bark rolls. Quipus were used by the Peruvians for a similar purpose (cf. Garcilasso de la Vega, Royal Commentaries, Eng. tr., London, 1869-71, bk. iii. ch. xxvii.), and in both Peru and Mexico there appear to have been bards whose business it was to transmit the national song. The germ of this institution appears in many a tribe where rites and mysteries are transmitted in traditional songs. Since the study of Indian music has become serious, many records have been taken, either by transcription into the European notation or by the use of the phonograph.

3. Song.-Indian music is primarily vocal. To be sure, melodies are played upon the native flute, and Garcilasso is authority for the statement that the Peruvians knew a kind of ensemble performance of flutes or pipes; but, even where so used, the flute seems to be regarded as merely a sub

stitute for the voice.

"They played upon these flutes,' says Garcilasso (bk. ii. ch. xxvi.), 'airs of which the words were rimed, composed in a spirit of gallantry upon the rebuffs or favours received from their mistresses.' In N. America this instrument is commonly called the 'lover's flute,' since its music is regarded as a substitute for vocal song from the bashful lover (cf. Burton, American Primitive Music, pp. 83-86). While their music is thus vocally conceived, the Indians generally appear to have no clear conceptions of the difference between the verbal and tonal materials of their songs.

To the Ojibwa, says Burton (loc. cit.), whatever departs from plain prose is nogamon, song, which means that his poetry is not only inseparable, but indistinguishable from music.' Burton found difficulty more than once in convincing an Indian who sang new words to an old melody that new music was not being offered. But the words themselves need not be significant apart from the air. Frequently they are meaning. less, sometimes because the original meaning is forgotten, sometimes because archaic or foreign, sometimes because originated as mnemonic syll. ables or ejaculatory refrains, or as pure nonsense. It is the common thing for an Indian song to require a story to explain its meaning. In short, the verbal elements of Indian music form a kind of notation, dimly analogous to our do re mi method of denominating the scale.

Frances Densmore, writing of the Midé society songs, says:

The songs of the Midé represent the musical expression of religious ideas. The melody and the idea are the essential parts of a Midé song, the words being forced into conformation with the melody. To accomplish this it is customary to add meaningless syllables either between the parts of a word or between the words; accents are misplaced, and a word is sometimes accented differently in various parts of a song; the vowels are also given different sounds, or changed entirely. . . . The writer has even been informed that it is permissible for different members of the Midéwiwin holding high degrees to use slightly different words for the songs, but the idea of the song must always remain the same. The words serve as a key to the idea without fully expressing it. Sometimes only one or two words occur in a song. Their literal translation is meaningless, but to an instructed member of the Midéwiwin they bear an occult significance. Many of the words used in the Midé songs are unknown in the conversational Chippewa of the present time' (Bull. 45 BE, p. 141.).

Le Jeune's Relation of 1634 contains an interesting account of Montagnais music (Jesuit Relations, ed. R. G. Thwaites, Cleveland, Ohio, 1896-1901, vi. 182 ff.), indicating the same general character:

"They use few words in singing, varying the tones and not the words.... All their religion consists mainly in singing. Not one of them understands what he is singing, except in the tunes which they sing for recreation,' etc.

Interesting too, though chiefly as a token of the fundamental identity of N. and S. American conceptions, is Garcilasso's naive comment on Inca music: Each song has its particular air, and they could not sing two to the same tune; for the lover who would serenade his mistress indicated the state of his passion upon his flageolet by the diversity of the sound, sad or gay, making known to his beloved the joy or sorrow of his heart. But were he to give two different songs to the same air, each would be confused,

and the gallant defeated in his effort to make his feeling known (Royal Commentaries, bk. ii. ch. xxvi.).

4. Mode of composition. — Aristotle's remark that poetry first takes form in hymns to the gods and personal lampoons seems to have a fair illus Various writers tration in native American song. have noted the Eskimos' love for contests in satirical song, with which they while away their long winter evenings; while satire and lampoon are not uncommon in Indian song. These songs naturally vary indefinitely with occasion and mood. On the other hand, religious song (and, as Le Jeune says, Indian religion is mainly song) is strongly conservative, even while it marks almost every phase of native life.

Advocates of the communal origin' of poetry find good materials in the recreative type of song, but the evidence in the case of the more serious compositions points in another direction. Songs which have to do with tribal traditions, with rites and ceremonies, with love and death, and even cradle-songs, while they tend to assume a traditional form and come to be sung in chorus, in many if not all cases owe their composition to the stress of an inspirational moment on the heart of the individual composer (cf., e.g., the accounts of the origins of the Hako and Midé songs in 22 RBEW ii. and in Bull. 45 BE). Song, with the Indians, is, in fact, much more than verbal music; it is a part of life itself, and is efficacious in altering life's destinies. They use their songs, says Le Jeune (loc. cit.), for a thousand purposes; they sing in sickness and in health, at the feast, and in peril and suffering:

'During the time of our famine I heard nothing throughout these cabins, especially at night, except songs, cries, beating of drums and other noises; when I asked what this meant, my people told me that they did it in order to have a good chase, and to find something to eat.'

