k. :<■ ; n ^ ( HI'" ^ ^Ri' 1 1 k 6eR ^t^xrwc^ \^^^ HARVARD UNIVERSITY LIBRARY OF THE MUSEUM OP COMPARATIVE ZOOLOGY JUL 24 1922 OCCASIONAL PAPERS 01" THE BERNICE PAUAHI BISHOP MUSEUM OF POLYNESIAN ETHNOLOGY AND NATURAL HISTORY VOLUME VII honolulu, hawaii Bishop Museum Press 1922 ^ .^ CONTENTS Number Page 1. Director's report for 1918, by John F. G. Stokes i 2. The languages of the Pacific, by J. MacMillan Brown 13 3. New Hawaiian plants, by Cliarlcs X. Forbes 31 4. A new Cyanea from Lanai, Hawaii, by Charles N. Forbes and George C. Munro 41 5. Notes on Marsilea villosa Kaulf, by Charles N. Forbes 45 6. A new variety of Partulina horneri, by J. J. and A. Gouveia 51 7. New species of Sierola with explanatory notes, In' David T. Fullaway. (Issued October, 1920.) 55 8. Director's report for 1919 161 9. Edible mollusca of the Oregon coast, by Charles Howard Edmondson 177 10. Fish poisoning in the Hawaiian islands, by John F. G. Stokes 217 11. An archaeological survey of Haleakala, by Kenneth P. Emory 235 12. Notes on Hawaiian Zonitidae and Succineidae, by C. Montague Cooke, Jr 261 13. Stomatopoda in the Bernice P. Bishop Museum, by Charles Howard Edmondson 279 14. Dermaptera and Orthoptera of Hawaii, by Morgan Hebard. (Issued April 4. 1922.) 303 ILLUSTRATIONS Pr.ATE P^se I. Fiber caskets of Kings Liloa and Lonoikamakahiki 4 II. Holman feather cape, 6 III. Hibiscadelphus bombycinus 40 IV. Viola kauaiensis wahiawaensis 40 V. Cyrtandra olona 40 VI. Cyrtandra hii 40 VII. Cyrtandra propinqua 40 VIII. Cyrtandra munroi 40 IX. Cyrtandra georgiana 40 X. Clermontia samuelii 40 XI. Argyroxiphium caligini 40 XII. Cyanea baldwinii 44 XIII. Marsilea villosa 50 XIV. Marsilea villosa 50 XV. Partulina horneri var. kapuana 54 B. P. B. Mus. Occ. P. \'ol. \II Pl.ATK Wl. XVll. XVI II. XIX. XX. XXI. XXII. XXllI. XXIV. XXV. XXV !. XXVII. New species of Sicrola Making hola .1, llola in grass "spoons;" B, Applying liola J, The catch; A'. Aulnilm ( 'i\>plir'isia piscatoria). ILLUSTKATIONS Page 72 234 234 234 A. Platforms in I'uu Xane craUr; />'. Excavating in the south platform, Pmi Xaue group 260 J, Burial ahu in Kamoa O Pele from northeast; Z^, Large single terrace, llalalii grou]). A, Platform and series of three terraces, llalalii; B, Series of hve terraces, Halalii A, North platform Hanakauhi group; B, North platform, Laie group 260 Shells of Zonitidae ^^S Shells of Succineidae ^78 Drawings of Hawaiian Dermaptera and Orthoptera 377 Drawings showing sections of Hawaiian Orthoptera 37° 260 .260 Figure. I Page Distribution of shellfish in Tillamook Bay, Netarts Bay and along adjacent shores ~°^ Distribution of shellfish in Yacptina Bay and River -'03 Distribution of Mya arenaria Linnaeus in the Siuslaw River ^07 Distribution of clams in Coos Bay ^^^ Sperm and ova of edible mollusks 212 Sperm and ova of edible mollusks Map of Haleakala Sketch of the ahu in the crater of Kamoa o Pele showing manner of burial Map of the crater of Halalii 214 239 Godwinia caperata Godwinia caperata Godwinia liaupuensis, new species. Vitrina tenella Nesovitrca pauxillus 240 242 265 266 268 270 27?, Accessory organs and segments of Stomatopoda 282 Coronida ^inuosa, new species. 294 I. Paratrigonidium rosenm 334 OCCASIONAL PAPERS OF XHE BERNICE PAUAHI BISHOP MUSEUM OF POLYNESIAN ETHNOLOGY kM,'- . , NATURAL HISTORY Vol. VII— No. i. With Plates I-II. Dire6lor's Report for 1918 HONOI.UIvU^ HAWAII BisHot» Museum Press 1920 Bemice Pauahi Bishop Museum I9J8 BOARD OF TRUSTEES Am'I;'! ' I ■ Ml - - - - - President E - - - - Vice-President J. M. DowsETT ------ Treasurer WiiXiAM Williamson . . - . . Secretary 1Ii:niotany Otto H. Swezev - - Honorary Curator of Entomology John W. Thompson - - - Artist and Modeler Miss E. B. Higgins ----- Librarian Miss L. E. Livingston - - - Assistant Librarian John J. Grekne ------ Printer M. L. Horace Reynolds - - - Cabinet Maker EXHIBITION STAFF Mrs. Helen M. Helvie - - - - Superintendent John Lung Chung _ . - . . Janitor Thomas Keolanui - - . - . Janitor John Penchula -_.-.- Janitor OCCASIONAL PAPERS OF THE BERNICE PAUAHI BISHOP MUSEUM OF POLYNESIAN ETHNOLOGY AND NATURAL HISTORY Vol. MI— Xo. i. With Plates I-II. n •» Director's Report for 1918 honolulu, hawaii Bishop Must^um Prlss 1920 Director's Report for 1918 TiiK retirement of Dr. \\'illiani T. Brighani as Director of the Bishop AInseum. foreshadcnved in his annual report for 1917, took place at the beginning- of the present year. By vote of the Trustees, he was given leave of absence for the year 1918, and appointed Director Emeritus and Curator of Anthropology beginning with the year 19 19. Dr. Brigham's retirement as the active head of the museum after nearlv 30 vears of service is regarded with regret by those under his direction, as they appreciated the helpful interest he displayed in their work. The advantages of his many accomplishments and wide experience, he was always ready to share with his assistants, while his courtesy and consideration are among the characteristics that have endeared him to them. It is a matter of great satisfaction, however, to those formerly under Dr. Brigham's charge, that his valuable services have been recognized by the Trustees with the title of "Director Emeritus" and that his association with the Museum will continue. During Dr. Brigham's leave of absence, and pending the ap- pointment of a new Director, ]\lr. John F. G. Stokes, Curator of Polynesian Ethnology, was requested to assume administrative responsibility for the Museum, with the title of Curator in Charge. Before leaving for the mainland on his well-earned vacation. Dr. Brigham was able to make another contribution to the study of Hawaiian featherwork. which included all the information he had gathered up to date. This was published as Memoirs \'olume \'II. No. I. Second Supplement Hawaiian Featherwork. He also com- pleted his Annual Report for 191 7, which was, as usual, published in Occasional Papers. The first Territorial Fair, conducted largely as a food conser- vation measure, took place in Honolulu in June. The ]\Iusei'.m was invited to exhibit from its extensive collection of fruit and fish ' The Director's Report for 1918 was prepared by John F, G, Stokes, Curator of Poh-nesian Ethnology. [3] 4 Dircilar's Report for !(ji8. casts. As nn ])n)visiiin was iiiade for tlu- ])r()tccti(m nt truit casts fR)ni hciii^;' liandlcMl 1)\' the i)ul)lic, this ]K)rti()ii of tlic rxliihit was with(h-a\\n. h'or the lish scctimi, howexxT, casts of thirty of the hir,u;cst cihhlc fishes, es])ecian_\- of tll:l^e --omewhat ne^iccted hy the fish-eatin.