i inch thick). It was fastened throiigh holes drilled in its side to the alihi. Another form was a section of a branch about 4 inches long, and 2.7 inches in diameter, which was pieced on the side and so attached. The heart wood in the kinds used was soft, and could be easily removed to allow the pikoi strung on the alihi, but it is uncertain if this kind of pikoi was of ancient use. Sinkers of pohaku (stone) and kepau (lead) were shaped to suit. Sticks used to support or dis- tend a net were called k/ikii, and when curved or bent were also called kaka. To make the necessary length for the kuku or kaka two or more sticks were fastened together by overlapping and binding the ends. In referring to the nets b^- their native names, it should be mentioned that one form of fish net is occasionally used for various kinds of fishing and has a different name in each, and that the same name is sometimes applied to different nets in various parts of the islands. I Va runts Ncis. 157 UpENA Papai, Crab Net. — One specimen, No. 762. This is a flat, circular sieve from 20 to 23 inches in diameter. The frame is composed of two kaka securely fastened. Stretched across the hoop is a set of parallel heavy cords one inch apart, at right angles to which is a second set the same distance apart and merely knotted to the first. This sieve of course cannot be called netting. Such an implement was mainly useful in throwing the crabs ashore after they had been enticed within range b}' a piece of bait. Fishermen report another style of upena papai, illustrated b}' Mr. Cobb, 5" in which the cords of the sieve give place to a conical net. Upena Paoo, or Upena Iao. — One specimen, No. 761. A small, fine-meshed, tanned dip net suspended from an ovate-acuminate frame, which shape is called /cw. The frame is rigid and is composed of kaka bent until the ends meet at the point, where they are bound together and form the handle. A cross piece is tied to the frame seven inches from the handle. The net tapers irregularly to a point, which is diredlly below the handle. The mesh varies somewhat ; adjoining the frame is a band of netting 4.5 inches wide of .75 inch mesh joined at the handle. Then come several re(5langular pieces of nae averaging .2 inches in mesh. All the connexions are with aea. These pieces have no selvage, and have probably been cut from the remains of a larger net. Length and breadth of frame, 32 and 17 inches. Depth of net 23 inches. Upena K.\EE, or Kaee Paoo. — One specimen. No. 765. Thenameis pradlically the same as that previousl}- mentioned. The specimen is a small fine-meshed untanned scoop net with a frame similar to but longer and narrower than No. 761 and without cross piece. The kaka are two long pliant wands, resembling willow, twisted together. The frame is so pliant that it may be readil}- elongated by a little pressure at the handle and the opening thereby reduced in size. The net is shaped like a very broad sack, and is gathered very much at the upper margin. In this specimen the netting has been done for the purpose. The mesh of the top row is i inch, and has been re- duced from .4 inch in the second row to .2 inch at the bottom. As netted, it was long and narrow, and to reach its present form, was doubled and netted together at the bottom and one side. Length and breadth of frame 27 and 9 inches. Length of net at bottom, 47 inches, depth 10 inches. Upena Ahuulu, or Upena Puni. — Two specimens, Nos. 763 and 764. In this style of net the kaka are two parallel sticks about 3.5 feet long. The net, of medium sized mesh, is a single piece of netting folded together and sewed at the bottom and side by the second or third method mentioned above. The top edges of the bag thus formed are attached one to each of the kaka for about three-fourths the length, the balance of the kaka remaining bare for handles. Where the two sides of the bag meet -"■Ibid, pi. 23. 