The Eskimo evidence proves that cat's-cradle may, in part, have a magical significance and suggests a line for future inquiry, for we know that all over the world strings, cords, and knots enter largely into magical practices. The information at present available is too scanty for us to discuss these questions with profit.

ALFRED C. HADDON.

CAMBRIDGE, ENGLAND, July, 1905.

STRING FIGURES

CHAPTER I

DISTRIBUTION OF STRING FIGURES-NATIVE NAMES-METHODS FIRST RECORDED-ETHNOLOGICAL VALUE OF STRING GAMES-RELATIONS BETWEEN FINISHED PATTERNS MADE BY DIFFERENT RACES-RELATIONS BETWEEN NATIVE METHODS HOW STRING FIGURES ARE MADE-DIFFERENT OPENINGSNOMENCLATURE ADOPTED-TYPICAL MOVEMENTS DESCRIBED-EXPLANATION OF THE DESCRIPTIONS AND THE DRAWINGS.

N our childhood we have all doubtless enjoyed the fascination of the game of Cat's-Cradle, and experienced a sense of being hopelessly baffled, when, after completing the series of familiar movements, we were at the end of our knowledge, and all our attempts to go on further ended in a complete tangle of the string. We did not know that the game, as we then played it, is one of a host of similar games played with a loop of string by savage or primitive people all over the world, and that, while our childish game is also known in many and widely separated lands, it is possibly only a survival of others now lost, and crude enough compared with the intricate and beautiful patterns devised by savage races.

For many years travellers have been calling attention to the fact that a game resembling our Cat's-Cradle is played in various parts of the world; hence we now have some idea of its geographical distribution.

We know that certain simple patterns are common in Great Britain, Europe, and the United States (in addition to the Indian games), and have been reported from India, Japan and Korea (Culin, 2, p. 30 and Weir) and China (Culin, 2, p. 30; Fielde, p. 87). Ehrenreich (p. 30) tells us that string games are played in South America, and I have found a few figures among the Batwa pygmies from Africa. Reports of their occurrence come chiefly, however, from explorers of the various islands of the Pacific Ocean, and from observers of the North American Indians. Thus we learn of string figures in Java from Schmeltz (p. 230); in Borneo from Wallace (p. 183), Haddon, and Furness; in Celebes from Matthes (p. 129); in the Philippines from my own studies at the St. Louis Exposition; in Australia from Bunce (p. 75), Smyth (Vol. I, p. 178), Eyre (II, p. 227), and Roth (p. 10); in New Guinea from Finsch (1891, p. 33), Rivers and Haddon (p. 151), and Turner (p. 483); in Torres Straits from Rivers and Haddon (p. 146); in New Ireland from Finsch (1888, p. 143); in the Solomon Islands and the New Hebrides from Codrington

(p. 341); in the Loyalty Islands from Rivers and Haddon (p. 148); in the Fijis from Buchner (p. 269); in New Zealand from Dieffenbach (Vol. II, p. 32), Taylor (p. 172), and Tregear (1, p. 115; 2, p. 58); in the Hervey group from Gill (p. 65); in the Hawaiian Islands from Culin (1, p. 223) and Brown (p. 163); in the Caroline Islands from Furness; and in the Marshall Islands from Stephen. In America we learn of their prevalence among the Eskimos from Boas (1, p. 229; 2, p. 85; 3, p. 569; 4, pp. 151, 161), Hall (1, p. 316; 2, p. 129), Klutschak (p. 138), Murdoch (p. 383), Nelson (p. 332), and Tenicheff (p. 153); among the Salish from Smith (p. 281); the Tlingits, Tsimshians, and Kwakiutls, from Boas (2, p. 85); the Clayoquahts from Haddon; the Tewas and Zuñis from Culin; the Pawnees, Omahas, and Cherokees from Haddon (5, p. 217); and among the Navahos from Haddon (5, p. 219) and Culin. Mr. John L. Cox has gathered games for me from the Klamaths, Tewas, Omahas, and Onondagas, and I collected string figures from the Navahos, Osages, Chippewas and Apaches at the St. Louis Exposition.

