- 273

In October 1990, the authors began a project to identify the biological components of kāhili preserved in the Bishop Museum, to determine techniques of construction, and to reassess their ethnohistorical significance. 1 This paper presents the results of that study. We begin by reviewing the ethnohistorical context of these distinctive ensignia of rank in 18th- and 19th-century Hawai'i, then examine the process of kāhili construction, focusing particularly on the raw materials used and the techniques used to assemble the standing kāhili surviving from the 19th century. Finally, we examine the particular bird species used in the manufacture of kāhili, and conclude with some speculations on the possible impact of the collection of feathers for kāhili on the bird species involved.

This paper is based on a detailed examination of the entire collection of 75 tall, or standing, kāhili (kāhili kū) preserved in the Bishop Museum (Rose MS 1991). Over half of the collection (46 kāhili or 61% of the total) originate from heirs of the Kamehameha dynasty: Queen Emma (29 kāhili), Bernice Pauahi Bishop (10), Ruth Ke'elikōlani (4), and Victoria Kamāmalu (3), heir to the throne during the reign of Kamehameha V. These kāhili survive from among those assembled for the funerals of Victoria Kamãmalu in 1866, Kekelaokalani (mother of Queen Emma) in 1880, and the funerals of Ke'elikōlani, Bernice Pauahi Bishop, and Queen Emma between 1883 and 1885. They were among the first items placed in the Bishop Museum in 1891 and comprise the initial 46 entries in the catalogue — an incidental footnote of their contemporary esteem. 2 Of the remaining 29 standing kāhili, 17 were received between 1910 and 1924 from the heirs and estates of Kalākaua (5), Lili'uokalani (7), Kapi'olani (1), and Ka'iulani and Likelike (4). All but two or three of the remaining 12 were probably associated with Lili'uokalani, but this is not fully verified in the museum records. The last two kāhili, received in 1946 from the Edgar Henriques estate, descend through Lucy K. Peabody, former lady-in-waiting to Queen Emma. A collection of some 175 small kāhili and raw materials used in featherworking — as well as historic photographs, prepared bird skins and mounted specimens in the ornithology collections at the Bishop Museum and Brigham Young University, Hawai'i Campus — were also used for reference and comparison.

- 274

William Ellis, surgeon on Cook's third voyage, grasped something of the variability and complexity of Hawaiian kāhili when he observed in 1778-9:

Kāhili were observed and occasionally collected by members of Cook expedition and by subsequent visitors, but their cultural contexts were rarely described in useful detail. From the early contact period to the 1820s, individuals of high chiefly rank, males and females of all ages, were observed on many occasions accompanied by servants and retainers, one or two of whom usually carried a kāhili. Favoured attendants held kāhili over the sacred heads of their masters or waved them gently to and fro as fly brushes to stir the air and keep troublesome insects at bay. Although kāhili provided practical benefits of personal cooling, especially inside the close and stifling grass houses, they served primarily to mark the presence and spiritually protect an important individual of ali'i or chiefly rank. Kāhili bearers were of either sex, but early accounts suggest that females tended to be bearers for female ali'i and males for male ali'i.

Kāhili bearers, along with other members of a chief's retinue such as the ubiquitous spittoon bearer, occupied honoured positions; called pa'a kāhili or lawe kāhili, they functioned essentially as personal servants. Jules Remy noted in the mid- 19th century: “The duties of this officer were continual and most fatiguing, for he must constantly remain near the person of his master, armed with his kāhili, whether the king was seated or reclining, eating or sleeping.” (Nordhoff 1974:229-30). Historian David Malo wrote: “Where the king went, there went his kāhili bearer…and where he stopped, there stopped also the kāhili bearer. When the king slept the kāhili was waved over him as a fly-brush…” (Malo 1951:77). Malo added that servants who handled kãhili about the king's sleeping place were called ha'a e or kua-lana-puhi (pp.59, 61); yet another name for them was olu-eke-loa-ho'o-ka'a-moena — a poetic allusion to the rustling of the chief's tapa robe or blanket as he turned from side to side while lounging on his mat.

- 275

Apart from the times when chiefs ceremonially greeted visiting ships, outsiders had few opportunities before 1820 to witness — or record — purely indigenous formal events in which kāhili may have played a conspicuous role. When New England missionaries arrived in 1820, their headquarters at Kawaiaha'o Church soon became the venue of pageants and colourful spectacles mixing the old and the new. During the 1820s, school examinations, the occasional dedication of a new church structure, or annual celebrations commemorating the death of Kamehameha I and accession of Kamehameha II provided suitable opportunity for ceremonial display. Kāhili of all sizes and hues, feather capes and helmets, elegant tapa and imported silks and satins, mele and hula were conspicuous components of the display on these occasions.

The missionary C. S. Stewart witnessed one such procession in May 1823 that concluded a two-week commemoration of Kamehameha II's accession. In full view of thousands of spectators, his ranking wife Kamāmalu, “seated as a Cleopatra” (Bingham 1847:184-5), was carried on the backs of men in a large whaleboat shaded by an immense Chinese scarlet silk umbrella. Two kāhili bearers, high chiefs Nahi and Kalanimoku, proclaimed her rank by each carrying

Except for height of the pole, this description closely matches the appearance of kāhili in the Bishop Museum collection.

As a highly conspicuous form of status-rivalry, pageantry such as this waned through the coming decades of change confronting Hawaiian society. As a consequence, kāhili assumed new functions in increasingly Westernised contexts. Whether borne as fly-flaps by personal retainers as in the past or set up as symbols of authority of the evolving state, kāhili added colour to formal audiences and receptions whenever the royal court and chiefs received visiting foreign dignitaries. When Kamehameha III convened the newly created Hawaiian Legislature for the first time in Honolulu on May 20, 1845, he was accompanied by two kāhili, “their feathers…worn and rumpled by age…”:

They were, however, about twenty feet high, with massive and rich staffs — the one surmounted by black feathers on a white ground — the other by orange and crimson. These were placed so as to tower over the throne, over which was - 276 thrown the only really rich remnant of royalty left. This was a feather cloak, made of very minute yellow feathers… (Judd 1928:119-20).

Thereafter, State kāhili appeared at annual openings of the Legislature until it was reorganised in 1850, and subsequently throughout the 19th century on special occasions. Until the overthrow of the monarchy in 1893, kāhili highlighted numerous state events, such as the accession of Lunalilo in 1873, the coronation of Kalākaua and Kapi'olani in 1883, and the 50th birthday jubilee of Kalākaua in 1886.

Throughout the 19th century and into the 20th, kāhili graced formal social functions such as receptions, banquets, dinner parties, lū'au, poi suppers, church services, and quasi-official events wherever ranking members of the Hawaiian ali'i were present or memorialised. One event relevant to note is the grand birthday lū'au, banquet, and ball that Ke'elikōlani hosted from April 9 to 11, 1882, to celebrate the housewarming of Keōua Hale; some of the kāhili made for her stately new mansion are preserved in the Bishop Museum (see figure 1). By the turn of the century and the demise of the old ali'i, kāhili became relegated more and more to stage and costume props for historical pageants and tableaux, such as the 1920 centennial of the arrival of the first missionaries, the 1928 sesquicentennial of Cook's discovery, various Hawaiian civic society functions and, of course, the annual Kamehameha Day parades still held each June.