1 58 Haiuaiian A^cts and Nciiing. the kaka, the latter are fastened and prevented from spreading by cords about two inches long. The kaka are of about equal strength, and by holding the points of the handles in the palm of the hand and springing the sticks apart with the fingers, the kaka, being tied at the other end, become bowed and so open the mouth of the net wider when necessary. No. 763 is of olona, and No. 764 of waoke, both tanned. The measurements are : — 763. Length of kaka 42 in., width apart 2-2.3 i"-; length of net 33 in. or 25 meshes, depth 16 meshes, mesh 1.4 in. 764. Length of kaka 44 in., width apart 2.-2.5 "'-i length of net 36 in. or 16 1-2 meshes, depth 8 1-2 meshes, mesh 2 in. UpKna Holoholo, or UpEna Poo. — One specimen. No. 5176. A large tanned dip net with a frame the same shape as No. 761, but three times as long and twice as broad. The cross piece is 16 inches from the point. The net has a mesh of 2 inches, and is composed of a piece of netting folded together and netted along the sides to form a bag. The net is suspended from the frame as far as the cross piece, the balance of the kaka being left uncovered for the handle. Length of frame 8.3 feet, width 2.8 feet. Length of net, 81 meshes, depth 24 meshes. Upkna Uhu. — Four specimens, Nos. 766, 767, 76S and 5175. These specimens are each composed of a square piece of netting stretched flat by alihi, which are fast- ened by the corners to the ends of two kaka tied at right angles to each other at the middle. The diagonal of the netting is less than the length of the kaka, so that the latter are somewhat arched when in position. Weights of stone, or lead, are fastened to the ends of each kaka. Specimen No. 766 is labelled iipcna akiikii^ and has a mesh con- siderabl}' smaller than the other nets of this kind. Upciui pakiikiiis probably the same. 766. Upena akiikii, length of kaka 6.5 ft., size of net 21X22 1-2 meshes, mesh 2.3 in. 767. Upena uhu, length of kaka 8.3 ft., size of net 10 1-2 XiS meshes, mesh 6 in. 768. Upena uhu, length of kaka 8,.5 ft., size of net 16X17 1-2 meshes, mesh 5.3 in. 5175. Upena uhu, without kaka, size of net 15X15 meshes, mesh 4.5 in. Mr. Cobb's description" of an upena uhu does not seem to agree with the specimens in the Museum, nor with the evidence of fishermen consulted. The piece of netting is hardly bagged at all, and is almost as flat as a piece of netting can be in hori- zontal suspension. The kaka, which Mr. Cobb states were made "to swing around and lie parallel" and thereby close the mouth of the net to retain the fish, are fixtures when the implement is set up and cannot be moved from their relative position without drawing the net to the surface where the hand can reacli and untie them. \\\ drawing the implement rapidly through the water the net would of course act as a drag and bow the "Ibid. Various Nrls. 159 kaka somewhat, and bag tlie net, but tlie sticks are too stiff to bend much. The secret in catching the uhu, as described to the writer, is the knowledge that the instinct of the fish when disturbed is to dart to the bottom, and the uhu after being enticed into the net b}' the decoy spends its time in trying to swim through the meshes underneath. But one fish is caught with each dipping of this net. The scoop net described by Mr. Cobb on page 407^' is known to the natives as knee. There is a scoop net in the colledlions. No. 9052, without name or history, which from its appearance might well be used as a kaee also. It has a frame composed of two kuku 3.7 feet long, forked at one end ; at the forked end the}' lie 2.3 feet apart, and come together at the others for the handle. Near the handle is a cross piece 8 inches long. Resting in the two forks is a kaka 6 feet long made by join- ing two slight wands. The net, .8 inch mesh, was a piece 51 meshes deep and 153 wide, and this has been sewed at the bottom and side by aea, making a bag 76 meshes wide. The top edge of the bag is fastened to the kaka which is somewhat bowed bj- the tension. About the middle of the two kuku, the sides of the bag are tied to the sticks and from the bottom of the bag is a cord reaching to the end of the