Burton tells of an old pagan whom he could not persuade to sing hunting-songs out of season. The priest who gave Alice Fletcher the Hakowhich is a ceremonial prayer for life-told her that it must not be sung in winter, but only in spring, summer, or autumn, when life was stirring; similarly, the several songs must be sung at suitable times :

'Sometimes the songs of the nest and the wren are sung early in the day, as these songs were made in the morning. The song of the owl must be sung toward night' (22 RBEW ii. 168). tide over the exigencies of life and invented to The song is, in short, a kind of spell, helping to meet them.

The Navaho myth of the man who discovered the use of corn as food begins: A man sat thinking, "Let me see; my songs are too short; I want more songs; where shall I go to find them?" Hasjelti, the white-maize spirit, appeared to him, and led him to a country where he learned the use of corn and the appropriate songs, which he brought back to his own people (8 RBEW (1891), p. 278).

An interesting feature of Indian song is the sense of personal proprietorship which attaches to compositions. Burton states it thus:

'A has no right to sing B's songs; B did not compose them, but they came down to him through his family, or from some chief who taught him, and B alone should say whether they So in Le Jeune's Relation of 1636: might be given to another' (p. 118 f.).

Each has his own song, that another dare not sing lest he give offense. For this very reason they sometimes strike up a tune that belongs to their enemies, in order to aggravate them' (Jesuit Relations, ix. 111).

The personal song is evidently a portion of the 'medicine' which every Indian owns, and it is potent in the same way. A special type of song comprises the spontaneous melodies coming in dream and dance, to which the Indian attaches especial significance as revelations of superhuman power, many of them being associated with some animal, revealed as the tutelary of him to whom the song is given. Perhaps the most picturesque and affecting aspect of Indian life is the death

song with which every Indian seeks to face his end, sometimes composed upon the spot, sometimes prepared in advance; if possible, every Indian dies singing; and the breath of life goes forth to the spirit-world as a breath of song.

5. Structural traits.-The study of the structure of Indian music, in spite of the serious attention devoted to it, is hardly more than begun. From investigations already made it is evident that there are many types of Indian music, perhaps as many as there are linguistic stocks, each with its structural peculiarities. On certain points, however, there is an approach to agreement, and these we may broadly summarize.

his

(a) From the Arctic southward, American music is primarily (though not exclusively) drum-song. The voice and the drum are the Indian's instruments, and his music is a kind of concord of the two; the drum is the instrument of his rhythmic, the voice of his melodic, expression. But his drum-rhythms are not primarily guides to the rhythms of the voice; rather they are the rhythms of movement, of the dance, to which the rhythms of vocal utterance may, or may not, conform. White observers have been astonished at the apparent duality of the Indian's time-sense — drum being beaten to one time, his song sung to another time, and this even when the drum is obviously used as accompaniment. In sixty-three out of ninety-one Midé songs recorded with the drum accompaniment, Frances Densmore found the metric units of voice and drum to differ from each other. In one instance the metric units of voice and drum were so nearly alike that the same metronome indication was used for each; at the beginning the drum-beat was slightly behind the voice, but it gained until with the voice, and then gradually passed it. Normally, when drum and voice have the same metric unit, they are slightly in sequence in beat (Bull. 45 BE, p. 5f.).

'I have heard Indian music wherein the conflict between

voice and drum was much more marked than is the case be

tween 3-4 and 4-4' (Burton, p. 46).

The voice is sometimes made to mark its own time by rhythmic pulsations (cf. 22 RBEW ii. 282); while, again, both voice and drum employ rhyth mic accents, or beats, to mark the metric units. The vocal rhythm frequently changes even in the course of a brief song, but the recurrence of rhythmic figures, or motives, serves to give outline and unity to otherwise formless compositions.