^' ])uhhc, were seK-cted ami exhihiti'ort for iqiS. 5 The number of etlinological accessions during the year was 365, classified as follows: gifts, 180; purchases, 126; loans, 15; gifts of relics, 44. This is less than the average for the preceding five 3ear period, in the early part of which large collections were made, given and purchased. GIFTS. Mr. Bruce Cartwright, Jr., presented an accumulation of small loans made by him from time to time and amounting in all to fifty-two specimens ; Judge S. B. Dole — one of the rare Neckar Islands stone idols, which makes a total of nine now brought to light; Judge H. E. Cooper — a tobacco pipe, apparently of slate, whicli was dug up in Manoa Valley, Oahu, the same valley in which an Indian arrow head was found many years ago. The wa'u ipii (scraper of cowry shell) lost siglit of until 1916 when specimens of it were found by Messrs. Judd and Cooke on deserted house lots on IMolokai (described in Occasional Papers Vol. VI, p. 232) has now been turned up in great numbers, as shown by the gifts of Messrs. Judd and Munro. Another gift worthy of special mention (although outside the province of the museum) was a collection of Moro weapons captiu'ed by men of the Twenty-fifth Infantry, U. S. A., in the Philippines. It was presented by the men of the regiment when vacating their station at Schofield Barracks, Oahu. LOANS. Among other things deposited in the Museum by the various interests concerned, were two cocoanut fibre caskets ( PI. I ) , believed to contain the bones of Liloa, king of Hawaii about 1500 A. D., and Lonoikamakaliiki, a great grandson who reigned later. Examples of this form of l)ody encasing have not been observed before and undoul:)tedly are extremely rare. They will be described in a later puljlication. A feather cape (PI. II) and three lei (ornamental strings of feathers) which were taken to Xew England in 1822 by Mrs. Lucia Ruggles Holman, one of the earliest missionaries, were lent l)y the present owner, a descendant. The cape was given liy Queen Kaahumanu to Mrs. Holman, probably for her daughter, claimed to be the first white child born in the Hawaiian Islands. It is remarka1)ly well preserved, tlie feathers showing very little wear. The colors are yellow (00), red ( iiwi, ) and black (00), and the measurements are: length of back, 355mm.; length of fronts: right 272, left 260 ; greatest width 820. PURCHASES. A collection of stone implements made by Mr. H. Schullz, during his residence of a quarter of a century on tlie island of Kauai, was the most; important purchase of tlie year. It consisted of eighty-five items, whicli brought several new forms to our knowledge. [5] Director's Report for jqiS. RELICS. Durint; the \car two royal standards were received, for each of whicli was claimed llie distinction of being llic royal standard lowered at the over- throw of the 1 lawaiian monarchy in 1893. One was given liy !\[r. A. A. Brown of San Francisco, and the other hy Mr. George \i. Smitliies of Honolulu. The flags arc of similar design hut differ in other respects. In the hope of establishing the identity of this historical specimen, incpiiries ha\e been made among persons who witnessed tlie revolution, but with no deiuiitc results. Mr. Smithies also presented, in the name of his wife, many other relics including the sword of his late Majesty Kalakaua. Mrs. Smithies was the daughter of the late Colonel Saiuuel Nowlein, wdio commanded the body guard of the Queen, at the time of her deposal. Acknowledgments of gifts have been sent to Mesdames C. M. Cooke and C. X. Forbes; Messrs. L. J. Bouge, A. A. Brown, Bruce Cartwright, Jr., C. Montague Cooke, Jr., C. M. Cooke IK, H. E. Cooper, S. B. Dole, A. F. Judd, James Munro, R. Xui, G. W. Paty, A. Perry, J. W. Pratt, H. Roberts, W. S. Rycroft and G. E. Smithies; the Hawaiian Evangelical Association and the Twenty-tifth Infantry, U. S. A. While Prof. J. Macmillan Brown, vice-dean of the L'ni\ersity of New Zealand, visited the Hawaiian Islands, to continue Iiis Polynesian researches, it became the good fortune of the curator of ethnology to conduct the noted anthropologist to various parts of the island of Oahu. PaOIONATA. Dr. C. M()ntaj:(tie Cooke, Cttiator of I'tilinoiiata, reports for his (lepartir.ent as follows : During the past year, fewer shells have been added to the collection than in any year since the curator has been employed by the Museum. The reason for this is that a little more than half of the year was spent in preparation of a manuscript dealing with the Hawaiian Pupillidae. The whole of the Museum collection of our species belonging to this family ( catalogued up to December 31, 1917) has now been classified and arranged. .V large number of new species was found in the collection, as about two-thirds of the species and varieties dealt with are new. The numuscript has beeti slightly enlarged by Dr. Pilsbury and will appear shortly in the Manual of Conchology. Five thousand one hundred and ninety-eight (5198) specimens were catalogued during the year. These specimens are distributed in five hundred and sixty-seven (567) catalogue numbers. Specimens have been received from the following: Mrs. G. W. Bryan, Miss M. Burbank, :\Iiss :\I. Clough, Messrs. J. S. Emerson, C. F. Mant, A. Gouveia, J. C. Bridwell and E. R. Davis. [6] Bernice P. Bishop Museum. Occasional Papers Volume VII, Plate II. THE HOLMAN FEATHER CAPE. Director's Report for icjiS. 7 Dr. Cooke also prepared in the early part of the year, a pre- liminary catalogne of the Lepaehatinae and Tornatellinidae in the Bishop Museum. The paper, while illustrating the extent of his work in these families, serves the wider purpose of a eheck list of the species, and an indication of their distril)ution. The catalotjue appears in Occasional Papers, XDlume \ 1, Xumher 5. It mig-ht be mentioned that the rapid expansion of the collec- tions in this curator's charg-e, had so limited the space for the re- classifications made from time to time, that an additional rcjom was assigned to him — the room that had been used for instruments. BOTANY. The report of the Curator of Jiotany, Mr. C. N. Forbes, covers a period of two years, 1917 and 1918. ^Ir. Forbes observes that 191 7 was exceptionally dry on most of the Hawaiian islands, with the exception of Lanai. and was, on this account, a very favorable year for the botanical collector. On Kauai, during Juh' and August, the section west of the Waimea River, covering- a large portion of the country near the Kokee stream was explored. The region is so extensive and so cut up into numerous ridges and gulleys, that it was hardly more than skimmed over in the time that I was there. Tlie lower ridges represent one of the dryest habitats on the island of Kauai, while the vast dissected upland plateau known as the Alakai Swamp is a region of perpetual rainfall. In a region of such diversified hahitats, the flora is naturally very rich and varied and possesses additional interest as being the type locality for most of the species collected ])y Valdemar Knudsen and described in Hillebrand's Flora, and also many of Wawra's species. I received much kindly help from Hon. E. A. Knudsen, and through his assistance covered more territory than would otherwise have been possible. During Septemljer a short visit was made to Lanai as the guest of xMr. G. C. Alunro. Under Mr. ]\lunro"s system of plowing and seeding paddocks, coupled witli unusually ramy weather, the country was quite green. Owing to the decrease in the number of wild goats, the forest under- growth on the main mountain was mucli thicker tlian on my last visit in igi2. I was especially fortunate in lieing able to hnd a single specimen of Hrsfcro/iniiiiiia urhorcscciis in its type locality. Part of September and the month, of October were spent on INIaui, in Honokohatt valley, where T camped at the in-take of the plantation irrigation ditch. Several trips were made to the summit of Eke, a small isolated crater with a peculiar and very interesting bog flora. The flora of the valley itself showed the effect of an unusual drought. A fern, Dryoptcris cvaflicoidcs, which was very common [7] 8 niicctdi-'s /v('/i('/7 for KjiS. (Ill tlio steep sides oi the guleli, hun,t; down in dr\' festoons, while even patches of tlie comparatively seinixerophilic Clcichcnia linearis were with- ered in places. Trees showed varying degrees of resistance, species of Siraussia apparently' sntYering the nmsl. .Man\' thanks are due tn .Mr. Da\'id I'deniing for permission to camp i.i tlie valley. The most notahle ac(|uisition to the herl)arinm was a collection of plants nanietl hy 1 lille])rand, already mentioned in the Director's .Vnni'al Report for 1917. Rettiniino in the hitter ])art of vSepteniber, 191(8, after nine months leave of absence, the t^reater ])art of ]\Ir. Forbes' thne was occupied by the routine work which had accumulated. With the aid of a teni])()rar\- assistant the ])tilk of the material has been pois(.)ne(l, and mtich of it mounted. .