handles. This net in position lies on the kuku with a wide mouth on the ground held wide open by tlie kaka which is bent around in a semicircle. Behind the mouth and slojjing up towards the hand is the bag. UpEna Nae Kuku. — Three specimens, Nos. 756, 757 and 6578, untanned. This is a scoop or bag net, of very fine mesh as indicated by the name. It is shaped like a sack, broader than long, with the end and top open, and gathered at the upper edges. Along part of these edges. Fig. 165, two kuku are tied, with their ends resting 5= Ibid. Fig. 164. SAMPLE OP .4 INCH MESH. i6o Hawaiian Nets and Netting. in rings at the corners. In operation these ends are placed on the sea bottom as far apart as the mouth of the net Avill allow and the sticks elevated at an angle of about 45°. The closed end of the bag drops behind like a pocket. These nets are composed of several pieces joined by aea and growing finer in mesh towards the pocket. In the figure is given a plan of No. 657S showing the meshes and sizes of the varions pieces. This specimen is called upeua nae hinana, or upena hinana and is of par- ticularly fine mesh. Fig. 164 shows a sample of .4 inch mesh in No. 757. The twine in these nets is much heavier than might be expedled in those of such close mesh. 756. Total length 7.5 ft., depth 4 ft., mesh .2^-.\ in., length of kuku 4 ft. 757. Total lengtli 25 ft., depth 7 ft., mesh .3-5 in., length of kuku 17.5 ft. 6578. Total length 13.3 ft., depth 4.6 ft., mesh .05-.35 in., length of kuku S.6 ft. It is probable that the nets described as upcna neliit, itpcna pua, kapuni nchn and upcna iiao b}- Mrs. Beckley and Mr. Cobb would all come under the above description. Upena Kaili, or Upena Aai-opelu. — One specimen. No. 760. This is a large semiglobular bag net suspended from two kaka. In position the ends of the kaka are overlapped and tied, making a circular opening. The length of the kaka is the same as that of the fishing canoe. Three men handle the net, one to draw the net to the sur- face, and one at each end of the canoe to untie the ends of the kaka and allow the sticks to spring to. The bottom of the net opens into a small bag, to which is fastened the stone sinker. In the specimen the kaka are wanting. This specimen was made of many pieces of netting of various meshes. On the top is a narrow band 7 meshes wide of 2.5 inch mesh. Then follows a band of 35 meshes wide and 2 inch mesh. The rest of the netting is 1.2 inches mesh, and, with the exception of the bag at the bottom, was made in two vertical sections. Each seAion was made up of strips of netting drooping at the middle, gathered at the upper edge and joined at the ends to the netting of 2 inch mesh. B3' this means the native secured the oval shape. Where portions of the ends overlapped, the larger sized net was cut and the piece fitted. The strips and sections were run together by aea. To attach the bag, the two bottom strips were cut in the middle of the lower edge and the bag fastened by the same cord as before. The bag was a broad piece of netting, netted at the sides, with an open bottom which was closed when fishing by tying with cord. It might be men- tioned that the bag is 105 meshes round, while the opening in the bottom of the larger net is only 76 meshes in circumference. This specimen is 48 feet in circumference at the top; the depth, hanging flat, is 19.5 feet and that of the bag 2.9 feet. Upcna kaihi is said to be similar to this net. Upena Eke. — One specimen. No. 5310. This specimen is an old untanned bag net, of which most of the bottom is missing. When in operation it is a long hori- Upcna Eke. i6i zontal bag, tapering but slightly away from the mouth, which is at one end. At the other end in the specimen is an opening less than half the size of the mouth, where probably was fastened a closed bag of smaller size. In sedlion, the bag is semicircular, with the arc above. The bottom part of the net called Jiomia is anchored by stones at the corners of the mouth and at two points along each side about ten and twenty