(b) The question of the range of tone-material and of the existence of scales in Indian music is still moot. Early observers, struck by the strangeness of the intervals which they seemed to detect in Indian singing and by the constancy with which these intervals were repeated, jumped to the conclusion that the American Indian possessed an ear of unusual delicacy of discrimination, and scales of a refinement which the white man's could not duplicate. More careful studies have led to the reverse view, that Indian music is built upon scales in the making, or even that it is essentially adiatonic. The latter is the view of Hopi music adopted by Gilman. He characterizes Pueblo song as a kind of rote-song having no fixed intervals:

'The singer's musical consciousness seems restricted to a few intervals of simplest vibration ratio approximately rendered, and to melodic sequences formed by their various analysis and synthesis and rendered with a certain loose fidelity. If a scale were in his mind, even dimly, it should make itself known in a more uniform interval production and in a more impartial use of the tones continually at hand in the fancy. The hearer seems witness to a wholly strange method of musical thought and delivery. The total complex of tone, timbre, and articulation -doubtless at times movements, and other noises also-moves on apparently without guidance by any vanguard of fancied tones at fixed intervals Hopi Songs,' in Journal of Am. Eth. and Arch. v. [1908] 5).

Other authorities do not go to this radical extreme, but credit the Indian with an adumbrate

consciousness of a scale or scales, which his lack of musical standards prevents him from clearly finding, or, if found, from holding. Burton says: 'The Ojibways recognize all the intervals of our major dia. tonic scales, but the fourth and the seventh rarely occur in the and he indicates a pentatonic major and minor and same song' (p. 41), a hexatonic major and minor scale as in use among them. Frances Densmore's study of the music of the same people is in substantial agreement with this. A similar view-interesting, as coming from S. America-is taken by F. J. de Augusta with respect to Araucanian song (Anthropos, vi. [1911] 685 f.). Alice Fletcher says of the Omaha music: The octave is seemingly the one fixed interval. The songs are not built on any defined scale. Nevertheless, the Omaha have and by their recognition of good singers' (27 KBEW, p. 374). 'a standard of musical tones,' as witnessed by their drum-tuning Another interesting observation is that the Omaha object to the rendering of their songs on the piano as unsupported arias; when rendered with harmonic chords, 'That sounds natural,' was their comment. Burton found the Ojibwa lively in their appreciation of harmonizations of their melodies. This aptitude of the native ear for the appreciation of harmony seems to lend some colour to the supposition that they do actually (even if imperfectly) think their music in harmonic intervals. the predominance of downward progressions in (c) A third structural feature deserving note is their melodies; like the ancient Greek, the Indian thinks his music mainly in the descending order. Melodic form is fluid and undeveloped, but repetition, balance, and antithesis give form to their better pieces. Further, uncertainty of form is partially offset by definite conventions of execution, amounting to artistry. The accomplished singer affects a vibrato, or wavering of the voice, ornamentation of grace-notes. In many of the a drawling tone, a portamento or slurring, an Omaha societies a fine was imposed if a member made mistakes in singing (A. C. Fletcher, 27 RBEW, p. 373). Finally, the inner harmony of a general tonality is set by the mood of the composition, or perhaps keyed by the incessant drum, whose drone serves as a tonal background for the melodic broidery.

6. Adaptations.-No account of Indian music would be complete without mention of the use of themes caught or imitated from Indian melodies as the foundations of artistic compositions. These vary from simple harmonizations of Indian songs (many of which have been published by the WaWan Press, Newton Center, Mass.) to elaborate symphonic compositions. The most notable works of this nature yet achieved are doubtless Dvořák's New World Symphony,' MacDowell's 'Indian Suite,' and Coleridge-Taylor's 'Hiawatha' music.

LITERATURE.-F. R. Burton, American Primitive Music, New York, 1909, is a careful and readable introduction to the subject; it contains harmonizations of twenty-eight Ojibwa songs. N. Curtis, The Indian's Book, New York, 1907, contains melodies from some eighteen tribes. Thorough studies are F. Densmore, 'Chippewa Music,' forming Bulletins 45 and 53 BE (1910, 1913), and B. I. Gilmore, Zuñi Melodies' and Hopi Songs,' in Journ. of Amer. Ethnology and Archaeology, i. [1891), v. [1908]. Alice C. Fletcher, A Study of Omaha [1893], Indian Story and Song, Boston, 1900, The Hako, 22 Indian Music,' in Papers of the Peabody Museum, vol. i. no. 5 RBEW ii. [1904], 'The Omaha Tribe,' 27 RBEW [1911], and 'The Study of Indian Music,' reprinted from Proc. of Nat. Acad. of Sciences, i. [1915] 231 ff., contain indispensable material. Special mention may also be made of: J. Reade, 'Some Wabanaki Songs,' 'Aboriginal American Poetry,' Trans. Royal Soc. of Canada, v. ii. [1888]; W. Matthews and J. C. Fillmore, Navaho Music,' Mem, of the Amer. Folklore Society, 1897, pp. 255-290; F. Boas, The Central Eskimo,' 6 RBEW [1888], p. 409 ff.; W. J. Hoffman, 'The Mide'wiwin,' 7 RBEW (1891), p. 149 ff.; J. Mooney, The Ghost-Dance Religion,' 14 RBEW [1896] 653 ff.; S. Culin, Games of the North American Indians, 24 RBEW (1907); F. Russell, The Pima Indians,' 26 RBEW [1908], p. 3 ff.; F. G. Speck, 'Ceremonial Songs of the Creek and Yuchi Indians,' Anthropological Publications of the Museum of the Univ. of Pennsylvania, vol. i. no. 2 [1909]; A. T. Cringan, 'Description of Iroquois Music,' Archaeological Report of Ontario for 1898. T. Baker, Über die Musik der