Of the name "cat's-cradle," which is confined, of course, to the English game, no satisfactory derivation has ever been given (see Murray, N.E.D.). Comparatively few of the native names have been recorded. The Eskimos of Cumberland Sound call it ajarar poq; the Navahos, na-ash-klo, "continuous weaving"; the Japanese, aya-ito-tori, “woof-pattern string-taking"; the Caroline Islanders, gagai (the word also employed for the pointed stick used to open coco-nuts); the Hawaiians, hei, or "net"; the New Zealanders, he-whai, huhi, or maui; the natives of Lepers' Island in the New Hebrides, lelagaro, and of Florida in the Solomon Islands, honggo. In the eastern islands of Torres Straits it is known as kamut, in the western islands as wormer. The Bugis and Makassars of the Celebes call it toêká-toêká, from toêká, 'a ladder." In some places in Australia it is named cudgi-cudgick. In North Queensland the various tribes of blacks have different names for it. The Kokoyimidir of Cape Bedford call it kápan (used also for "words, letters, writing," etc.); the Ngaikungo and Ngatchan of Atherton, etc., morkuru; the Nggerkudi of the Pennefather and Batavia rivers, ane-inga; the Kungganji of Cape Grafton, manjing; the Koko-lama-lama of the Hinterland and coast of Princess Charlotte Bay, yirma; the Koko-rarmul of the same, mianman; the Koko-wara of the same, andaiibi; the Mallanpara of the lower Tully River, kumai or kamai.

66

Although the existence of the game has been known for years, no one had described how the figures are made until Dr. Boas, in 1888, recorded the methods employed in two Eskimo games. In 1900 Mr. Harlan I. Smith figured certain stages of two games played by the Salish Indians of British Columbia. To Dr. W. H. R. Rivers and Dr. Alfred C. Haddon, however, must be attributed the real impetus given to the study of string games: their paper, published in 1902, gives us the first plan whereby all these intricate and difficult figures may be described so that anyone can repeat them; their simple and accurate nomenclature now makes it possible to record all future discoveries. A second paper, in 1903, by Dr. Haddon (5) on the string games of the American Indians, and a paper, in 1903,

by the Rev. John Gray on several Scotch cat's-cradles, include, probably, all the descriptive records on the subject.

My brother, Dr. William Henry Furness, 3rd, in his recent trip among the Caroline Islands, by following Drs. Rivers and Haddon's directions and nomenclature, was able to record fifteen new and extremely interesting string figures; and at Dr. Haddon's suggestion I visited the St. Louis Exposition several times in 1904, and was fortunate enough to secure thirty-one additional games from natives of the various races and tribes there congregated. Of the ninety-seven figures set forth on the following pages seventy-one are now described for the first time. To this list have been added drawings of a number of finished patterns obtained by other observers who did not record, however, the methods by which they were made.

Just what value the study of the string games of different races will have to the ethnologist, it is difficult to say at this time. That evidences of racial or tribal relationship, or of migration, may be found in them, is not unlikely; that they bring us in closer touch with the folk-lore of savage people is already clear. While games in general of native races, and their connection with folk-lore, have by no means been neglected, string figures have appeared so difficult and require relatively so much time and such intimate relations for their collection, that, as yet, few careful observations have been made. Gradually, however, we are learning more about them; we know that many are closely connected with racial history and mythology, with traditional tales and fortune-telling; some are accompanied by muttered chants or songs; in others a consecutive story follows from movement to movement, or perhaps a touch or a word is associated with a certain turn or twist of the string.

Concerning the relations, which the finished patterns produced in the string games of different countries, bear to one another, we know that a few simple figures are practically universal, that several others are formed by widely separated races, but that the great majority are peculiar to definite localities. We cannot suppose that the natives set to work deliberately to form figures of familiar objects, but rather that of the many patterns-formed by chance, in sheer idleness or from an inventive turn, whether under tropical suns or in ice huts during long arctic winters-only those were kept up and named which bore resemblances, however slight, to something connected with their daily life or prominent in their thought. How far tradition has preserved the figures unchanged, or time and constant repetition have altered their original form, of course it is impossible to tell. In the finished patterns we find, among all races, representations of men and women, parts of the body, articles of dress, of commerce, and of warfare; and of stars, and natural phenomena such as storms, darkness, and lightning. Animals and plants are frequently reproduced, the names of course being conditioned by the fauna and flora. of the locality, as, for example, the coral of the Pacific Islands, the cariboo of the Eskimos, and the owl, snake, and coyote of the American Indians.

The methods employed by different races in making the figures and a com