feet from the mouth. Between these points, stretching the honua wide are sticks called piilii. There is another stone at the tail end of the bag. To the upper curved part, called laiia^ the pikoi are fastened in great numbers in irregular rows. Heavy ropes of hau run the whole length of the bag, one along each edge and one on each slope of the lana t II ^ett ,, about 8 feet from the edge. The total length of the lana is 39 feet, the first 21 feet from the mouth being light cord of .08 inch in thickness and of 1.5 inches mesh, and the rest of heavy cord .17 inch thick and 1.2-1.4 inches mesh. The lana is 29 feet in width at the mouth, measuring the ropes, and the lighter portion is made up of five pieces of wide netting averaging 33 meshes long and decreasing from 407 meshes wide in the first piece at the mouth to 23S meshes in the fifth. The pieces are run together by aea, in this net called iwi piilii. This method of tapering a net seems to have been preferred by natives to the use of the knot known as makakukai, already mentioned. The heavier part of the lana is in two pieces, 112 and 70 meshes long, and respedlively 170 and 165 meshes wide. Of the honua, there remains but one piece, of the heavier cord, at the rear end. It is about 11 feet or 113 meshes long and 130 meshes wirle. Leading away from the sides of the mouth, there should be two long nets called pakii^ the same depth as the bag, at a wide angle to each other, as a drive. The net on the right is said to be 15 fathoms long, and that on the left 10 fathoms. Upcna ko/o is said to be a net on the same lines as the preceding, but on a much larger scale. The same is reported of the net used in Laic kapalili. It is probable that the nets called />rt/>c^/^«//^^' and au-mai-ewa and used in conjundlion with each other are also similar. Memoirs B. P. B. Museum, Vol. II, No. i. — ii. NO. 1^5. 1 7> ft. i t/nt. DIAGRAM OF UPENA NAE KUKU. l62 Haiuaiian Nets and Netting. Hano Ohua, or Upena Ohua. — One specimen, No. 7072. This specimen has been made of pieces of netting purchased from and netted by Chinese from Chinese fibre, and run together by natives by aea in what is probably one of their ancient forms of fish nets. The knot is what is known as the square or reef knot and is a poor one for iish nets. The mouth is semicircular, 11. 8 feet at the base and 19.5 feet along the arc, and from there tapers to a small hole 2 inches in diameter. The length of the upper side is 1 1.8 feet, and of the lower 9 feet. The mesh is about .5 inch. The mouth is leaded at the bottom, and along the upper part is closely strung a row of pikoi. Fig. 166. Hano I'ao and Iiano vialolo are re- ported as having this shape, the latter being a much larger net. Upena Kuu, Upena Paloa, or, generally speaking, seines or gill nets. Two specimens, Nos. 758 and 759, tanned. 758. Length 82 ft., depth 7 ft. or 41 meshes, mesh 2.2 in. 759. Length 66 ft., depth 10.5 or 48 meshes, mesh 3.3 in. The names of some of the nets re- ported^'' ixnder this heading are akuikui or pakuikui, ulunlu, ka- ^'g. 166. hano ohua. waa.^ hipo, luelue, pahu, papaolewalezva, pouono. Occasionally a kuku was fastened to each end of the net, and was of great assistance when drawing the net through the water. Upena Mano and Upena Honu. — Large meshed roughly made nets of about 12 or 6 in. mesh respecftively. They are reported to be about 100 ft. long and 6 ft. deep. There are several other names,''" attributed to fish nets, for which the writer has been unable to find descriptions, viz.: apai.