\ brief monograph on the Genus Lage- nophora was published with Occasional Papers \'olume VI, Num- ber 5. The accessions to the herbarium for the two years are as follows : 1917. .-\. S. Hitchcock, U. S. Grasses hy exchange 104 C. M. Cooke. Oahu 8 J. C. Bridwell, Oahu 6 H. Graves, U. S. hy exchange 13 G. C. Alunro, H. I. purchase 134 J. W. Thompson, Oahu i L). Thaanum, U. S 25 J. AI. Lydgate, Ilillehrand collection purchase 589 CuR.vroR's Collection Oahu 316 Kauai 2765 Lanai 597 Maui 1300 Total 5858 1918. Mrs. A. H. B. Judd. teratological papaia i rUireau of Science, Manila, e.xchange 247 (j. C. Alunro. I^anai and Alaui 33 E. L. Caum, Alaui 2 J. F. G. Stokes, teratological papaia i .\. I). \\. Rimer, Philippine Islands, purchase 1318 J. R. Weir, U. S 13 Total 1615 [8] Director's Report for igiS. 9 ENTOMOLOGY. The Honorary Curator, Air. O. H. Swezey, lias been much handicapped through the departure of Mr. Bridwell, his former assistant in the Museum. Mr. Sweze}- has comparatively little spare time to devote to collections of the Aluseum. nevertheless he has been able to increase them during the year by 1260 specimens, comprising 415 species. The courtesy of his employers, the Ha- waiian Sugar Planters" Association, in permitting him to take the honorary position and to devote certain hours to the work of the Museum, is very highly appreciated. The Honorary Curator re- ports that all of the Helms collection has been placed in the Museum cabinets, with the exception of the New Zealand Coleoptera vet remaining to be transferred from the old boxes containing them on arrival. The Hawaiian insects formerly in the gallery of Hawaiian Hall have been removed to the metal cases, so that now they are con- venient of access for reference or study. ■ During the year man}- insects were collected m the native forests and prepared for study. The localities and the approximate number of specimens and species follow : Specinifns Species lao Valley, Maui, August 8th 200 70 Haiku, IMaui, August 27th 6o 15 ?Ialeakala, Maui, August 25th, 26th and 29th 303 70 Kaivviki, Hawaii, September 22nd 260 75 Palolo and Mt. Olympus, Oahu, October 20tli 160 6c Pacific Heights and Kaumuahona, Oahu, Nov. 17th.., 60 35 Lanihuli Ridge, Oahu, November 24th 120 40 Kuliouou, Oaliu, December 22nd 100 50 1260 415 This material includes many species not hitherto represented in the collection, also a number of species not previously collected or described. When time or circitmstances make it possible to work up this material in connection with the 3250 specimens collected by Mr. Bridwell in 1917, it will make a beginning towards supplying tlie species lacking in the Hawaiian collection, which contains but 64 per cent of the number of species listed in the Fauna Hawaiiensis. bi February, 1918, the para-types of the Hawaiian Heteroptera were lent to Air. E. P. Van Duzee, Curator of Entomology of the Museum of the California Academy of Sciences, San Francisco, for his assistance in working up the Hawaiian Heteroptera in the collections of several persons and institutions of Honolulu. (9] 10 J li rector's h^c/^ort for I(^l8. IJl'.KAKV. The Kilirarian. Miss E. l'>. Ili^i^ins, reports an tiuijrccedcnted iiuiiiher of iK'\vs])a|)C'rs in tlie native lan^ua^c received from var- ions sonrccs dnring" tlic year. L'litil tlie l)i.\i;iniiin.L; of llie tuciitii.'tli eHMit\n'\', the \-cTnacular news- papers coin])risctl alnmsl llu- sole natixe literature piihlislied, and they covered a transition jieriod in nati\e lil'e hut Ii.t;lul_\' touched u])on l\v foreign ohserxers. Ileing also tin.' niean> (jf tlu' nati\es" exposilir)n of their own eusnans and folk-lore (although their intellectual leaders were ec'ucatcd and inspired by foreigners) the Librarian has made many efforts during the past few years to complete the early hies. The largest contri- bution so fear recei\'e(l came this year from the liasement of Kawaiahao Church in Honolulu, from which thirty-four sacks and thirty-nine trunks of papers and native school liooks were taken. These were cleaned and sorted, .and the duplicates — the greater bulk — were returned. A smaller accession yielding greater proportionate results was the i)ur- chase of the files of a native minister which included in addition tc) many other fragmentary sets the following thirty li\e .-omplete or nearly complete volumes: Ka Nupepa Kuokoa, 24; Ke Alolia Aina, 2; Ka Makaainana, 7; Ka Loea Kalaiaina, 2. Local periodicals are now represented in the Library b\- seventy-five titles, classified as follows: English. 41; U.awaiian, 32; Portuguese, i ; Chinese, i ; the earliest date is 1834. The tiles of scientific journals ha\e ])een added to by purchase, gift and exchange, but numerous \acancies were still left as many parts and numbers were out of ])rinl. The lack w;is partl\' supplied by approaching the contributing authors — a mode suggested by Dr. Herbert E. Gregory on a visit early in the year. Twenty-one journals were added by subscription. The distribution of scientiiic journals in the ^Museum Library, is appro.\imatel\- as follows : Archaeology, anthropology and ethnology 41 Natural History 38 Botany 15 Conchology i Zoology, General 26 Ornithology 2 Mollusca 2 Microscopy 1 Geography 6 General Science 15 Geology 7 Museums 1 1 Miscellaneous 9 174 LioJ Director's Report for iqi8. it The following institutions have l)een added to our list of exchanges: University of Chicago; University of Illinois; Newark Mviseum ; Societe d'Etudes Oceaniennes. This last named society has 1)een recently formed in Papeete for the study of the Pacitic. Acknowledgment of author's separates, and other items, is made to the following: Messrs. L. J. Bouge, C. Montague Cooke, Jr.. W. M. GifYard, Ivan C. Hall, A. F. Judd, Vaughan MacCaughey, Sidney Powers, T. G. Thrum and Harold Wright, and the Haw^aiian Association through Mrs. R. W. Andrews, A miscellaneous lot of books and pamphlets was purchased from the Rowell Estate — 50 titles, including some duplicates. Among them were MS notes on the Hawaiian P>il)le, 1843, by Rev. G. B. Rowell. Summary of accessions for the year is as follows : Volumes Parts and Pamphlets Exchange 181 411 Purchase i/O 217 Gift 13 56 Total 364 684 Grand Total— 1048. This summary excludes incomplete volumes of local newspapers re- ceived, dating from 1846 to 1918, and representing 65 volumes and 47 titles — 19 English, 27 Hawaiian and i Portuguese. More than 200 volumes have been prepared for the binder; of these 165 volumes have been bound. MODELING. Dtirini^- the year, Mr. Thomp.son, Artist and Modeler, finished 70 models: 16 of fishes, i of a mammal, 16 of frnits, and 2i7 o^ ethnological specimens. He has 21 unfinished models on hand. He has also painted and ]M"epared three coconnt crabs from Fanning Island, An excellent representation was secured of a porpoise 7 feet, 7 inches long, weighing 375 pounds. EXHIBITION HALLS Mrs. Helvie, Superintendent of Exhibition Halls, reports that the attendance of visitors has kept up very well in spite of the lack of tourists — the military visitors from the various posts being greater than usual. The total attendance for the year was 14,029, a decrease from 15.145 the average for the five previous years. The [iij 12 Director's Rcf'orl for loiS. daily average was 54.4, — the halls lia\iii.ii ])vvn n])vu for 258 days in the year. The attendance ])\ nmnths. and classified according to race, is given in the acconi])an}ing lahle. The total includes 1367 j)ni)i!s, representing 4*) classes from schools in I lonolnlu. TABLE OF ATTKXDAXCl-: 1918 ^ ^0 1 a CO January . February March ... Ai)ril May June July August ... September October November December Totals... 