Kāhili are perhaps most commonly associated with the imposing state funerals staged throughout the second half of the 19th century for members of the Kamehameha and Kalākaua families. Although a newspaper account in 1921 4 purports to describe the funeral procession by canoe of Kalani'ōpu'u in 1782 — in which the kāhili named 'Ele-ele-ua-lani, Kaua-ka'a-honua, and Hawai'i-loa were erected — there are no eyewitness accounts yet found of pre-Christian Hawaiian mourning practices involving kāhili. On the death in 1823 of Keōpūōlani, mother of Kamehameha II and Kamehameha III, she was accorded the first native Christian funeral in Hawai'i. Her remains were borne in procession to a newly constructed mausoleum at Moku'ula, Lahaina, on a bier surrounded by six pallbearers — the five wives of Kamehameha II and the wife of Boki — “each carrying a beautiful black kāhili…” (Stewart 1830:348; see also Bingham 1847:197).

As Christian burial practices supplanted traditional rites, ali'i funeral processions became increasingly elaborate, and public. In May 1825, Lord Byron of HMS Blonde brought home to Honolulu the bodies of Kamehameha II and Kamāmalu, who had died on a state visit to London the year before; their remains, in magnificent coffins furnished by the British Government, were

- 277

Figure 1. Some of the Kamehameha family kāhili assembled in front of Keōua Hale, the house of Ke'elikōlani and Bernice P. Bishop, c.1890. All these kāhili are preserved in the Bishop Museum. Photo: W. T. Brigham, Bishop Museum Archives.

- 278

interred with military honours. The procession from the Blonde to Kawaiaha'o Church, then to temporary shelter in a grass house belonging to Kalanimoku, was preceded by male bearers carrying nine kāhili, “about thirty feet long, and from one to two feet in diameter, some of black, some of crimson, others of green, and others again of yellow feathers…” (Stewart 1830:348; see also Loomis MS).

Western military processions of honour, provided initially by visiting naval ships and which proved to be so pleasing to Hawaiians of the 1830s and 1840s, became a standard feature of state funeral ritual. Before 1850, contemporary eyewitnesses, and early newspaper accounts when they exist, do not always mention the number or appearance of kāhili, or even their presence, but there is clear evidence for increasing numbers of kāhili being displayed throughout the second half of the 19th century (see Table 1). Kamehameha III, laid to rest in the old Royal Tomb at Pohukaina in 1855 in the most imposing funeral cortege yet witnessed, was accompanied by at least four “large” kāhili (reportedly four black, or two yellow and two green — accounts are ambiguous) and 16 “small” kāhili. 5 Most of the individuals interred (or transferred there later) at the Royal Mausoleum at Mauna'ala were similarly honoured; their biers were accompanied in procession to their final resting place surrounded by kāhili. High-ranking female chiefs seem to have merited a greater number of kāhili than their male counterparts during the second half of the century. Notable are the funerals of Ke'elikōlani, Bernice Bishop, and Queen Emma (whose rituals marked the culmination of 19th-century royal funerals; see Fig. 2.)

Funeral obsequies for ranking ali'i throughout most of the 19th century lasted ideally a full month, during which period the deceased lay in state constantly attended by mourners. Watching over the body in pairs of three or four, they waved small, often black kāhili slowly and rhythmically over the coffin in solemn vigil day and night. From time to time chanters wailed in traditional manner and recited mele in honour of the deceased. Tall kāhili were assembled from components previously stored away and added to the scene, their numbers culminating

- 279

on the day of the funeral and procession to the mausoleum. At the conclusion of ceremonies at Mauna'ula, it was customary to place half a dozen or so kāhili around the coffin and apparently outside the mausoleum as well. They were left behind to weather slowly as the feathers drifted away in the breeze one by one, or until another royal funeral required that they be removed. The other large kāhili were carefully dismantled and their feathers stored until needed again, usually when the next ali'i death occurred.

The word kāhili, defined in part by the Pukui-Elbert Hawaiian Dictionary (Pukui and Elbert 1971:105) as “Feather standard, symbolic of royalty…to brush, sweep, switch”, has cognates in many Polynesian languages, including Tahitian, tahiri (Davies 1851:242); Marquesan, tahi'i (Phelps 1976:96); Aitutaki, tahirihiri (Buck 1907:207); and Maori, tawhiri (Buck 1927:207). Although the physical referents differ, ranging from finely plaited fans to fly-flaps of various sorts, most connote some sort of insignia of rank and/or authority. 6 In Hawai'i today, two kinds of kāhili are recognised. Kāhili kū are the “standing or tall kāhili, and kāhili pa'a lima or kāhili lele are small, much more common “hand-held” or “waving kāhili.

Like other highly esteemed articles such as feather cloaks and helmets, kāhili also were given personal names. Such names are recorded for 29 (39%) of the standing kāhili in the Bishop Museum, mainly on old paper labels glued to Queen Emma's kāhili poles (Brigham 1892:10-5; see also Brigham 1899:21-4). Kāhili 9, for example, is Ka-'ula-ho'āno-lani The “Red Awe of Heaven”, named by its makers “for the redness ('ula) of the feathers and the heavenly sacredness (ho'o-lani) attached to the chief for whom the kāhili was made” (HEN 1:2751). The

- 280

Figure 2. Kāhili in the funeral procession of Queen Emma from Kawaiahao Church to the Royal Mausoleum, May 17, 1885. Photo: Bishop Museum Archives.

- 281

engraved inscription on its metal-sheathed pole proclaims that it was “Presented to the Prince of Hawai'i by his native subjects of Foreign Parentage” in 1859 to celebrate the birth of the only child of Emma and Kamehameha IV on May 20, 1858. It was refurbished, probably with albatross feathers, before being taken to the Bishop Museum in 1891. One of Bernice Bishop's kāhili (1), made of black 'ō'ō feathers, is named 'Ele-'ele-ua-lani “Black Rain of Heaven”; it is a very old name descended from the chiefs — “rich and beautiful in color as the greenery of the forest” (Kamakau 1961:184). Besides physical characteristics, other kāhili are named for famous ancestors, such as several in Queen Emma's collection: La'ie-lohelohe (6), wife of Pi'ilani; Ka-neoneo (24), a chief of Kaua'i killed in the battle of Nu'uanu in 1795; and Ke-ku'i-apo-iwa (27), a wife of Kekaulike.