^ n/iina, kai\ kaii\ kiilcohua^ Iitahia, makui. Some of these names no doubt refer to forms of the nets described above. In concluding, the writer wishes to express his appreciation and thanks to Mr. L. G. Blackman for his careful and painstaking drawings, and to Dr. W. T. Brigham, Professor Otis T. Mason and many others for suggestions and information cheerfully given and the generous loan of specimens. "Hawaiian Antiquities, by David Malo, pp. 277, 279, 280, and Andrews' Hawaiian Dictionary. '"Ibid. Publications of the Bernice Pauahi Bishop Museum. A Handbook for the Bishop Museum. Oblong octavo, 94 half-tone illus- tnitions. Price 50 cents ; postiige 9 cents. Occasional Papers, Vol. I. Octavo. No. I. Director's Report, iSgS. Visits to Ethnological museums in a journey around the world. Price 50 cents; postage 5 cents. No. 2. Director's Annual Report, 1S99. Mat Sails of the Pacific (.Stokes). Kay-skin Rasps (Walcott). Notes on the Birds of Oahu( Scale). Price 50 cents ; posbige 4 cents. No. 3. Director's Ann"al Report, 1900. Visit to the American HI '.scums (Stokes). Mission to Guam (Scale). Notes on the Birds of K.uai (Bryan and Se:ile). Price Ji ; post;ige 8 Cents. No. 4. New Hawaiian Pishes. By Alvin Scale. Price 25 cents ; p(>s1;ge 2 cents. No. 5. Director's Annual Report, 1901, and Index to Vol. I. Illustrations of sjcim whale and bird groups. Price 50 cents ; postage 4 cents. Occasional Papers, Vol. II. No. I. Director's Annual Report, 1902. Noteworthy Hawaiian St- n.; Implements (lirigham). Fibres of the Hawaiian Islands ( Black- ni.n>. Uistribulion and Variation of Achatinella multizonata (Cooke). M -ingraph of Marcus Island ( Bryan). Price fi ; postage 7 cents. No. 2. Director's Annual Cleport, 1903. Remarks on Phallic St Uvts from Rap:inui (J. L. Young). Aboriginal Wooden Weapons of Australia V Black mt;ng H;Lbits of the Hawaii;-.!! Owl (Bryan). Description of the Nest jMi'i l.gg.; v>f Clil^rodreoanis virens (.Gmel. ) (Bryan). Notes on the .■\i;!eric n Birds C'ilecteif in the Ilawiiian Isla!ids by Mr. Gerrit Wilder ( Bryan). A Birds Nest of Peles Hair (Bry.in). Two Ui!described Nests an;l an Egg of a Hawaiian Bird ( Bryan). Price 50 cents; postage 5 cents. No. 4. Director's Annual Report, 1905. Old Ha-waiian Can-ings (BngliaH!). Three New Ha-waiian Fishes (Bryan. ) Report of a Visit to MIdwiy Isl.-nd (Brj'ai!). Nest of Hawaiian Hawk (Bryan). In press. O jcasl.^nal Papers, Vol. III. Reprints of Original Descriptions of Achati- !iella.'. By iC. W. Thwing. In press. Occasional Papers, Vol. IV. Fishes of the South Seas. By Ahan Scale. In press. Memoirs, Vol. I. Qu .rto. No. I. Hawaiian Feather Work. By Wm. T. Brigham. Price 5 .50; postage 14 cents. No. 2. Index to the Islands of the Pacific. By Wm. T. Brigham. Price fi.50; post;'.ge 20 cents. No. 3. Key to the Birds of the Hawaiian Group. By Wm. .\lanson Bryan. Price fi.50; postage 13 cents. No. 4. Ancient Hawaiian Stone Implements. By Wm. T. Brig- ham. I'rice 12.50; postage 20 cents. No. 5. Supplement to Hawaiian Feather Work, with Indeir to Vol. I. By Wm. T. Brigham. Price 75 cents; postage 8 cents. Memoirs, Vol. II. No. I. HawaiianMat and Basket Weaving (Brigham). Hawaiian Nets and Nettings (Stokes). Price J3 ; postage 20 cents. MEMOIRS OF THE BERNICE PAUAHI BISHOP MUSEUM OF POLYNESIAN ETHNOLOGY AND NATURAL HISTORY. Vol. II — No. 2. Old Hawaiian Carvings. honolulu, h. i.: Bishop Museum Press. • 1906. BOARD OF TRUSTEES Sanford B. Dole, LL.D President Henry Holmes - Vice-President Alfred W. Carter Secretary Joseph O. Carter Treasurer Samuel M. Damon William O. Smith E. Faxon Bishop MUSEUM STAFF William T. Brigham, Sc.D. (Columbia) . . Director William H. Dall . . Honorary Curator of Mollusca William A. Bryan .... Curator of Ornithology John F. G. Stokes . . Curator of Polynesian Ethnology Leopold G. Blackman . . Assistant, and Acting Librarian C. Montague Cooke, Jr., Ph.D. (Yale) . . Assistant John J. Greene Printer MAV -' i9nR OLD HAWAIIAN CARVINGS FOUND IN A CAVE ON THE ISLAND OF HAWAII FIGURED AND DESCRIBED BY WILLIAM T. BRIGHAM Memoirs of the Beniice I'aualii Bishop Museum Vol. II — No. 2 honolulu Blshop Museum Press 1906 ILLUSTRATIONS IN THE TEXT. FROM PHOTOGRAPHS BY THE AUTHOR. Two Aumakua placed to guard the cave deposit Title page Plan and sedlion of the Cave, sketched by David Forbes 4 First portrait Statuette 5 Second portrait Statuette 6 Profiles of Statuettes 7 Side view of Aumakua 8 Back view of Aumakua 9 Bowl carved with female figures and inlaid with human teeth 10 Top of carved Bowl 11 End of carved