025 IHS 76 :r> 787 117 HI 14.S 602 88 47 "2 561 194 64 S(l 442 221 77 1(1(1 , 410 92 109 102 584 59 81 62 491 91 58 76 555 239 97 90 426 181 98 118 408 130 72 105 561 292 29 70 6452 1867 919 1098 241 22(1 1:1:! 10S7 44 21 2(11 14 21 108 2(1 21 212 24 23 268 45 21 196 11 22 273 33 21 191 3 22 3446 247 258 54 8 1205 70.2 1414 47.7 1048 50.2 11(16 93.9 1971 44.2 1928 43.5 914 40.2 952 (•0.2 1294 46.9 1030 48.7 1021 52.1 114G 54.4 14029 The steel feather-cloak case, ordered closed by the Trustees at the beginning of the year, was reopened in December for semi- monthly exhibitions of half an hour each, under the charge of Mrs. Helvie. The establishment of a War Saving Stamp sub-agency in the halls was a great success. I 12] The publications of BBRNICB PAUAHI BISHOP MUSEUM include : MEMOIRS, \^olumes I-VII. OCCASIONAL PAPERS, Volumes I-VII. MISCELLANEOUS PUBLICATIONS, Numbers i-6. A descriptive list of publications with prices will be mailed on application to the Librarian. • H'j. .- • -v-, ..■• ,■•<'.-■• )*:»>T; v-T'-jS-J «W OCCASIONAL PAPERS OF THE BERNICE PAUAHI BISHOP MUSEUM OF POLYNESIAN ETHNOLOGY AND NATURAL HISTORY Vol. VII— No. 2. The Languages of the Pacific By J. MacMillan Brown honolulu, hawaii Bishop Museum Press 1920 OCCASIONAL PAPERS OF THE BERNICE PAUAHI BISHOP MUSEUM OF POLYNESIAN ETHNOLOGY AND NATURAL HISTORY Vol. VII— No. 2. The Languages of the Pacific By J. MacMillan Brown honolulu, hawaii Bishop Museum Press 1920 The Languages of the Pacific' By J. Macmillan Brown One of the fallacies that (lo,y the science of language is that there are three types of language, isolating, agglutinative and inflectional, separated strictly from one another. It was one of the too early generalisations of ]\Iax Midler wln^, coming from Germany with a knowledge of Sanskrit, ruled with absolute author- ity the science of philology in the English speaking world during the latter part of the nineteenth century. This theory together with the idea that all classical myths have a philological origin, is now discounted. It is found that almost all languages have some trace or relic of each type. The Chinese is taken as the typical instance of the isolating language ; each word may be used in various grammatical relations without any formal element to indicate these relations. But modern English has become practically an isolating language with only particles to indicate these relationships and a few relics in the pronouns of the old inflectional system. Polynesian is on the same footing; a worrl may be a noun, a verb or an adjective without any distinctive formal mark : and particles indicate the relationship, whilst in the pronouns, as in English, there remains a few relics of inflection. The Japanese is the Pacific (^cean language that best illustrates the agglutinating t}-pe. The formal elements retain so much of their original indei)endence that a(lverl),s and honorific words may he thrust in between them and the words they pilot grammatically. Ihit the language has nnich that may be said to be inflectional and has some trace of the isolating. English, likewise, shows a tendency to the agglutinative in, for example, the frequent separation of the formal to of the infinitive by an adverb, or even a phrase, from the verb. So in Polynesian the ia, a. that added to the verb makes the passive, shows in some groups a tendency to assert Lecture delivered before The Hawaiian Historical Society September 5, 1918. [3] (15) l6 '/'//(' lAin_i:;iiao;cs of fhc Pacific. its indcpendctu-e and be separated from its verb. ( )f the inflectional t\'pe the best instances are fonnd in the Indi i-b.nn ijjcan tongues. Latin is hiL^hh inllectional, (.reek still more .so, and vSanskrit most of all. in Tolynesirm the inflections of the dual and i)lnral pers uuiknu or kakoii and of the third hihoii: this ai;'ain shows its origin in kohi, three. These inflections for the plural were manifestly formed at a most primitive linguistic stage when the ancestral speakers of Polynesian did not count beyond three ; one and two were delinite, three was all beyond, the indefinite. This must have been before they launched out into the Pacific, for there ior the first time they counted up to five ; //;/;(/ for five is practically imiversal in the so-called Malayo- Polynesian languages: but they had been able to count up to four before they left the sphere of influence of the Indo-European languages. "(3ne" varies most of all the numerals. IV)lynesian rua for two is the Latin duo, English t-Zi'o ; for the sound-law that makes / or r and (/ interchangeable existed as strongly in early Indo-European as it does in Polynesian and Malay. Latin lacr\um. Old Latin dakniiiia. is Greek dakni, Gothic tai^^ra. .\nglo-Saxon tcai^^or. tear. Polynesian torn or foli( is the uncontracted form of Latin trcs, German drei. English three; whilst Polynesian wha, four, is Latin quatuor. Sanskrit catvar, Anglo-Saxon fcoTcr. There is no trace of "five" in Polynesian or of lima as a numeral in any Indo-European tongue. Yet the Polynesian must have retained some consciousness of the old European for one. (Latin junis. Old Latin oiiios), for in counting on the second hand, six is ono, i. e., number one of the second hand. JJ'hitii for "seven" retains a trace of "septem" (from sa-pita). It is probably a modification of "u'hiti," to cross over, Hawaiian Itiki, to come, to rise, just as "trcs" is from "tara" to cross over. The Polynesian forms for eight (warn) and nitie diva) belong to that by no means uncommon method of counting from the highest number counted downwards by subtraction ; thus e. g. the inimbers between five and ten in Yap are ten minus one, two, etc., which may be compared with the Roman numeration IV, IX, XL, XC. Malay also expresses 99, 98, 97, etc., by "hundred minus i, 2, 3, etc." IVa is a common Poly- [4/ The Languages of the Pacific. 17 nesian word for interval or "space between"; 8 (zvari{)=io — (space between or minus ji*, 9 (iz^'a )^=onii — from 10. Lima, the word for five, is not without trace in European, though not as a numeral. There is a European root form, "ri>iia," meaning "row, iniiiibers, z'erse." In Old Norse riin=icalciidar, I'd'se. In old Irish riui= n limber. In Old High German Rim^roi^', niimher, German Reim= English rhyme. Compare the Greek arithiiios, a number, iierifos, countless. It is from the same root as ra, to arrange, to fit, Latin reri, to think, ratio, read, reckon, hundred. The original form is ra, Maori rarangi, line, rank, row. The ri form is seen in zchakarite, to arrange, put in order, Hawaiian like, to be like. Many of the languages use lima not only for "five" but for "hand," evidently meaning "the counter,"" but in Maori and Fijian the word for hand is "ringa," implying that "ma," was felt to be an affix, just as nga is. We may say then that the Polynesian ancestors were only feeling their way up first beyond three and then beyond five. They were feeling their way towards "tekau" which first meant "7/;^ company," "the lot." and, when they counted