¹ For the distribution of shell-appliqué ornaments in Oceania, see Reichard, Gladys A., Melanesian Design, New York, 1933, pp. 88–124 and plates 93–125, 128–130, 134–139.Google Scholar

² The two kapkaps illustrated in Figs. 1 and 2 (the latter, as often happens, with its fretwork broken) are the most fully developed specimens of their type known to me. Three additional examples of the type—two in Sydney, Australian Museum, A 14226 and A 19744, and one illustrated by Edge-Partington, James, An Album of the Weapons, Tools, Ornaments, Articles of Dress, etc., of the Natives of the Pacific Islands, Manchester, 1890,Google Scholar iii, pl. 44, fig. 2—appear to be debased variants of the ornithomorphic norm evidently represented by our two examples. Probably related to the type, but badly debased, is a specimen in Berlin illustrated by Reichard, 1933, pl. 103, fig. 3 29. For a kapkap from Nissan in which four highly conventionalized birds appear to converge upon the center, see Krause, Fritz, “Zur Ethnographie der Insel Nissan,” Jahrbuch des städtischen Museums für Völkerkunde zu Leipzig, vol. 1, 1906, fig. 28.Google Scholar

³ Cf. Luschan, Felix von, Beiträge zur Völkerkunde der Deutschen Schutzgebiete, Berlin, 1897, pl. 37, figs. 9, 10, and p. 81Google Scholar a. These designs are identified as frigate birds by Nevermann, Hans, Admiralitäts-Inseln (Ergebnisse der Südsee-Expedition 1908–1910, edited by Thilenius, G., II. Ethnographie, Melanesien, A., Band 3), Hamburg, 1934, p. 393.Google Scholar Actually, many more designs can be found in Admiralty Island art which so clearly represent the frigate bird that a native name is hardly necessary for identification.

⁴ For additional examples of the Ceramese oiale, see Visser, Herman F. E., “Over Ornamenckunst van Seran,” Koloniaal Institute te Amsterdam, Mededeeling No. IX, Afdeeling Volkenkunde No. 3, Volkenkundige Opstellen I, Amsterdam, 1917, pp. 93–104,Google Scholar 14 pls.; and Schuster, Carl, “Dr. Carl Schuster on Bird-Designs in the Western Pacific: Indonesia—Melanesia—Polynesia,” Cultureel Indië, vol. 1, Leiden, 1939, pp. 232–235.Google Scholar See also Figs. 9 and 10 of the present article. It should be said here that our comparison of Ceramese and Melanesian designs is supported by an extensive and notable similarity in cultural institutions between the two areas. Cf. Duyvendak, Johan Philip, Het Kakean-Genootschap van Seran (Leiden thesis), Almelo, 1926;Google Scholar and Deacon, Arthur Bernard, “The Kakihan Society of Ceram and New Guinea Initiation Cults,” Folk-Lore, vol. 36, 1925, pp. 332–361CrossRefGoogle Scholar.

⁵ Cf. Schuster, 1939. citing Aalst, H. Krayer van, Liefde-Macbt, den Haag, 1915, pp. 10, 54.Google Scholar

⁶ See, for example, our Fig. 10, where the forked tails are represented at the outside; and compare the oiale illustrated on the cover of Krayer van Aalst's book, 1915, in which the forked tails extend inward so as to touch the central disk—in such fashion that they suggest the beginning of a lozenge-shaped diagram like that of our Admiralty Island kapkaps.

⁷ The question may be raised whether in both instances we do not have to do with the confused reminiscence of an original double head. In another Melanesian archipelago, that of the Solomon Islands, the double-headed bird is a definite entity in native design; and there are considerable grounds for believing that the four birds in this type of composition were originally double-headed birds.

⁸ This object was collected in 1936 by a New Zealand missionary, A. H. Voyce, at the village of Ilitopan (Iltapan, Yeltupan) at the northern extremity of Buka, the northernmost large island of the Solomon group. In a letter of May 29th, 1937, Mr. Voyce wrote me: “Parkinson referred to such ornaments on North Bougainville and Buka during the last century, but as I am a very keen collector of ethnographical material of that sort, and during more than eleven years’ residence had not seen or heard of any such ornaments, I thought that any that had existed must surely have been traded from the Southern Solomons, where they are fairly common.”

The passage referred to by Mr. Voyce is evidently Parkinson, R., Dreissig Jahre in der Südsee, Stuttgart, 1907, p. 492Google Scholar, reading: “Other ornaments [i.e., ornaments other than those of the kapkap type, in two layers] consist of perfectly round, but sometimes elongated, plaques of tridacna shell with the engraved design of a stylized frigate bird; these plaques, called kini, which are made on Buka and Bougainville, are very highly regarded and are produced only by a few artists.” Parkinson then refers to Meyer, A. B. and Parkinson, R., Album von Papua Typen, Dresden, 1894, ii, pl. 45Google Scholar, showing a native wearing a neck-pendant of tridacna shell with engraved design. This pendant, however, is nothing like the paparah, our Fig. 5, but unmistakably represents a type of shell ornament well known from the Southern Solomons, particularly from Malaita (e.g., Ivens, W. G., Me lanes tans of the South-East Solomon Islands, London. 1927,Google Scholar pl. 15, top and right; Reichard, 1933, pls. 126, 127). It is thus likely that the piece worn by the Bougainville native in Meyer and Parkinson's Album was, as Mr. Voyce suggests, traded to Bougainville from the Southern Solomons. The same conclusion is reached by Emil Stephan and Graebner, Fritz, Neu-Mecklenburg, Berlin, 1907, p. 190,Google Scholar who attribute the piece mentioned by Parkinson and illustrated by Meyer and Parkinson to the island of Malaita in the Southeast Solomons. Nevertheless, it is still possible that our paparah represents the round (rather than the pointed oval) plaques referred to by Parkinson. However this may be, the writer has not seen any specimen even remotely resembling that of our Fig. 5 in any European, American, or Antipodean collection.