According to some writers (Brigham 1899:15; Chinen MS 1980; Scott 1930:16), the name of a kāhili was applied to its pole or shaft. When its feathers were removed and stored after the specific occasion for use had passed, or when the feathers became worn and were replaced by others (not always of the same kind), the name remained for the pole alone. Perhaps once the rule, particularly for poles incorporating human bones from known individuals, there are exceptions in the Bishop Museum kāhili. The feathers of 'Ele-'ele-ua-lani, for example, were removed at some unknown time and remounted on a turned and polished, imported ash pole, yet the name remains. Ka-'olohaka, a ruler's kāhili from the time of Kamehameha I to Kamehameha V, was one of those placed by Ke'elikōlani at the head of the coffin of Kamehameha III at the time she inherited the estate of the Kamehamehas. According to Lucy K. Peabody (HEN 1:2749), the kāhili was taken from the Royal Mausoleum by Kalākaua then acquired later by Lili'uokalani, who buried the bones of Ka-'olohaka at Mauna'ala and commanded her principal featherworker, Naheana, to divide the feathers for use in two kāhili. This pair (1922.08.02-.03) — with yellow 'ō'ō axillary (shoulder) and red 'i'iwi contour (body) feathers on similar walrus ivory and turtle-shell poles — were often displayed on either side of Lili'uokalani at royal receptions and other special events. They stood guard at her bedside until her death, and finally at the head and foot of the bier upon which she lay in state at Kawaiaha'o Church.

Like the second incarnation of Ka-'olohaka, kāhili were frequently made in pairs — sometimes surprisingly matched duplicates, other times differing only in size, feathers, or type of pole. Of the seven closely matched pairs in Queen Emma's collection, both members of four pairs have the same name (8 and 33, Ka-lele-ho 'ānoāno; 10 and 29, Ka-maka-alaneo; 12 and 31, Ka-leo-aloha; 16 and 25, Lele-'ō'ili). Individual kāhili in three other physically similar pairs bear different names (2, Noelani and 4, Ka-maka-ma'o) and (14, Ku'imaka, and 27, - 282 Ke-ku'i-apo-iwa; for 20 and 22, see below). Four nearly identical kāhili (40-43) are named Pili-aloha; made of large, dyed red feathers on kauila poles, they were displayed at the four corners of the bier of Queen Emma's mother Kekelaokalani in 1880, and five years later over her own remains in Kawaiaha'o Church (Brigham 1899:24).

The pair with two different names, a famed and still handsome pair, combines trimmed black 'ō'ō contours and yellow axillary feathers with red 'i'iwi contours and central tail feathers of red-tailed and white-tailed tropicbirds. La'i-kū “Extreme Calm” (20, see Fig. 3) is mounted on a turned and polished, reddishstained kauila pole, while Malu-lani “Heavenly Shade” (22) is on a walrus ivory and turtle-shell pole. According to Lucy K. Peabody (HEN 1:2753-4), these kāhili were made and given to Queen Emma during her childhood at Lanikeha in Lahaina, Maui, but their names, bestowed by the chiefs, were not ancestral. Both kāhili were erected at the heads of the coffins of Kamehameha IV and the Prince of Hawai'i at Mauna'ala, in 1864, then placed at the head of Queen Emma's bier during her funeral obsequies in 1885. It is interesting to note that Malulani Beckley Kahea, a celebrated featherworker of this century, was named for one of these kāhili, as was a daughter of Nalaniwahine. 7 Lili'uokalani herself is said to have quoted often the expressions “La'i-kū” and “Malu-lani” (Webb n.d.).

Because of the fragile and irreplaceable nature of the kāhili in the Bishop Museum collection, it was not possible or desirable to disassemble any kāhili or their component parts in order to discover the construction techniques and materials used in their manufacture. As such, construction techniques had to be inferred through close visual inspection of the assembled components, and especially of areas where deterioration had revealed important details of the assembly process. Much information on construction and materials was also derived from examination of 40 of the standing kāhili disassembled (the branches taken off the pole) for safer storage between 1967 and 1974. The ability to examine the individual components of these kāhili allowed us to infer the construction techniques with much greater accuracy than otherwise would have been possible. The construction processes detailed in this paper represent the most plausible techniques that we can infer from both visual inspection and the few ethnographic sources that exist on kāhili construction.

Hawaiian kāhili are constructed from three basic components: the 'au or branches which make up the hulumanu or feathered cylinder; the kumu or pole

- 283

Figure 3. The kāhili called La'i-kū (20, “Extreme Calm”), one of a pair of kāhili, the other being Malulani (“Heavenly Shade”), made and given to Queen Emma during her childhood at Lanikeha in Lahaina, Maui. The feathers in this kāhili are the central tail feathers of koa'e-ula (red-tailed tropicbird), 'ē'ē (axillary feathers) of the Hawai'i 'ō'ō and body feathers of 'i'iwi. Photo: C. Meyer.

- 284

to which the branches are tied; and the optional trim consisting of the pāpale or cap, the pā'ū or skirt and the streamers, originally of kapa or bark-cloth. Although construction techniques of each of these components differ only slightly among individual kāhili, the kāhili makers of old 8 were able to produce a tremendous variety of shapes, colours and styles by varying the form, materials and combinations of the three types of components. One factor which made this diversity possible was the availability of a large assortment of different raw materials, both native and imported, whose great diversity is only now beginning to become apparent. The following discussion of raw materials and construction is based on the 75 standing kāhili kū in the Bishop Museum. Small hand kāhili (kāhili pa'a lima, or kāhili lele) are more variable in form, materials and appearance and are not discussed in this paper.

The most visually striking aspect of Hawaiian kāhili is the large hulumanu or feathered cylinder forming the crown. In most specimens, this cylinder is composed of several hundred separate 'au or branches each tipped with several bunches of feathers. Traditional kāhili branches are all made of one or more bundles of stiff plant fibre, henceforth “core fibres”, which are lashed together to form the finished branch. Analysis of these core fibres reveals that most of the branches are made from the split adventitious roots of the indigenous 'ie'ie vine (Freycinetia arborea), with coconut midrib (Cocos nucifera) and some other as yet unidentified plant fibres also present to a much lesser extent. Table 2 lists the botanical elements identified in kāhili branch cores and lashings. The predominant use of 'ie'ie as a branch core fibre accords with the few surviving historical accounts that we have of kāhili construction (Anonymous MS n.d.). From the late 19th century on, there is also some use of stiff copper or steel wire as a branch core material, as seen in four of the later Bishop Museum standing kāhili.