beyond, came to be "ten" or in some "eleven." (Compare kaii, company, \ot,= iigahii-ru, Gilbertese tengaiin=^io, Tongan //^bundle, kehiii, flock. ) The Hawaiian "umi" easily meant at first "the measure."" The usual Polynesian ngahuru, for "ten," becomes in Malay "sapiiloh" by prefixing "^a"=one, to piiloh, ecjuivalent to Jiiirii, the hair. Sapuhh means the bunch of hairs, nga, the plural article in Maori, being replaced by sa, and hunt, brushwood, coarse hair, in English "wool." The true classitication of linguistic affinities is not by their grammar, but by their phonology, i. e. the range of sounds and sound laws that belong to them. The organs of speech do not change unless the climatic environment is changed, or the mothers. To shift from the temperate zone to the tropics relaxes all the tissues, including the tissues of the speech organs ; to shift in the opposite direction gives them greater tensity and vigor. And if at the age of the moulding of a man, i. e. from infancy to seven years old, he is set in a difl:'erent speech-environment from that of his ancestry his speech organs will be different. It is the mother or nurse that creates the phonological capacity of a man or woman. The speech organs are set practically for life during the first seven years, the period when it is the mother that is the dominant influence. [5] iS I' hi- I j)ii;^uai^cs i>i I he' I'aciUc. If tliereh irc Uktc i^ a (lillereiicc l)L'l\\cL-n the Polynesian phonology and thai ol those to the west ol it we may assume that it was a change of mothers that caused it: t'or right through the seven thousand miles from Tonga to the coast of India the climatic enviromnent is j^ractically the same, moist heat governed hy regular- 1\- hliiwing winds. Xow the phonology of the Polynesian dialects dilTers hy a whole world from that of all the languages to the west of it. The former have only twelve to fifteen sounds, the five vowels and seven to ten consonants, the most primitive outfit that any language in the world has. As soon as you step out of Polynesia westward, say from Tonga to the neighboring Fiji, the language has from twenty to thirty sounds, and this holds right to the coast of India and all through India. Further, there are sounds to these languages to the west that no Polynesian could l:)y any training be made capable of pronouncing, nay that no European could, i.e. the speech organs are ab.solutely different in the two regions. One instance is the c}=kpz^< of Melanesia. I)Ut the fundamental principle that divides Polynesian plionologically from all to the west is that it must close a syllable or a word with a vowel, and it cannot pronounce twd consonants together. All the languages to the west can not only close a word with a consonant, but many of them (including Malay) prefer to do so. The only tw'o languages in the Pacific Ocean that have the same phonological laws are Japanese, away to the north- west, and Quichua. away to the southeast ; but the former is grammatically of a different type, the agglutinative, and the latter, though almost grammarless, like Polynesian, has inflections only in the pronouns, including the strange Polynesian characteristic of a different form in the first person for the plural that includes those spoken to and the plural that excludes those spoken to. I have found Init a small i)ercentage of Ouichua words or roots the same as in Polyne-sjan, while the range of sounds in Japanese is nearly the same as in Polynesian. There is one other characteristic of Polynesian phonology that almost puts out of court the accepted theory that the Polynesian languages came from India or the Malay archipelago. They are divided into / languages and r languages. In P'olynesia / has a little of the trill of the r and the r has somewdiat of the licjuidity of 16] llic iMiii^uai^cs of the Pacific. 19 of the / so that it is easy for one to pass into the other. But the southern groups have a preference for the r sound, so that the missionaries have always written this consonant in their language as r, whilst the northern have a preference for the / sound ; these are Tonga, Samoa, Futuna, Tokelau and Hawaii : all the rest except Marquesas use r; that group has a rule neither r nor /. If we step out of Polynesia and g'o west, every language uses both r and /. I should like to have explained to me how, if the Polynesian languages came east into the central and eastern Pacific, they were • able to divide off the / speakers and the r speakers after coming through seven thousand miles of languages that used botJi r and /. Undoubtedly in the now submerged fatherland, Hawaiki, probably lying well to the south of the equator and to the east of Samoa and Tonga and the Tokelau group, the peoples in the north and north- west of it preferred the /, those in the south and southeast preferred the r; though the preference had not grown as pronounced as it is now, it had been made perhaps through that contradictoriness which dictates the fashions of neighbors, probably more pronounced because the northern tribes were nearer the equator and preferred the sound that needed less tensity and energy in the organs of speech That Hawaiki was to the east of Samoa and Tonga is evident in the fact that the spirit land of the two groups is not Hawaiki, but Bulotu, which is probably from the Fijian biilubulii, the grave, and /;/(///, the abode of departed spirits, modified by the Polynesian purotu, pure, pleasant, agreeable, soft, delicate, beautiful. Burotn is in Fiji the residence of the gods and the place of spirits; so it is in Samoa and Tonga. Next to the northern tribes of Hawaiki must have lived the Tahitians, for they, like the Samoans and Hawaiians, eliminated the guttural k that had come with the primeval Polynes- ians from the colder north and continued in all the languages that, like those of Tonga, the Maoris, the Paumotus and the Austral Islands, Mangareva and Easter Island, drifted further south into a colder zone. But, to show the influence of climate on the organs of speech, the Hawaiians, when they got up to the borders of the temperate zone, though they did not restore the primeval k, began to substitute for it the t of all the other Polynesian dialects. The Marquesans had already begun on Hawaiki to avoid the rolling r and the liquid / and when they reached the steep-to islands in which 20 'J III- lAiiigitagc's of llie Pacific. they afterwards sellled, they ahnost thoui^h Hdl ((uite completed the ])rocess; there are onl_\ a tew words in their hm^ua^e that retain the r. They also showed the same tendenc} as the Hawaiian to substitute k for t, though the tendency did not proceed to the full length of the northern language. Kanlni is the Marquesan salutation ecjuivalent to the Hawaiian aloha. Yet the k sometimes disappears in Marcpiesan ; for it is only from eight to ten degrees .south of the equator and has sufficient moist heat to create languor in the organs of speech. Thus we have in the different branches of this, the most primitive of languages, fully developed a phonological law- as strict as Grimm's Law amongst the Indo-European and far wider in its application; it dominates not jnerely the explosive consonants, (t, p, k) as in the Indo-European language, but the liquids and sibilants, r, /, .y, sii and /;, and even the nasal con.sonants, ;/, ng. If we know the form that a word common to most takes in any one of the Polynesian languages, we know the form it takes in every other, provided we know this strict sound law. There is one exceptional sound, ch or iz, w'hich appears