⁹ Wheel-like motives of this type are common in certain classes of Western Melanesian design; and in the Solomons, at least, they seem to have a specifically solar connotation. Cf. Thurnwald, Richard, Forschungen auf den Salomo-Inseln und dem Bismarck- Archipel, Berlin, 1912,Google Scholar supplement, pl. 12, figs. 145, 146, 148; and especially vol. i, p. 537, where a motive of this type (fig. 110) from Buin at the southern end of Bougainville, is designated a “kreisförmiges Sonnenbild,” evidently in translation of a native term, bobotanke ru.

¹⁰ An alternative possibility which should perhaps be investigated is whether the designer could have had in mind a masked dancer.

¹¹ These scallops perhaps have their equivalent in the two zones of scallops in the Ceramese oiale. Fig, 3.

¹² The fact that the Buka plaque has two peripheral bands, one of scallops and one of triangles, may eventually prove to be decisive in determining the time and place of its Asiatic origin. For the present, however, this double border provides us only with a puzzle. For, since the paparah faithfully reproduces so many features of ancient Chinese mirror designs, it would seem reasonable to suppose that its double border also reflects a corresponding feature in a Chinese prototype. However, a cursory examination of many Huai and Han mirrors illustrated in several standard publications leaves us with the impression that the scalloped border ends in the later Han dynasty, at about the same time that the saw-toothed border begins to appear. Evidently the two types of border are never combined on the same mirror, even though they seem to overlap somewhat chronologically. These circumstances suggest that if the paparah goes back to a single prototype in ancient Chinese mirrors, that prototype might be of the later Han dynasty, and might be one which is still archaeologically unknown. The writer regrets that the circumstances in which this article was written make it impossible for him to search exhaustively among the ancient Chinese material for the closest possible prototype of the paparah; but he hopes that others may take up the challenging task.

¹³ The selection of motives in Fig. 6, B to F, is not intended to represent a sequential series, either in terms of chronology or of evolution, but merely to suggest a general analogy, especially in the matter of circular “perforations,” with the Melanesian birdmotive, Fig. 6, A. Each of the five Chinese motives here shown is repeated at the four corners of a lozenge-shaped framework surrounding the central knob of a mirror. By the inclusion of the motive F, we wish to raise the question whether such motives, which are repeated in many variants four times radially around the centers of Huai and Han mirrors, often without a lozenge-shaped framework, may not likewise be derived ultimately from the figure of a bird. This is obviously a question requiring an extended study of ancient Chinese mirror designs, and involving also comparative considerations, upon which we cannot embark in the present place. (See the second paragraph in note 26).

¹⁴ A perforated or ring-like body is perhaps the chief attribute of the “Sunbird,” a motive to be studied in a larger publication by that title now in preparation by the writer.

¹⁵ See Salmony, Alfred, “On Early Chinese Mirrors,” Art in America, vol. 30, 1942, pp. 190–198;Google Scholar and “Chinese Metal Mirrors; Origin, Usage and Decoration,” Hobbies, The Magazine of the Buffalo Museum of Science, vol. 25, no. 4, April, 1945, pp. 36–45.Google Scholar

¹⁶ See the second paragraph of note 26, below. On the appliqué effect of Chinese mirror designs, cf. Bachhofer, Ludwig, A Short History of Chinese Art, New York, 1946, p. 51.Google Scholar

¹⁷ See Hoop, A. N. J. Thomassen à Thuessink van der, “De Praehistorie,” in Stapel, F. W. (editor), Geschiedenis van Nederlandsch Indië, vol. 1, Amsterdam, 1938, pp. 9–111Google Scholar; and compare Riesenfeld, as cited at the beginning of our text.

¹⁸ Hoop, 1938, pp. 61–93.