- 285

Construction of the branches was accomplished using two slightly different techniques, each of which produced a similar appearance in the finished branch. In the most common technique, the 'au (branch) was made using a single bundle of core fibres evenly divided to form branch-like “forks” (i.e., it was bifurcated at set intervals) up to three times before ending in the twig-like tips, the pe'a or “tines”, into which the feathers were incorporated. At each bifurcation, the bundle of fibres was divided approximately in half, usually by making a half-twist of the right half of the bundle over the left, and the shape of the fork secured either by bending back one or more of the core fibres to hold the two sides of the fork apart, or by tying it into shape during the lashing process—or both. Feathers were attached usually by first tying them into a small bundle (sometimes with a splint of core fibre to make them easier to attach) and then incorporating the bundle into the pe'a at the end of the 'au. Once the feathers had been inserted, the entire 'au was lashed together by wrapping the lashing material helically along the full length of the branch. Traditionally, the preferred lashing material was olonā (Touchardia latifolia) bast, but imported cotton thread was used in the construction of most surviving kāhili. In some specimens, the branches, once lashed, were given an overlashing of coloured yarn or ribbon to give the branch a specific colour. In a few specimens, the entire branch was dyed (usually red), again, perhaps for decorative effect.

In the second construction technique (which is much rarer than the first), the branches were first made in several individual subunits which were then assembled to form the finished branch. For example, to produce a double-forked branch, two single forked subunits would be made using the technique described above. These two subunits were then lashed together, along with some extra fibres that formed the stem (by which the branch was attached to the pole), to form a double-forked branch. Rather than being a separate technique for making new branches, this second technique probably was developed in order to reuse miscellaneous shorter branches from older kāhili in the construction of a new specimen — a practice well documented in historical records.

By varying the number of times a particular branch forked, as well as the number and colour of feathers used, kāhili makers were able to produce considerable variety. Most branches in large kāhili were of four basic types: unforked (in which the initial bundle of fibres was not divided, resulting in a straight branch with a single tine); single-forked (in which the bundle was divided once resulting in a branch with two tines); double-forked (in which the bundle was divided twice, resulting in a branch with four tines); and the rare triple-forked (in which the bundle was divided three times resulting in a branch with eight tines). Figure 4 shows examples of unforked, single- and double-forked branches. Most branches seen in large kāhili are of either the single- or double-forked variety, with triple-forked branches occurring in only two of the Bishop - 286 Museum kāhili; these two specimens (26 and 32) are associated with Victoria Kamāmalu, heir to the throne who died in 1866, and are, thus, among the oldest in the Bishop Museum collection.

On completion, the branches could be stored for future need or used immediately. For assembly, the stems were bent to the appropriate length and then lashed to the pole beginning at the top and working downwards. Traditionally, either twisted olonā cordage or bast was used; however, ordinary cotton twine and commercial jute or hemp cordages are more common on surviving kāhili. In most kāhili, the 'au are lashed on in a single continuous descending spiral, a technique that required both great skill and patience to perfect. In others, the 'au are lashed on in a series of tiers, each consisting of a single ring of branches encircling the pole. Although overall shape of the hulumanu varies considerably, the most common form is a uniform cylinder of feathers which varies in compactness or density from one kāhili to the next. Rarer forms include squat or truncated cylinders, spheres, and spiky cylinders. Kāhili made of the two central tail feathers of the koa'e-ula or red-tailed tropicbird (Phaethon rubricauda) or koa'e-kea or white-tailed tropicbird (Phaethon lepturus) somewhat resemble pinecones. In some specimens, feathers of different colours are used to create a cap-like ring of branches on top of the hulumanu or a band of trim at its base.

Particularly notable are those kāhili with globular hulumanu which often resemble rosebuds or suggest the shape of ice cream cones. Such globular kāhili have an internal frame of bent wire or metal umbrella ribs to which the branches are tied to keep the hulumanu in its proper shape. A unique pair of kāhili, resembling English topiary and made not from feathers but from the curled shavings of an as-yet-unidentified plant, was presented to Queen Emma by a hui of women from Wailuku on her birthday in 1883. These two kāhili (only one of which survives) are the only examples in the Bishop Museum collection which do not use feathers. Other kāhili, made of kapa, flowers, or ferns are known from accounts of ali'i funerals but, as far as is known, none of these has survived.

Traditionally, the most important element in a kāhili was said to be the pole or kumu from which the kāhili derived its personal name (Brigham 1899:15; Chinen MS 1980:39). The length of the poles used in standing kāhili determined their final height and varies from about 2 to more than 5 metres. Although many different types of poles were produced from the late 19th century onwards, the earlier kāhili poles were of two kinds. The simplest of these were made by reusing an old spear to which the branches were tied. If a separate pāpale or cap (see below) was used, the tip of the spear was notched to allow it to be tied on more securely, but otherwise, little or no modification was needed. Several sources cite kauila (Alphitonia ponderosa) or 'ōhi'alehua(Metrosideros polymorpha) as the

- 287

Figure 4. Kāhili branch types (left to right) unforked (red-tailed tropicbird central tail feathers), forked (peacock eye feathers) and double-forked (Hawai'i 'ō'ō axillary feathers). Photo: S. Conant.

- 288

type of wood most frequently used for such spears/kāhili poles (see Brigham 1899:15-16; 1892:11; Handy and Handy 1972:287). Identification of thin sections of wood samples confirms that all of the “spear” type poles in the sample collection are made of Alphitonia ponderosa, one of two botanical species jointly known as kauila to the ancient Hawaiians.

The second variety of kāhili pole is more complex. In these specimens a central wooden shaft was used — again usually made of kāhili or, in at least one case, of 'ulu (breadfruit, Artocarpus altilis) and another of wauke (paper mulberry, Broussonetia papyrifera) — around the lower portion of which an elaborate handle was assembled. These handles, among the most finely crafted of Hawaiian artefacts, were constructed by alternating stacks of perforated turtle-shell disks and cylinders of human bone or ivory. Analysis of the turtle-shell has shown most of it to be from the 'ea or hawksbill turtle (Eretmochelys imbricata). Two different colours of turtle-shell are used; a dark brown from the dorsal scutes on the back, and a lighter brown or amber from the plastral scutes on the underside of the shell. Ivory was obtained primarily from walrus tusks (Odobenus rosmarus) imported mostly by whalers after 1820, and probably also from sperm whale teeth (Physeter macrocephalus) either brought in by passing ships or obtained locally. Human bone (Homo sapiens) was taken from the long bones of the arms and legs, one of the traditional residing places of a person's mana or spiritual power. Table 3 provides a summary of the biological materials identified in the kumu examined and some of the different pole types are identified in Figure 5.

In later historic times a wide variety of different types of poles was made and used. The most common were long slender poles of imported conifer wood painted in shades of brown, black, maroon and white to resemble the more traditional bone/ivory and turtle-shell poles. Analysis of several of these poles shows them to be made from Pinus lambertiana. Other later kāhili poles were made of koa (A cacia koa) and kauila, turned on a lathe and then shellacked or French polished. The most elaborate kumu were made by laminating a number of short segments of various Hawaiian woods to a central core and lathe-turning the assembled pole to a uniform diameter. One reason for creating this type of pole was apparently to display the beauty and diversity of native Hawaiian woods. Laminated poles in the Bishop Museum collection include woods from endemic tree species such as kauila and koa, Polynesian introductions such as kou (Cordia subcordata), and kamani (Calophyllum inophyllum), and at least one postcontact introduction, the monkeypod (Samanea saman).