Bowl showing face of one figure 11 Other end of carved Bowl showing face of other figure 11 Ipu aina or Slop Bowl inlaid with human teeth and bone 12 Similar Ipu aina in the Bishop Museum 13 Papamu for playing the game of Konane 14 Carved Bowl in the British Museum from Hawaii, showing similar supports 14 Carving tool of shark tooth mounted in the end of a human clavicle 15 Shark tooth carving tool in British Museum 15 Hue wai curiously moulded in growth by net work 17 Bare net work of feather cape once covered with red feathers 18 Funnel-shaped carving of wood, use unknown 19 Painted Fan 19 Fragment of a porcelain dish (Chinese) 20 PLATE. Enlargement of the carved portion of two Aumakuas Plate XVII Old Hazvaiian Carvings found in igo^ secreted in a cave on Hawaii. Figured and described by William T. Brigham, Sc.D. (Columbia), DireBor of the Bishop A'Inscuni. IN a cave on tlie western portion of Hawaii, of which the main opening was not concealed, a number of human remains were deposited, some of great antiquity, others certainlj' after the advent of the white men. So far there was nothing to distinguish this from scores of other burial caves, some abandoned, others still in use in the rural districts : all are bubbles or channels in ancient lava streams, — streams which have gradually built up the islands and are still occasionally poured forth from the adlive volcanoes of Hawaii. In recent years streams have flowed the same way as the ancient lava torrents which were used as burial caves, and the Bishop Museum contains articles snatched from these caves before the advancing flow. To one familiar with these caves the vestibule, as it ma}' be called, is the least interesting portion ; it is a tunnel of varj'iug size and irregular seAion, often shrink- ing to a passage hardly large enough to admit a man's bod}', then expanding into a hall of considerable height, and with numerous ramifications, some small as rat holes, others of a size to admit a horse. In the old and important caves these side passages are generally walled up, and originally this was done so effeAually as to conceal the entrance, but earthquakes and time have loosened the dry stone wall, and it is now possible to detect it, unless the deposited and decaying bones cover it, and these one hesitates to disturb unnecessarily. Behind these walls lie the interests of the cave. One may pull down with great labor in the dark and ill ventilated cave the carefully placed wall and find in the chamber behind it perhaps nothing, perhaps a few crumbling bones or a decayed calabash. Of three explorers of the cave around which centres the interest of this account, one had often explored such caves in search of water, and coming from a side chamber where was a corpse in a canoe raised on a stand with choice mats and kapa, he noticed a wall not far from the entrance to this burial chamber (Fig. i). After much hard work a small opening was made and the recess entered. There were fragmentary human remains, but on one side of the chamber were the articles which are the subject of this paper wrapped carefully in kapa. What connedlion they have, if any, with the bones in the main cave, or those in the branch chamber I cannot say: perhaps they were simply hidden here as the safest place of deposit known, — in the guardianship of the (3-165) I IflG. I. SKETCH PI^AN Olf PART OF THE CAVE. 1. Opening to cave from gulch. 2. Branch opening to B, closed with rocks. 3. Branch opening to c, closed with rocks. 4. Canoe with skeleton, feet to east. 5. Wooden bowl with carved figure. 6. Papamu or konane board. Two human figures, heads to east. Two auniakuas leaning against wall. vSkulls and bones rolled in kapa. Ipu aina with human teeth. Large gourd containing skeleton of infant. Old Hawaiian Carvings. 