¹⁹ Goloubew, Victor, “L'âge du bronze au Tonkin et dans le Nord-Annam,” Bulletin de l'Ecole française d'Extrême-Orient, vol. 29, 1929, pp. 1–46;CrossRefGoogle Scholar and Przyluski, Jean, “Sur deux miroirs de bronze,” Revue des arts asiatiques, vol. 9, 1935, pp. 165–167.Google Scholar

²⁰ In this connection it may be of interest to consider a wooden club from Matty Island, off the north coast of New Guinea, published by Luschan, Felix von, “Neue Beiträge zur Ethnographie der Matty-Insel,” Internationales Archiv für Ethnographie, vol. 12, 1899, p. 125,Google Scholar fig. 7, which Parkinson, 1907, p. 424, regards as an imitation of an ancient Chinese sword; and, again, a Micronesian taro-spade of wood with the appearance of a metal sword, illustrated by Eilers, Anneliese, Inseln urn Ponape (Ergebnisse der Südsee-Expedition 1908–1910, edited by Thilenius, G., II. Ethnographie, Mikronesien, B., Band 8), Hamburg, 1934, p. 230, fig. 97.Google Scholar

²¹ Cammann, Schuyler, “Suggested Origin of the Tibetan Mandala Paintings,” The Art Quarterly. Spring, 1950, pp. 106–119.Google Scholar The author's argument centers around the observation that the animals of the four directions are represented on the mirrors in inverse order. Dr. Cammann has called my attention to the fact that- in our mirror, Fig. 4, the inscription (in itself banal) likewise reads backwards, i.e., in counter-clockwise rather than in the usual clockwise direction.

²² Cammann, , “The ‘TLV’ Pattern on Cosmic Mirrors of the Han Dynasty,” Journal of the American Oriental Society, vol. 68, 1948, especially p. 165;CrossRefGoogle Scholar and Cammann, , “A Rare T'ang Mirror,” The Art Quarterly, Spring, 1946, pp. 92–114.Google Scholar Cf. Schuster, , “Das Vogelmotiv in der chinesischen Bauernstickerei,” in Strzygowski, Josef, Spuren indogermanischen Glaubens in der bildenden Kunst, Heidelberg, 1936, feg. 278 and p. 334Google Scholar.

²³ An extensive study of carpet-motives and of cognate designs in other modern Asiatic popular traditions is planned for the publication mentioned in note 14.

²⁴ Though we cannot at the moment point to any obvious Melanesian or Indonesian counterpart for the dragons in the Asiatic four-bird mandalas, it is possible that certain perforated shell ornaments (hinuili) of New Georgia in the Solomons preserve a reminiscence of this arrangement. In these (as generally throughout Melanesia, where the frigate bird is represented) fishes take the place of dragons as prey of the birds. An extended consideration of the hinuili of New Georgia is planned for the publication mentioned in note 14.

²⁵ Sachse, F. J. P., Seran (Mededeelingen van het Bureau voor de Bestuurzaken der Buitengewesten bewerkt door het Encyclopaedisch Bureau, vol. 29), Weltevreden, 1922, p. 113.Google Scholar

²⁶ The custom of suspending the tjidako with its four-bird oiale from the highest point in the club-house of the secret society in Ceram, and the custom of suspending the Chinese mirror in the same way from the highest point in tombs and temples, suggests that the original function of the four-bird motive must have been that of a ceiling decoration. This inference is supported by the occurence of what is, in effect, a_ four-bird mandala painted on the ceiling of a (seventh century ?) cave-temple at Ming Öi near Qyzil in Chinese Turkestan. See the color-plate at the end of Grünwedel, Albert, Altbuddhistische Kultstätten in Chinesisch-Turkistan, Berlin, 1912.CrossRefGoogle Scholar At the four sides of the ceiling four double-headed birds seize eight serpents (a typical carpet design); in the center is a blank space obviously representing (in practical terms) a smoke-hole, or (in symbolic terms) an opening to the sky or replica of the “Sun-door.” (For architectural prototypes of this painted ceiling, see LeCoq, Albert von, Bilderatlas zur Kunst- und Kulturgeschichte Mittel-Asiens, Berlin, 1925, pls. 99–103, pp. 31–33;Google Scholar and Strzygowski, Josef, Asiens bildende Kunsr, Augsburg, 1930, pp. 152–155).Google Scholar

Having identified the “four-bird mandala” as a ceiling design intended to be applied around a central hole symbolic of the sky, we find the way open to consideration of a significant feature of modern Asiatic nomad dwellings—specifically of the appliqué of felt in the form of a collar which is tied around the outside of the smoke-hole on the typical Mongol yurt. While these smoke-hole collars obviously serve the practical function of holding the roof-felts in place, it cannot escape our notice that they are cut in a pattern essentially identical with that surrounding the knobs of innumerable Huai and Han mirrors. The fact that the ancient mirrors were used as magic markers for the centers of ceilings, and that they show unmistakable traces of derivation from an appliqué of some pliable material, suggests to the writer that we may have, in the smoke-hole collars of the Mongol yurts, the modern survival of an ancient functional form which ultimately inspired the mirrors themselves.

Such an hypothesis obviously implies many questions—not the least important of which is whether we are justified in deriving the form of the modern smoke-hole cover, and the form of the motives surrounding the centers of many Huai and Han mirrors of somewhat different types than those here illustrated, all from an original “four-bird mandala.” This is a difficult question which should, however, eventually be comtemplated. The symbolism of the Mongol smoke-hole “collar” has recently been studied by Cammann, Dr.. Cf. “The Symbolism of the Cloud Collar Motif” The Art Bulletin 33 (1951) 1–9.CrossRefGoogle Scholar