Trim on Hawaiian kāhili consists of three separate components: the pā'ū or skirt, the streamers (which may have been known as kapa), and, sometimes, a - 289 pāpale or cap on the top. Because trim was often renewed each time a kāhili

was reassembled, it is difficult to determine what type of trim, if any, might have been present on the earliest kāhili. Skirts served to hide the unsightly lower ends of he branches of the hulumanu and may have originally been made of kapa cloth, which is found on one of the specimens in the Queen Emma collection. Most pā'ū consist of a cardboard cone to which either fabric or dyed feathers are sewn with cotton thread. Fabric skirts were made in a variety of different styles and colours with yellow, white, red, and blue being the most common. Analysis of skirt fabrics reveals that those skirts appearing to be “original” are made of silk or cotton.

The feathered pā'ū were made by sewing dyed feathers (usually of some domestic bird such as chicken, duck, or goose) to a cardboard cone that had,

- 290

Figure 5. A sample of kāhili kumu (pole) types. From left to right the pole numbers and the materials of which they are made: 12 (painted conifer), 143 (ivory and light-and dark-coloured turtle-shell; the light coloured rings are made from the ventral shell), 36 (koa wood), 18 (laminated native Hawaiian woods), 4 (reused kauila spear). Photo: E. P. Kjellgren.

- 291

in most cases, first been covered with ordinary balanced plain-weave linen or cotton cloth. Most feathered skirts feature two colours in patterns that include alternating vertical or horizontal stripes, crescents, or checkerboard squares. Predominant colour combinations are yellow and red, but yellow and white, yellow and green, purple and white, and other colours also occur. Feathered pā'ū are usually accompanied by smaller fabric skirts which are attached to the base of the cardboard and hide the juncture with the kāhili pole.

One, or more commonly two, streamers made of various types of ribbon (the older ones usually silk), are tied around the base of the skirt, the knot being either a bow or half-bow of varying length. The free ends of the ribbon form dangling streamers and are often cut at a 45° angle or in a fish-tail pattern for decorative effect. The colour of the streamers is sometimes different from (but complementary to) that of the pā'ū and streamers of two different colours are often used. Streamers were frequently renewed, so it is difficult to find out what types or colours originally decorated many of the surviving kāhili.

Like the pā'ū, the pāpale or “cap” was constructed from a piece of cardboard to which fabric and/or dyed feathers were sewn. The pāpale was stitched to a frame, consisting invariably of a cross made from two short lengths of old umbrella ribs tied together with twine, then secured to the top of the pole, which was notched to accommodate it. Colours and patterns of pāpale almost always match those of the pā'ū. Most pāpale are decorated only on the top side, all but invisible from the ground, whereas, in a few others, the decoration appears on the underside only, presumably because pāpale were seen from below in normal use.

Unfortunately, there are few historical sources to shed light on the precise ways in which the components were assembled into a completed kāhili. Nonetheless, by examining the assembled extant specimens, the process can be inferred. The various components — 'au, kumu, pā'ū, streamers, and pāpale — were probably assembled as follows: First, the top of the pole, in some cases, was wrapped either with kapa or, later, cloth. Next, the 'au, starting at the top and spiralling down, were tied on with either olonā cordage or commercial twine. Proper spacing and attachment of the 'au must have been by far the most difficult aspect of the assembly, since this had to be done precisely so as not to leave unwanted gaps in the finished hulumanu. Once the 'au were tied in place, trim was attached, probably in no particular order. The pā'ū could be slipped up from the bottom of the pole to cover the ends of the lowest 'au. The pā'ū was tied in place, using cotton or some other convenient twine, over which the streamers were tied to complete the decoration. In attaching the pāpale, the umbrella-rib frame was probably first tied to the top of the pole and then the ornamented disk stitched to it with thread.

- 292

Because none of the components was permanently attached to any other, each could easily be combined and recombined into an almost infinite variety of possible configurations. This flexibility is responsible for the great abundance of forms and innovations seen in Hawaiian kāhili, particularly those surviving from the later historic period.

In the course of examining each kāhili and describing its structure, any feathers already detached, or that became detached in the process, were labelled and set aside for comparison with bird specimens. Comparison of feather samples with bird specimens resulted in many positive identifications, but some samples remain unidentified, particularly those from presumed domesticated birds (e.g., chickens, ducks, geese). Although some museum catalogue entries for kāhili specified feather sources, many of these entries proved to be incorrect.

At the time many of the standing kāhili we examined were originally assembled, kāhili makers would have had access to feathers from commercial sources, because use of feathers in the millinery trade was extensive during the late 1800s and early 1900s (Berger 1981:8; Wetmore 1925:86). As a result, “exotic” bird feathers could have come from virtually anywhere in the world. This situation made the task of identifying feathers from non-native bird species difficult, though we made many tentative identifications.

Of the species of native birds available to them, the Hawaiians selected relatively few to provide feathers for the kāhili we examined. Bird species whose feathers were used in kāhili can be separated into three major biogeographic groups: exotic, indigenous and endemic. The feathers of exotic species (those birds not naturally occurring in Hawai'i) were most difficult to identify, and some remain without positive identification. Indi-genous birds are those that naturally occur in Hawai'i and elsewhere; all were seabirds. The endemic species (species found only in Hawai'i) whose feathers were positively identified in kāhili were forest songbirds.

It was not possible to determine when the feathers used in a particular kāhili were actually collected. As mentioned above, standing kāhili were often assembled for a particular occasion of State and then disassembled and stored for future use. The materials may have been stored as assembled branches or simply as loose feathers, core and lashing fibres. If adequately protected from insects, exposure to extremes of light and humidity, etc., feathers may retain their colour and shape for decades, even centuries. Thus, it is possible, though perhaps unlikely, that some of the native bird feathers that occur in kāhili assembled as late as 1928 for the Cook sesquicentennial were collected before Cook's arrival in 1778.

- 293

It seems likely that both feather colour and some characteristics attributed to the source bird species by the Hawaiians influenced the choice of feathers used in kāhili. The significance of colour remains obscure, although we know from historic usage, as well as pan-Polynesian parallels that red and yellow were associated with the Hawaiian ali'i. Some individual ali'i seem to have preferred certain colours. For example, black is the predominant colour in the kāhili of Lili'uokalani, and reputedly one of her favourites. Two pure white kāhili (15 and 21) were made for the funeral of Mrs Bishop, allegedly as symbols of her pure character. Black, red, white and yellow are the most frequently used colours of feathers. When feathers are dyed, they are most often dyed red or yellow. Further research is needed to clarify the importance of feather colour and its iconographic significance in Hawaiian culture.