5 dead. It lias been suggested that tliej'- form the paraphernalia of a temple and were hidden, as so many of the idols were, at the time of the general destrudlion of the idols in 1819 in the hope that the storm would blow over and better times ensue, but there is absolutely nothing in the colledlion to support such a view. The two gods or auniakita were household deities, the other articles might be the private property of FIG. FEMAI.K PORTRAIT STATUETTE. some chief or priest, and two things, — the fan and bit of porcelain are such keepsakes as were commonly deposited with the dead to whom the articles had belonged. The objeAs were carefully wrapped in kapa, but the explorers unfortunatel3' removed all this without especiallj- noting the kinds used on each, and we are deprived of any clew this disposition might afford. We have the letter but the addressed The cave opening was about 30 feet above the bottom of a drj' gulch running northeast and southwest. It was about 3 feet wide and the passage into which it led (a) was about 55 feet long. At about half the height a walled-up opening (2) was found leading to a narrow passage, 2 feet wide, soon opening into a chamber (b) approximately 10 feet high and wide with a length of 30 feet. This chamber (b) contained a human skeleton placed in a canoe and raised on a stand (4). As will be seen in the lower diagram this chamber is below the passage .a. Another walled-up door (3) led to a passage about 2 feet wide and 7 feet long opening into a chamber (c) perhaps 20 feet long and 7 feet wide and high. In this last chamber were the articles here described. [167J 6 Old Haivaiiaii Carvings. envelope was torn off. In passing it should be mentioned that the kapa was in fair preservation, some of the specimens being of verj^ choice quality, such as was gener- ally made only by women of high position. First and most important in the find are two wooden statuettes, both female, perhaps mother and daughter, which I consider portraits," and if this conjeAure be t'lG. PORTRAIT .STATUETTE OF GIRI,. ' I am not vminincUul of lliu inlcresting stone fis^urt- already described and rit^ured in the essay on Hawaiian Stone Im- plements (Mem., vol. i, 4, p. 95), bnt that, if not of foreij^n workmansliip, was a portrait of a foreigner. All that is known of this curious bust is given in the essay referred to, but the illustrations may l)e repeated here to show the con- trast to genuine native work. [168] Old Hawaiian Carvings. 7 correct, they are the finest specimens known of Hawaiian portrait work. The measure- ments and description are as follow : — I. Height 28 in.; breadth of shoulders 12.3 in. Dark wood well polished. Vulva carefully carved, breasts conical. Teeth of lower jaw inserted as one piece of FIG. 4. STATUETTKS IN PROFILE. light colored wood. Hair neatl}- pegged in the usual waj-, reddish and ver}^ brittle. Eyes of shell, right with peg hole drilled, left entire and more recent. 2. Height 27 in.; breadth of shoulders 11.2 in. Hands and ears better finished than in No. i. Breasts hemispherical. Vulva deep cut, with deeper hole. Mouth open with hole at bottom as if to attach tongue. Eyes pearl, without holes. Hair pegged in with some fibrous wood. Same wood and workmanship as No. i. [169I 8 Old Haivaiian Carvhigs. Be3'ond the limitation of insufficient tools and primitive methods there is a type seen in all Hawaiian sculpture which is strongl}' developed here. In addition there is a freedom and individuality in these little statues, which are doubtless from the same hand and of the same date, seldom seen in the images of the gods. The relative age of the two subjects is shown not only in the breasts but is most stronglj' marked in tlie facial features. The limbs are rendered much truer to nature than usual, with less of the elephantine form, fully separated, and the balance is so well preserved that the images stand firmly on their feet. When found each image had ou an apron of kapa. The hair, which in its unkempt state gives a wild though realistic effect to the figures, is the usual coarse hair of the Hawaiian, in these speci- mens of a reddish tint, whether from coral lime dyes or from decay I cannot sa}', but the hairs are certainly very brittle. The method of attachment to the