in Tongan and Moriori, w'hilst Tongan has a h instead of the usual p. This must be due to the long intercourse of Tonga with Fiji which had a phonology more Melanesian than Polynesian. Strangely enough this ts sound also belongs to Japanese, whilst the ch form of it belongs to Ainu. But b is purely Fijian and is in fact in that language nih. This regularity of consonantal change in the various dialects of Polynesian is a characteristic that completely differentiates it from all those to the west, the Micronesian and especially Papuan and Melanesian. In these there is phonological chaos in their rela- tionships. As a rule neighboring villages in Melanesia and Papua cannot understand each other's language though only a few miles apart, whilst the Maori can understand the Rarotongan or Tahitian or Hawaiian after a brief acquaintance with the con.sonantal changes. And in Hawaiki this tendency to consonantal decay must have been widespread, the change that is complete in one or more of these groups occurs sporadically in all the rest. Take as an example the loss of h which is universal in Tahitian, Samoan and Hawaiian. In Maori it is quite common to find two words meaning the same, one with the h. the other without it; two or three will [8] The Languages of the Pacific. 21 suffice, kapo, to snatch, and apo, to grasp: kita, tight, fast, and /7a, tight, fast ; and koti, to cut, and oti, finish. Fornander points out how some, if not all of these, are paralleled in the Indo-European languages. The substitution of .!> in Samoan for the h of the other dialects occurs also in Sanskrit, Latin, Gothic, Iranian, Greek and Cynu-ic. The change of ng of Samoan, Maori and other southern dialects into ;/ in Tahitian and Hawaiian has its parallel in the substitution of ;; in Slavonic for the ng of Sanskrit, Zend, Latin and other European tongues. The r was interchange- able with / in Indo-European as in Polynesian, and both were often changed into d in both linguistic spheres. It is not infrequent to find roots in both spheres that have forms with and without the to b as it shows the change from t to ch or t::. Primeval Aryan as it is seen in Tocharish has the same range of sounds as Polynesian and practically the same sounds and number of sounds. It showed the same tendencies to drop k, to make t and k interchangeable, to elide r or make it interchangeable with / or d, to substitute ,? for h, f for ich, and // for ng. Its fundamental vowel was a; and so it is in Polynesian. Look in the Hawaiian dictionary and you will find ten [iij 24 '///(• Liiiii^itagcs (if the Pacific. times as many words with a as the vowel as those with any one of the other four vowels, c, i. o. it. h'rom a in l)oth languatJes there are two series of nuitations (• and /' on the one hand and o and ;(- on the other. if von look into l'"ick's indo-I'.nropean dictionary, yon will find ten times more roots with (/ as the vowel than those tliat have c or /, o or ;(. And as far as I can judge hy analysis of the roots of more than one syllable or two letters, they are all reducible to roots of one or two sounds, a pure vowel or a consonant and a vowel. In other words i)rimeval Indo-European had the same sound law as Polynesian, i. e. it preferred to close a syllable or word with a vowel. There is one other point that the discovery of Tocharish settles, it is that the western European type of language came east into Asia. Aryan languages are divided into two sections by a line drawn from the Baltic to the Black Sea. West of that all the languages retain the original A' sound. East of it all reduce it to a sibilant, at least all till Tocharish was discovered. The former are called by philologists the centiiiii languages from the Latin word for one hun- dred ; the latter the Sato languages from the Sanskrit word for one hundred. Tocharish retains the k unchanged to s, and must there- fore have come east long before Sanskrit hived ofif and traveled into Asia. Polynesian also retains the k unchanged into jr, and it too with the same sound-range as Tocharish and the primeval Aryan languages must have traveled from Europe west of the line between the Baltic and the Black Sea through Asia, long before Sanskrit began its long migration into India or even began its elatorate inflectional system. That inflectional system had begun before it completely separated from its cousins ; for many of its inflections have close kinship wdth those of Greek, Latin and the Teutonic languages. Even Polynesian, which shows an extremely primitive beginning of inflections in the personal pronouns (the dual in ua and the ])lural in on ), must have hived off and gone east before the inflectionalism had developed to any great extent. There could have been little or no formal grammar, as we can see in Tocharish ; the same word could be used as noun, adjective, adverb or verb; and particles supplied the cement or binding element of the sentence. Of course every dialect of Polynesia has a large percentage of its words and roots pecidiar to itself ; Hawaiian has, I should [12] The Languages of the Paeifie. 25 think, at least thirty per cent of such, but this is no proof of any ahen infiltration, but only of mitrrations from the sinking fatherland Hawaiki, to the group, separated by so long intervals of time as to allow of the disuse of one set of words in the mother tongue and the loss of aui^ther set in the new land. For they have all the same phonology, figurative application and transparency of composition that distinguish all the Polynesian dialects. The languages of Melanesia and coastal Papua, away to the west of Polynesia, have only a small percentage of their vocabularies in any way to be identified with Polynesian words, and as a rule these are greatly mutilated and often difiicult to recognize. I gave some few words in my previous lecture, which going right through to the Malay archipelago yet found their derivation only in Polynesian; as e. g. bia or pia the sago tree, but in Polynesian "exudation" from pi which is used in that language in the sense of " to exude." I will add one more; the Polynesian wahinc, a woman, comes from zva^^ "set apart" and hine. "a girl," but it goes away west into Indonesia in many dififerent forms as e. g. fafeii, vaiiie, aiiie. babiiieh. I could easily give scores of others. I doubt greatly if the implication in the term "Malayo-Polynesian" that these languages are all akin is correct. For though they are to some extent grammarless like Polynesian, they have much more formal grammar than Polynesian. In the Melanesian and coastal Papuan and to a small extent in the Micronesian and Indonesian languages there is a shorter form of the personal pronoun used as an afiix to the noun. These are so much more primitive in their linguistic and intellectual development that they cannot think of a thing but as belonging to some personality ; it is always mine or yours or his. The Polynesians have no mental primitiveness of this kind, they can think of a thing in itself and apart from its possession by a person. So in the Polynesian dialects (chiefly in Hawaiian) there is only a trace of a grammatical habit that is found largely in the Indonesian languages and is almo.st universal in the languages between Polynesia and the Malay archi- pelago. They cannot use the numerals except with classifying particles ; flat things have one special particle to themselves when being counted, and round things another aixl so on. A third characteristic of those languages to the west is the use of an infix, i. e. the insertion of a significant syllable right into the heart of a [13]' 2f> The /,(/// i^;/(?