Regarding the “character” of the bird whose feathers were used, an understanding of how the Hawaiians regarded the various bird species might explain, for example, whether a black feather from an 'iwa (great frigatebird, Fregata minor) had the same significance as a black 'o'o (Moho spp.) feather. Keauokalani calls the 'iwa “a royal bird.” In writing of the mamo (Drepanis pacifica), the “chief of small, mountain birds”, he notes that “this bird has a proud appearance, as though with pride and haughtiness”. The 'i'iwi (Vestiaria coccinea) “flies up and down as if it despises laziness and idleness” (Keauokalani MS 1859). Whether such notions of a bird species' “character” played a significant role in the use of its feathers in artefacts remains an enigma. Rarity or difficulty in procurement of feathers may also have determined their value to some extent (Berger 1981:105-6, Peale 1848:8).

Only two species of exotic birds could be positively identified from the feathers present in the kāhili sampled (see Table 4). These two were the peafowl (Pavo cristatus) and the domestic chicken (moa, Gallus gallus). The spectacular “eye” feathers that comprise the tail, tail coverts and lower back feathers of male peafowl (i.e., peacocks) are distinctive, and were used in 16 of the 75 kāhili examined. Sometimes barbs from below the “eye” of these feathers were stripped from the vane and used in bunches. Eight kāhili contained peafowl contour feathers (that is, feathers that give shape or “contour” to the body) of the following types: downy feathers and brown and white barred feathers, either of which could have come from any age or sex of bird; and black or bluish-black iridescent feathers from adults. Peacocks had been given as a gift to Kamehameha III and Queen Kalama at least as early as 1848 (Richards 1970:300), and the feathers that appear in the kāhili may have come from descendants of this or other domesticated stock, or from a millinery supplier.

- 294

Rooster hackles, or the neck feathers of male chickens, have a distinctive shape, and were identified in two kāhili. In seven other kāhili, feathers tentatively identified as chicken contour feathers were mixed in with feathers tentatively identified as some variety of domesticated waterfowl (probably domestic ducks or geese). Chicken contour feathers can be distinguished from those of waterfowl because they are structurally looser.

Numerous feathers were attributed to the category “domestic ducks”, which could include domestic ducks (Anas spp.), muscovy ducks (Cairina spp.), or small specimens of domestic geese (Anser spp.). Another category, “large waterfowl”, could include geese, or perhaps swans (Cygnus spp.), and a number of different seabird species (e.g., albatrosses, shearwaters, boobies). Seabird feathers are similar in structure to those of many species of ducks and geese. The characteristically curled shaft and strong hooks that hold together the barbs of these feathers afford the plumage some of its water resistance. We suggest that most of the feathers we attributed to the “domestic duck” category are, in fact, domestic duck or goose, which would have been readily available commercially or from the local poultry yard. The feathers in these two categories, nearly always white to begin with, were usually dyed red or yellow.

Two kāhili had what appear to be emu (Dromaius novaehollandiae) feathers. The feathers may have come through the commercial trade or possibly from a specimen brought to Hawai'i as a curiosity.

- 295

Five seabird species were positively identified as sources for kāhili feathers. Additional species may also have provided feathers, but the identifications remain unconfirmed. Black-footed albatross (moli, Diomedea nigripes), red-tailed tropicbird (koa'e-ula), great frigatebird ('iwa), sooty tern ('ewa'ewa, Sterna fuscata), and brown noddy (noio-kōhā, A nous stolidus) feathers were all identified from Bishop Museum kāhili (see Table 5).

Black-footed albatross contour feathers, with their distinctive shape and colour, were found in three kāhili. Two or three may also have had contour feathers from Laysan albatross (moli, Diomedea immutabilis), but these white feathers cannot be distinguished with certainty from similarly structured feathers of large domestic waterfowl. The catalogue entry for kāhili 9 notes that it was “restored with albatross feathers”, which, in this case, is probably correct.

Most of the tropicbird feathers found in kāhili are from the red-tailed tropicbird (koa'e-ula) rather than the white-tailed tropicbird (koa'e-kea). This judgment is based on the red colour of unfaded central tail feathers (tail feathers are also - 296 known as rectrices), and the size of the black and white flank and tertiary (the third row) wing feathers (which are larger in the red-tailed tropicbird). Some black and white feathers from immature tropicbirds were used; the black and white feathers from immature birds can be distinguished from adult black and white feathers on the basis of pattern, and size again allows identification of the species in many cases. The white wing and tail feathers of tropicbirds have black shafts, a character not obscured by dyes.

The two red, central tail feathers of red-tailed tropicbirds lend a distinctive appearance to the eight kāhili in which they were used. These slender feathers can reach more than 45 cm, and the unfaded feather is a striking combination of red, black and white. The white and black feathers of white-tailed tropicbirds are nearly as long, and faded feathers are difficult to assign to one of the two species. Red-tailed tropicbirds, which nest in dense cover on the ground or in natural cavities, might have been an easier source of feathers than their cliff-nesting, white-tailed relatives.

The iridescent back and neck feathers of breeding male great frigatebirds were used in more (16) than those of any other indigenous bird. Although the glistening black, green and blue hues are similar to those found in feathers of feral chickens (i.e., jungle fowl), frigatebird feathers taper to a sharp point, unlike chicken feathers. Interestingly, two later kāhili made for educational purposes (INV-007 and INV-008) use chicken feathers trimmed to look like frigatebird feathers. Black feathers either from 'ō'ō or frigatebirds were used in numerous kāhili, and even the most brittle and dirty frigatebird feathers still retain their iridescence.

We were able to positively identify the feathers of two other seabirds. Two kāhili (B.6850 and B.6851) contain the trimmed primary (first row) flight feathers of brown noddies (noio-kōhã, A nous stolidus). The catalogue entry for kāhili B.6850 and B.6851 reads “brood tern”, a term we have been unable to trace in the ornithological literature. Kāhili B.7250 and its mate B.7251 contained the trimmed primaries of sooty terns ('ewa'ewa, Sterna fuscata). The catalogue entry for kāhili B.7250 reads “kala”, which Titcomb and Gagné (1976:117-26) construed as pakalakala or gray-backed tern (Sternalunata). Interestingly, kāhili INV-O10 and INV-O11 contain complementary parts of the trimmed primary flight feather type also found in kāhili B.7251. The feathers in INV-O10 and INV-O11 retain the shaft and the leading edge, while the hulumanu of kāhili B.7251 appears to be made of the portions trimmed out of the same type of flight (i.e., wing and tail) feathers.