head is that commonl}' used in Hawaiian figures, viz.: a hole, round or square, is made in the head ; a lock of hair is bent in the middle and the bend iniserted in the hole and driven home b}- a flat wedge which is finall}' cut to be flush with the general surface and the hairs are thus held securely without anj^ cement. These holes are distindll}' seen on the back of the head of the first figure where the hair seems to have broken off but not come out. The empty square holes are more plainly shown in the carved bowl (Fig. 7). The eyes are of flat pearl shell, large and held in place without cement. Only one has been drilled, although it was common, especially in the feather and wicker work heads, to drill a rather large hole through which a round head peg held the shell fast and served to give the expression of a pupil to the eye. The Maori made a similar eye of paua shell, but the central peg was usually an integral part of the head. The older woman has a block of artificial teeth inserted in her lower jaw, a curious deviation from the method used in many of the god images of inserting dog teeth. I have never seen human teeth used in this way, although in one idol in this Museum (No. 7658) there are cavities in the mouth larger than would be [170] FIG. 5. SIDE VIKW OF AUMAKU.'V. Old Hawaiian Carvings. required for ordinary dog teeth. The younger woman has a hole drilled in the back of the mouth perhaps for attaching a tongue. The waist, especially in the younger woman, is contrafted, and there is a curious ridge between the abdominal and pubic regions. This ridge is also shown in an image in the Museum (No. 7656) which maj' be a poor attempt at a portrait. In the Museum figure the hands are better wrought than in either of the female im- ages here figured; the breasts are well modeled and show the large Hawaiian nipple. The calf of the leg is, however, affe(5led with elephantiasis, and the balance is destro3-ed. The head is remarkabl)^ brach3-cephalic, the ears small and high, nose prominent and curved, mouth open and with thin lips, no signs of teeth or tongue. The e3'es were of shell and the holes for the attaching pegs are neither central nor sj'mmetrical. The holes for attaching the hair are of the rudest form, and while the hair is gone the pegs generally remain. Another anatomical feature seldom shown is the diaphragmatic line separating chest from abdomen. This image was for many 3'ears in the cabinet of the American Board of Commissioners for Foreign Missions at Boston. I can hardly say whj', but these two female figures remind me strongly of the old-fashioned figureheads of ships," which, however, often had less _expression than these, and it is quite possible that the na- tive artist had lessons from some European or American sailor, for I should hardl}' place these carvings earlier than the early part of the nineteenth century, or possibly the last decade of the eighteenth. At the time of the destruAion of the idols, after the kapu was broken, many " I have compared them with photographs I once took at Nantucket of a number of these figureheads, once the pride of shipowners, now discarded by a changing fashion, and while I cannot trace the least resemblance to indi- vidual figures, the impression remains. [171] FIG. 6. BACK VIEW OF AIJIAKUA. lO Old Hawaiian Cari
  • -" .cl (U T^ o S & •SS C3 O "O V -« (U OJ Ph '~' ^ o 13 2 ^ ■4-* ""' a 3 tn o 3 «f . XJ > 3 t^ ■a >, A era n of b; OJ en 'o a p biO M S u O -M bG &( .-*^ t/} U-i a Memoirs Bishop Museum, Vol.. II. Plate I.XV. is .3 ^ vo" OJ ON ^ CO -M > i-T W o ii ^ s o O o oi O 3 a a Memoirs Bishop Museum, Vol. II. PLATE LXVI. 1) HI u M u a 'S< u u u (U 3 ^ J3 •a -*-• 4-t _ n 'o cl o h-f o a (LI ."2 o o s ,13 a M-« a a o ,^ o -*-• w •a s 13 ■4-< 1— ( a "a; a >— ( > X s o ■♦-> .5f i-r -a Q ,a a, w a rt XI o be ^a 4-1 ■♦-' a o o o a CI & 3 o ft s 2 aj S 3 ft 'o Jx< (U a •6 S o a o\ in a u 00 !fl o M s ^ 'o a 'o .2 >% c« -M »-< a a, tu a 3 •— < rt S Memoirs Bishop Museum, Vol. II, Plats LXVII /= ■+ ^Klt- 'Vmr*^- ']] I If