_i;('.s- (if Ihr J\icific. word. I '()1\ iK'sian and Aryan show no sii^n of this. These char- acteristics rcN'eal a (HfTerent linmii>tio attitude of mind from Poly- nesian and Indo-European. The hn^nistic attitude of l'(il\nesia faces nortli towards Ia])anese and Ainn which have ,L;ot no such restriction on their use oi noun.s and numerals. That the Polynesian vocahtilary looks also to some extent in that direction will he ai)i)arent from a few examples. ( i ) ilawaiiau //;///, kalo tops for planting", Samoan uU, sprouts of taro. .\ino chi tirip. the Japanese taro-vam, Ja])anese uri, a melon. (2) Aia, thimi;!] l\cv. J. M. I^\(latc informed mc, that he renienil)ere(l collectint;" this species at Kawaihae, and calhn^' I linel)rand"s attention to its jjccnl- iarities. It is (hfficult to nndcrstand \vh\- llint'l)rand slionld send specimens to Mann without retaining- (hiphcates, l)nt Mr. J. \\ l\ock informs me tliat he saw no specimens of 1 liljiscadelphns in tlie liinel)rand cohection at HerHn. As tlie native vei^etation at Ka- waihae has ])ractically (hsappeared, the s])ecies is probably extinct. Viola kauaiensis (Iray. far. wahiawaensis, var. nov. PLATE IV. Tliis well marked variety differs from the species in its leaves which are cuneate at the base, never truncate or reniform. As in the species the petinles \ary s.;"reatl\' in lens^th. and the fragrant dowers vary in size. All llie plants oI)ser\ed at the Waliiawa bog belong t(* tlie variety, tliose at the type locality, mountains aliove Waimea, Alakai swamp, and Leliua makanoe, to tlie species. Type in the B. P. B. ]\I. Herbarium, collected at the Wahiawa swamp, Kauai, T. H.. August, IQ09, by C. X. Forbes, X"o. 166, K. There are two unmnnltered specimens in tlie B. P. B. "SI. Herbarium, collected at the \\'ahi;i\\a swamp 1)y Rc\'. J. M. L>'dgate. Cyrtandra olona, sj). nov. I'LATE V. Stem 6 dm. higli. unbranched. glal)rous in all parts. Leaves opposite, brnadly ovate in outline, obtusely rounded at the apc.x. truncate at the base, a few sliglitly uneven-sided, crenate except at tlie l)ase, glal)rous, \'ery dark green alio\'e, paler I)elow, nerxes prciminent. 15.4- 16 cm. long, 12-12. 5 cm. wide, petioles 6-1 1.5 cm. long. Idnwers ( se\eral ? ) on a cdmmon peduncle of 7 mm.; the pedicels of the same length, liracts not seen. Calyx cu]i-shaped, sJTortly 5-clcft witli deltoid, acute lol^cs, glabrous, 8 mm. high. C(Trolla small, white, sliglitly curved, i.i cm. high. Ovary glabrous. Fruit not seen. Type in the !'>. 1'. I'.. M. Herb.-irium, collected at the Wahiawa moun- tains, Kauai, 'P. H., .\ugust. 1909, by C. ,N. Forties. No. 213. K. The mature. l)luish-gTeen. and the coarsely veined juvenile leaves are rather suggestive of the Olona. Tiniclutrdia lati folia. The species is distinct from any Cyrtandra known to me. but may prove to l)c distantly related to C. cyaiicoidcs R. P. V>. AT. Herbarium, collected at the Koolauloa moun- tains between Punalvui and Kaipapau, Oahu, T. 11., Nov. 14-21, 190Q, by C. N. Forbes, No. 256,?. (). This species differs from C. wdiolaiiii W'awra in its cordate leaves, differently shaped bracts, and somewhat in the pubescence of the corolla. ('. f^rof^iiujua has calyx lobes broader at the base than those of t". ■:c(ili)laiili. wliich are linear. The two species are closely related. Cyrtandra Munroi, sp. nov. TLATE VIII. A shrub with rust-colored tomentose liranclies. Leaves opposite, elliptical in outline, acuminate at the apex, acute at tlie Ijase, serrate, hirsute on both faces, especially along the rust-colored nerves and petiole, dark green above, paler below, 14.2-15 cm. long, 6.8-9.4 cm. wide; petiole 5. ,1-5. 7 cm. long, biflorescence rusty tomentose, the common peduncle 2.2-4.5 cni. long, the pedicels 1.4-.^.! cm. long, tlie bracts ovate, acute, i cm. long, 7 mm. wide, 1-3 flowered. Calyx cup-shaped, incised to the middle, lolies broad, acute, hirsute, 9 mm. high. Corolla erect, hirsute on the outside, glabrous within, projecting beyond the calyx; the tube 1.4 cm. long, the small lobes sub-equal. Ovary glabrous. Berry conical, rusty holosericeous, 2.3 cm. long, i.i cm. in diameter. Type in tlie B. P. B. M. Herbarium, collected at the eastern end of tlie mountains of Lanai, T. TI., June, 1913, l)y C. N. Forbes, No. 235, L. Also collected at the Pali above Waiopaa, Lanai, T. H., March 25, 1915, l)y G. C. ■\Iunro (No. 417). ('. Munroi belongs to the section Crotonocalyces of llillebrand, resembling" certain forms of C'. f^lafxphylhi, from Hawaii, in the shape of its leaves, but is more closely allied to the si)ecies belonging to this section which occur on the island of Maui. The leaves of the specimen collected liy ]\Iunro are ttne(|ual sided. Cyrtandra Georgiana, sp. now PLATE IX. Shruli. Brandies covered witJT a short hirsute pubescence. Leaves opposite, elliptical to obovate in outline, acuminate at tlie apex, cunate at the base, hirsute above, rusty tomentose below, especially along the veins, serrate, 16- 16.4 cm. long, 7.4-9.8 cm. wide, with pubescent petioles 2.7-4.5 cm. long. Inflorescence rusty tomentose, 3-6 flowered, some pedicels bearing 2-3 flowers, of wliich the prim.ary ;nid secondary pedicels combined ecpia! in length tlie primary single flowered pedicel, the common peduncle 3.9-4.3 cm. long, the pedicels 2-2.8 cm. long, the bracts ovate, .acute, i cm. long, 6 mm. wide. [6] A^ezc Ha7caiian Plaiifs. 37 Calyx deeply parted to near tlie base, the sepals ovate-acuminate, narrowed at the base, rusty pubescent on both faces, i cm. long, 5 mm. wide. Corolla (incomplete) sparingly pubescent on the outside, glabrous within. Ovary glabrous. Berry ovoid, glabrous, 1.2 cm. long, 8 mm. in diameter. Type in the B. P. B. M. Herbarium, collected in the mountains of Lanai, T. H., September, 191 7, by C. N. Forbes, Xo. 348, L. Although this species belongs to the Schizocalyces of Hille- brand, it is more closely related to C. Munroi Forbes of the Cro- tonocalyces than to any other described species. Clermontia Samuelii sp. nov. PLATE X. While collecting botanical specimens on the northern slope of Haleakala, Alani, T. IL, between the Keanae Gap and Kipahiihi, at an elevation of about 6500 feet, the writer observed a very beauti- ful species of an undescribed Clermontia, in the dense rain forest a sliort distance west of an old cone called Wai Anapanapa. In color and shape the Howers resemble those of the common form of C. g'randiflora Gaud, which is common on the western slopes of Hale- akala in the vicinity of Idvulele, but in other characters the plant dififers widely from any of the described species. While related to C. grandiflora Gaud, in character of flowers and in size of leaves, it is at once separated by the character of its inflorescence, and the pubescence of the leaves. This species is verv distinct from C. kakeana ]Meyen and related species, having pubescent leaves ; while the flowers at once separate it from C. parviflora Gaud, and its allies. It was the only Clermontia seen in the type locality, but the forest is very dense there, and practically unexplored. I take pleasure in naming the species in honor of Mr. S. A. l^)aldwin of Makawao, Alaui, who very generously aided the Mu.seum in the exploration of Haleakala, and to whom the writer is indebted for many kindnesses. Clermontia Samuelii has the following characters : Shrub, 12-18 dm. high, densely branched from the base, the branchlets covered with dense brownish pubescence which is purple on new growth. Leaves obovate to elliptical in