Feathers from two species of endemic forest birds were positively identified from a variety of kāhili (Table 6). Six kāhili contained the bright vermilion contour feathers of adult 'i'iwi, lashed in bunches with feathers from 'ō'ō, red- - 297 tailed tropicbirds and other species. The extinct Hawai'i 'ō'ō (Moho nobilis) provided black feathers from the wings (probably primaries), tail (rectrices), and

body (contour) and yellow feathers from the shoulder (axillary) and undertail regions. 'Ō'ō rectrices were called puapua, 'ō'ō, axillary feathers were called 'ē'ē, and uppertail or undertail coverts (those feathers covering the stiff rectrices) were called pue (Buck 1957:538). The rectrices of M. nobilis are distinguished from those of the nearly extinct M. braccatus, and the two extinct ō'ō species, M. apicalis and M. bishopi, by their size and the shapes of the feather tips. In general, M. braccatus has smaller flight and other contour feathers with a brownish cast; they are easily distinguishable from flight and contour feathers of the other members of the genus. Rectrices of M. apicalis were white-tipped, unlike those of other Moho. The wing feathers (primaries, secondaries and tertiaries) of nobilis, bishopi and apicalis would be difficult to distinguish from one another. However, because most of the rectrices and yellow axillary and undertail coverts were attributable to M. nobilis, we concluded that wing feathers were most likely to have been from this species. The axillary feathers of M. nobilis are distinctive in structure; the short shaft and long barbs render the feather soft and loose, even in comparison with the similar axillary feathers of Bishop's 'ō'ō (M. bishopi) from Moloka'i.

The 5th and 6th (outer) pairs of rectrices of M. nobilis were important elements in four Kāhili in our sample, as well as a number of the hand Kāhili we briefly examined. These feathers, which were the shortest pairs of rectrices, were - 298 distinctively marked, white, yellow and black feathers. Called pīlali, they are illustrated in a hand-painted plate (Plate LXXII) of M. nobilis appearing in Rothschild's lavishly illustrated work on Hawaiian birds (Rothschild 1893-1900). The hulumanu of several hand kāhili are composed entirely of pīlali (that is, rectrix pairs 5 and 6).

kāhili 26 contained some yellow feathers that may have been uppertail coverts (called ko'omamo) of the extinct mamo (Drepanis pacifica) from Hawai'i Island. However, the possibility exists that these feathers were undertail coverts (pue) of M. nobilis, M. bishopi, or M. apicalis (Buck 1957:538). Generally the yellow undertail and uppertail coverts of birds in the genus Moho are distinguished by colour, even with some fading, from the uppertail and undertail coverts of D. pacifica, as the shades of yellow are distinctive. However, the feathers in kāhili 26 were faded and extremely dirty; a positive identification will require additional effort and suitable specimens for comparison.

The Hawai'i 'ō'ō was the bird whose feathers were used in the greatest number of kāhili in our sample (20, or 27%). Feathers from all parts of the bird were used: wing, tail, body, and axillary (shoulder). Axillary feathers comprise the entire hulumanu of kāhili 19. Pīlali comprise the entire hulumanu of standing kāhili 5, as well as some hand kāhili and major parts of the hulumanu of kāhili 6, 36, and B.7250. The fact that so many Hawai'i 'ō'ō feathers were used prompted us to attempt to estimate how many birds would have been required to supply the feathers in some specific kāhili, as well as a larger sample of the 75 standing kāhili. We began by counting or estimating how many of each feather type could be obtained from a single bird. Then, by counting the numbers of feathers in a sample of branches and counting the numbers of branches in individual kāhili, we obtained an estimate of how many of each feather type were in each kāhili.

Conflicting accounts exist as to how feathers were collected from birds. Some sources report that the kia manu or bird-catcher plucked only a few of the desired feathers, releasing the bird to grow more feathers and to reproduce (Buck 1957:217-8). On the other hand, Keauokalani writes that “the koa'e is a delicious bird, tasting almost like the nene”, and calls the “uwanau” (dark-rumped petrel, Pterodroma phaeopygia sandwichensis), “most delicious of birds” (Emerson 1895:101-111). In his Hawaiian Antiquities, Malo reports that 'ō'ō and mamo “are used as food” (Malo 1951:38). Several authors argue that the difficulty of catching birds, the large number of suitable feathers offered by some birds (e.g., 'i'iwi, mamo) and the scarcity of protein make it likely that birds caught for feathers were killed and eaten (Emerson 1895:101-111; Perkins 1903:368-465; Henshaw 1902:54-106;1903:113-45, Munro 1944:87). Wilson (Wilson and Evans 1890-9:105) wrote, “The O-O is esteemed a great delicacy by the natives, - 299 and used formerly to be eaten by them…”. Cook (1784:2:206) reported that large numbers of small red birds (certainly 'i'iwi) were offered for sale in bundles of 20 or more on Kaua'i. Henshaw and Munro both speculate that the use of the shotgun in the 1800s contributed to extinction of the mamo and other birds (Henshaw 1902:52, Munro 1944:92). Today, destruction of habitat, the effects of introduced predators and avian diseases carried by alien birds are all recognised as important factors in the extinction of many Hawaiian birds.

Taking into account the controversy about methods of bird and feather collection, we made our estimates of numbers of birds required to make kāhili in two ways. Our minimum estimate is based on the assumption that all suitable feathers were removed from any captured bird, whereas our maximum estimate was based on the assumption that no more than six of each feather type were removed from a captured bird. We note that, if all feathers suitable for use in kāhili were removed from an individual Hawai'i 'ō'ō, it would surely have died, being unable to fly without wing feathers. Thus, the minimum estimates are based on the presumption that birds were killed for their feathers.

We chose five kāhili (1, 5, 13, 19, and 27), each containing distinct Hawai'i 'ō'ō feather types, to examine in detail for our estimates. Table 7 shows our estimates of the numbers of feathers in those kāhili and the minimum and maximum numbers of birds we estimate were required to construct the kāhili. Table 8 shows the estimated total number of Hawai'i 'ō'ō feathers of each type found in most (18), but not all, of our sample of standing kāhili that contained Hawai'i 'ō'ō feathers. Also shown in this table are minimum and maximum estimates of the numbers of birds required to supply the feathers, based on the assumptions described above.

Did the capture of endemic forest birds and the use of their feathers in Hawaiian artefacts — kāhili, cloaks, capes, helmets, basketry images, and lei — cause the extinction of any of these species? We believe it would be impossible to satisfactorily answer this question even if virtually all precontact Hawaiian artefacts are as closely examined as our sample of kāhili has been. Although some individual artefacts may have required what seem to be immense numbers of feathers (Tables 7 and 8), and although it seems likely that most birds that were caught were actually killed, it seems unlikely to us that feather-collecting by early Hawaiians caused species extinctions. We base this speculation on the facts that capturing small forest songbirds must have been extremely difficult, and that feathers, properly cared for, could have lasted for decades. Furthermore, the numbers of people permitted to wear or possess feather artefacts was relatively small. Some particular species (e.g., Hawai'i 'ō'ō) may have experienced population declines, but extinction due to feather-collecting seems improbable. The dramatic changes in habitat wrought by agricultural development and the effects of introduced predators (e.g., the Polynesian rat, Rattus exulans)

- 300

- 301

as well as predation by the Polynesians themselves, are more likely to have contributed to the extensive extinctions of birds now known to be associated with the colonisation and occupation of Hawai'i by the Polynesians (Olson and James 1991:7; James and Olson 1991:7). Although we chose to focus our attention on a single bird species in this discussion, one could estimate, with sufficient time, the numbers of birds required to supply certain other feather types (e.g., the central tail feathers of tropicbirds).

Our current research clearly shows that the construction of just the modest sample of kāhili we examined required the capture of many thousands of Hawai'i 'ō'ō. Even the use of firearms, about which we know very little, to collect feathers during the postcontact period would not have rendered this an easy task. It is no wonder that Malo wrote that “the feathers of birds were the most valued possessions of the ancient Hawaiians” (1951:76).

• Anonymous, MS n.d. A Description of A Kahili I Saw Made… Hawaiian Ethnological Notes. Archives, Bishop Museum, Honolulu.

• Berger, Andrew, 1981. Hawaiian Birdlife. Honolulu: University of Hawaii Press.
• Bingham, Hiram, 1847. A Residence of Twenty-One Years in the Sandwich Islands. Hartford: Hezekia Huntington; New York: Sherman Converse.
• Brigham, William, 1892. A Preliminary Catalogue of the Bernice Pauahi Bishop Museum of Polynesian Ethnology and Natural History, Part I., Kahilis, Feather Ornaments, Mats, and Kapas. Honolulu: Bishop Museum Press.
• ——1899. Hawaiian Featherwork. Bishop Museum Memoirs. 1(1):21-4.
• Buck, Peter (Te Rangi Hiroa), 1927. The Material Culture of the Cook Islands (A itutaki). New Plymouth: Thomas Avery and Sons (for the Board of Maori Ethnological Research).
• ——1957. Arts and Crafts of Hawaii. Special Publication 45. Honolulu: Bernice P. Bishop Museum.

• Chinen, Clarice, MS 1980. The Kahili: Royal Insignia of Hawaiian Royalty, [photocopy of paper for B. Harger, University of Hawai'i, Mānoa.] Department of Anthropology, Bishop Museum, Honolulu.
• Cook, James, 1784. A Voyage to the Pacific Ocean for Making Discoveries in the Northern Hemisphere in the Resolution and Discovery, Volume II. London: Nicol and Cadell.

• Davies, John, 1851. A Tahitian and English Dictionary. Tahiti: London Missionary Society's Press.

• Ellis, William, 1782. An Authentic Narrative of a Voyage Performed by Captain Cook and Captain Clerke, in His Majesty's Ships Resolution and Discovery During the Years 1776, 1777, 1778, 1779 and 1780. London: Robinson, Sewell, and Debrett.
• Emerson, Nathaniel, 1895. The Bird-hunters of Ancient Hawaii. Thrum's Hawaiian Almanac and Annual, 1895:101-11.

• Handy, Edward and Elizabeth Handy, 1972. Native Planters in Old Hawaii: Their Life, Lore, and Environment. Bernice P. Bishop Museum Bulletin, 233: 287.

- 302

• HEN, MS 1922. Kahilis. Lahilahi Webb Collection of Notes, data from Lucy K. Peabody, January 22, 1922, translated by Mary Kawena Pukui. Hawaiian Ethnological Notes (HEN), Vol. I. [typescript.] Archives, Bishop Museum, Honolulu.
• Henshaw, H., 1902. Complete List of the Birds of the Hawaiian Possessions, With Note on Their Habits. Thrum's Hawaiian Almanac and Annual, 1902: 54-106, 1903: 113-45.

• James, Helen and Storrs Olson, 1991. Descriptions of Thirty-two New Species of Birds from the Hawaiian Islands: Part II. Passeriformes. Washington, D.C.: American Ornithologists' Union.
• Judd, Laura, 1928. Honolulu Sketches of the Life Social, Political and Religious, in the Hawaiian Islands from 1828-1861. [Reprint of the 1880 edit.] Honolulu: Honolulu Star-Bulletin.

• Kamakau, Samuel, 1961. Ruling Chiefs of Hawaii. Honolulu: Kamehameha Schools Press.
• Keauokalani, Kepelino MS 1859. Short Notes Pertaining to Hawaiian Life, Descriptions of Hawaiian Birds, Paper II. Hai Manawa and other Catholic papers. [typescript] Archives, Bishop Museum, Honolulu.

• Loomis, Elisha MS 1825. Copy of the Journal of E. Loomis. Hawaii Mission Children's Society, Honolulu.

• Malo, David, 1951. Hawaiian Antiquities (Moolelo Hawaii) trans. Nathaniel B. Emerson. Honolulu: Bishop Museum Press.
• Munro, George, 1944. Birds of Hawaii (rev. edit.). Vermont: Bridgeway Press.

• Nordhoff, Charles, 1974. Northern California, Oregon, and the Sandwich Islands. Ten Speed Press.

• Olson, Storrs and Helen James, 1991. Descriptions of Thirty-two New Species of Birds from the Hawaiian Islands: Part I. Non-Passeriformes. Washington, D.C.: American Ornithologists' Union.

• Peale, Titian, 1848. United States Exploring Expedition, 1838-1842, Volume 8, Mammalia and Ornithology. Philadelphia: C. Sherman.
• Perkins, R., 1903. Fauna Hawaiiensis, Volume 1, part 4: Vertebrata. A ves. Cambridge: Cambridge University Press.
• Phelps, Steven, 1976. Art and Artefacts of the Pacific, Africa, and the Americas, The James Hooper Collection. London: Hutchinson.
• Pukui, Mary and Samuel Elbert, 1971. Hawaiian Dictionary. Honolulu: University of Hawaii Press.

• Richards, Mary, 1970. The Hawaiian Chiefs' Children's School. Rutland, Vermont and Tokyo: Tuttle.
• Rose, Roger, MS 1991. Bishop Museum Standing Kāhili Working Master List. Department of Anthropology, Bishop Museum, Honolulu.
• Rothschild, Walter, 1893-1900. The Avifauna of Laysan and the Neighboring Islands. London: R. H. Porter.

• Scott, Winifred, 1930. The Story of the Kahili. Paradise of the Pacific, 43(12): 16.
• Stewart, C. S., 1830. Journal of a Residence in the Sandwich Islands, During the Years 1823, 1824, and 1825. London: Fisher, Son, and Jackson.

- 303

• Titcomb, Margaret and Wayne Gagné, 1976. List of Hawaiian Bird Names. 'Elepaio, 36(10):117-26.

• Webb, Elizabeth, n.d. Editorial Correspondence. A lameda Times Star (California). [undated newspaper clipping of an interview with E. Lahilahi Webb, Archives, Bishop Museum, Honolulu]
• Wetmore, Alexander, 1925. Bird Life Among Lava Rock and Coral Sand: The Chronicle of a Scientific Expedition to Little-Known Islands of Hawaii. National Geographic Magazine, 48(1):86.
• Wilson, Scott B., and A. H. Evans, 1890-9. A ves Hawaiiensis. The Birds of the Sandwich Islands. [Reprint of the 1890-1899 edit. published by R. H. Porter, London.] Reprint Edition 1974 by Arno Press Inc., New York.

- 304 Page of endnotes