outline, acute or sub-acuminate at the apex, acute or cuneate at the base, finely serrate with callous pointed teeth, witli fine soft whitish tomentum on both faces but most so below, very dark green above but not shiny, dark but much paler below, rather thin chartaceous when dry. 6.4-7.4 cm- long, 2.t,-t,.2 cm. wide, on pubescent petioles of i. 4-1.6 38 Nezv Haz^'uiian I 'Units. cm. ill IciiKlh. Inflorescence a 2-5 flowered cyme, llic pedunclo 13-8 mm. louK, marked liy scars and one or more pairs of linear bracts ,5 mm. in length; the pedicels filiform, drooping, 2-1.6 cm. long, with subulate bracts near the base. Flowers slightly curved or sickle-shaped. Calyx glabrous, thin, slightly shorter tlian the corolla, the lobes linear acuminate, the upper one splitting to near the l)ase, tlie lower ones splitting one-third the distance of tlie tul)e, a delicate purplisli-jjink color; the ovarian portion turbinate, bright green, i.i cm. long. Corolla 4.6 cm. long, 6 mm. wide, glabrous, slightly lighter colored than the calyx. Staminal column and anthers glabrous, purple. Mature fruit not seen. Tlie type is mounted on sheet No. 1225, ^I, in the 15. P. L!. W. Herbarium at Honolulu, and was collected by C. N. Forbes on the north slope of Haleakala, Maui, T. H., at an elevation of 6500 feet to tlie west of Waianapa- napa, August 22, 1919. Another specimen is mounted on sheet Xo. 1215. M, and was collected near the same station on the same date. Argyroxiphium caligini .sp. iiov. PLATE XI. iJuriny the month of May, iQio, while hotaiiiziiio' on I 'tut Kiikiii the summit of West ^laui, a small colony of a species of Argyroxiphium was observed, but unfortunately all the plants were .sterile. While the plant has since been observed by Mr. J- F- Rock and others, no one seems to have collected Howering- specimens. - In Se])tember of 1917 while collecting on Eke an i-solated cone which is about 1000 feet lower in elevation than Puu Kuktii. this species of Silversword was found to be one of the most conspicttous plants of this boggy habitat. The plant proves to be a dilTerent species from that found on either the uplands of the island of Hawaii or of East Matii. It differs frotn the other described species in its much smaller size. The silvery leaves are about the width of those of A. virescens Hbd., but only half the length. The flower heads are smaller than those of A. virescens, which are described l)y llillebrand as smaller than those of A. sandwicense, DC. Field experience has proved that size of head is a variable character, the flower heads of many plants being fully as large as those of A. macroce])halum Gray. Another form of Argyroxiphium on Eke differed in having somewhat greenish leaves which however were "J. F. Rock in Annual Rept. Bd. Agr. and For.. T. If., for 1910, p. 80 (1911): also for 1918, p. 53 (1919). 181 Ncxc Hawaiian Plants. 39 niuch more pul^esceiit than tliose of A. virescens Hbd, and also differed in the glandular i)ubescent, not silvery raceme. The Pun Kukui plants probably belong to this form. This plant was not observed in any of the extensive bogs on Haleakala, and is probably not a cross between A. macrocephalum and A. virescens or Wilkesia Grayana. Argyroxiphium plants growing- in the bogs on 1 Taleakala are not depauperate. A detailed description of Arg'yroxiphium caligini follows : Herb with a stem of 2-2.5 cm. in diameter, soon tapering into a simple foliose raceme of 3-4 dm. in length. Leav-es linear, 10.5- 12.2 cm. long, 4-5 mm. wide, densely silvery-sericeous. Raceme simple; the rachis, leaves, and foliar bracts silvery-sericeotis ; the pedicels slender, green, viscous- pubescent. 5-9 cm. long. Heads nodding, 1.5-1.7 cm. in diameter. Invocural bracts viscous-pubescent, green, lanceolate, acitminate. Ray flowers few, ligulate, the ligules bihd, 3-4 mm. in length, i mm. or less in width, purplish colored as are the disk flowers. Disk flowers with the limb 4 mm. in length, narrow below and puberulous, widening above and glabrous. Achenes glab- -rous, 5-7 mm. long, the radial ones shorter and ctirved. Pappus of ray flowers reduced to a narrow truncate coronula. Pappus of disk flowers reduced to 5 or 6 corneous, acute, teeth-like paleae which are deciduous from a disk. Type in the B. P. B. M. Herbarium, No. 391, AI, collected on Eke, West Maui. T. H., October, 1917, by C. N. Forbes. Argyroxiphium caligini var Kai v?.r. nov. Leaves with a slight greenish hue to the pubescence. Rachis of the raceme glandular pubescent and not silvery. While of wider distribution than the species, it is much less common. The variety is named for Mr. K. Kai, head ditch man at Honokohau Valley, Matii, who aided in the exploration of Eke. Type in the B. P. B. M. Herbarium, No. 391, a. M., collected on Eke, West Maui, T. H., October, 191 7, by C. N. Forbes. Another specimen in the same Herbarium is No. 64, M, collected on Puu kukui, West Maui, T. H., Mav, 1910, bv C. N. Forbes. [9] Bernice P. Bishop Museum Occasional Papers Volume VII. Plate III HIBISCADELPHUS BOMBYCIXl'S FORBES. Beniice P. Bishop Museum Occasional Papers Volume VII. Plate IV VIOLA KAUAIENSIS VVAHIAWAENSIS FOKBES. Bernire P Bisliii]) Museum Occasional Papers Volume VII, Plate V CYRTANDRA OLONA FORBES. Bernice P. Bishop Museum Occasional Papers Volume VII, Plate VI CYRTANDRA HII FORBES. Bernice P. Bisliop Jluseuni Occasioiuil Papers Volume VII, Plate VII / CYRTAXDRA PROPIXQUA FORBES. Bernice P. Bishop Museum Occasional Tapers Volume VII, Plate VIII CYRTANDRA MUNROI FORBES. Beniic" 1'. Bishoj) Museum Occasional Papers Volume VII, Plate IX CYRTAXDRA GEORGIAXA FORBES. Bernice. 1'. Bishop Museum Ofcasiijiial rajieis Volume A'll. I'late X CLERMOXTIA SAMT'ELII FOHBES. Bernice 1'. BislKip .MusfUiii Occasidiial Tapers A'oluiiie VII, Plate XI ARGYROXIPHIUM CALIGIXI FORBES. OCCASIONAL PAPERS BERNICE PAUAHI BISHOP MUSEUM OF POLYNESIAN ETHNOLOGY AND NATURAL HISTORY \u),L;JllV \ ii -NU. 4 A NEW CYANRA FROM LANAl, HAWAII By CHARr and a more comjiact rootstalk than those collected b\' Cdiamisso. A station for this ])lant was fotmd in I'alolo valley, within the cit\' limits of Honolnln. in a small, poorh' drained area formerly taro ]iatches. but now overgrown with several species of ""rass and sedg-e, among which are a few large Kiawe trees (Prosopis jitHHora). Among- the sedges was an abundance of Blacocharis palustris (L) R. Brown, which was recorded from Oahu by Knnth, though its existence in the Hawaiian group was doubted by Hillebrand." When the station was first visited in March, an area of about two acres was flooded with water, on the surface of which were floating thousands of Alarsilea leaves. The plants gathered at that time were all sterile and were glabrous in all parts except the nodes, which varied from nearly glabrous to somewhat woolly (PI. XIII). The length of the petioles varied directly with the depth of the water in which the plants were growing — long petioles in deep water, short |)etioles in shallow water — and were of unusual length on plants growing in water amongst grass. Plants taken from shal- low water near the edge of the pond had petioles ranging in length from 3.5 to 4.0 centimeters; whereas many growdng in water with pjass had petioles of 23 centimeters in length. In jiroportion as the water dried up with the advance of summer, the plants became more and more conspicuously rusty woolly at the nodes, aufl the under side of the leaves became pubescent with whitish hairs. Finally when the water had disappeared, during the last week in April of the same year, the ]')lants were found t