nordamerikanischen Wilden, Leipzig, 1882, is one of the earliest scientific studies in this field. The Report of the U.S. National Museum for 1896, Washington, 1898, contains an account of native American musical instruments (pp. 561-664). Periodical articles of importance include: A Génin, 'Notes sur les danses, la musique et les chants des Mexicains anciens et modernes, REth iv. [1913] 301 ff.; F. Boas, 'Chinook Songs,' JAFL i. [1888] 220 ff., 'Songs and Dances of the Kwakiutl,' ib. p. 49 ff.; Washington Matthews, Songs of Sequence of the Navajos, ib. vii. (1894) 185 ff.; Alice C. Fletcher, Indian Songs and Music,' ib. xi. [1898] 85 ff.: E. Sapir, 'Song Recitative in Paiute Mythology, ib. xxiii. [1910] 455 ff.; W. Thalbitzer, 'EskimoMusik-und Dichtkunst in Grönland,' Anthropos, vi. [1911] 485 ff.; F. J. de Augusta, Zehn Araukanerlieder,' ib. 684 ff.; R. Lehmann-Nitsche, Patagonische Gesänge und Musikbogen,' ib. iii. [1908] 916 ff., containing a discussion of the origin of stringed instruments in use by American Indians; E. Fischer, 'Patagonische Musik,' ib. p. 941 ff. H. B. ALEXANDER. MUSIC (Babylonian and Assyrian).-As the inscriptions seem to furnish no easily recognizable indications of a musical notation, very little can be gathered concerning the music of these nationalities (see, however, below, § 6). The representations of musical instruments in their bas-reliefs, however, make it clear that they not only loved music, but had also made some progress in the art -progress fostered originally, in all probability, in their temple-ritual.

1. Its antiquity in Babylonia.-As their signlist implies that the Sumerians (the non-Semitic inhabitants of Babylonia) had musical instruments before they settled in that tract, they must have had music in practically pre-historic times. It is true that G. A. Barton, in his Origin and Development of Babylonian Writing (=BASS ix. [1913] pt. i.), gives (p. 186) only three forms, of which the third (balag) seems to be merely a variant of the second (tigi); but this number could probably be increased. The three characters in his list express various ideas derived from music, such as 'to be joyful,' 'to strike,' etc.

2. Musical instruments.-These were sufficiently numerous, judging from what we find in their sculptures, and may be classified as follows.

i. INSTRUMENTS OF PERCUSSION.-(a) Drum (Sumerian ála, Semitic álu, according to Langdon, though leu would seem also to have been used).The earlier forms were very large-indeed, one relief (about 2500 B.C.) shows an instrument nearly as high as the man who, with his left hand, seems to be striking it (B. Meissner, Grundzüge der altbabylonischen Plastik, Leipzig, 1914, p. 14). It looks like an enormous tambourine resting edgewise on the ground, and was probably moved from place to place by rolling. For the later (Assyrian) portable forms see G. Rawlinson, The Five Great Monarchies, London, 1862–67, ii. 160.

(b) Tambourine (Sum. liliz, Sem. lilisu, according to Langdon; tabbalu [F. Thureau-Dangin, Huitième Campagne de Sargon, Paris, 1912, p. 26 f., 1. 159]. The forms shown naturally do not differ much from the modern instruments, but the rim was in some cases not provided with rings or disks (see Rawlinson, ii. 158; E. de Sarzec, Découvertes en Chaldée, Paris, 1884-1912, pl. 39, no. 5).

(e) Cymbals.-Late forms are given in Rawlinson, ii. 158f. Those with rounded backs have a loop, those with conical backs a fixed rod, by which to hold them.

ii. STRINGED INSTRUMENTS.-(a) Harp.-The earliest example (c. 2500 B.C.) is probably that in de Sarzec, pl. 23, which shows considerable development, and has 11 or 12 strings. An Assyrian harp with 16 strings is shown in Rawlinson, ii. 153, where there is also an Elamite harp of 19 or 20 strings. According to Langdon, this was the Sum. balag, Sem. balaggu or balangu. Thureau-Dangin renders the group gis zag-sal by harp.' In WAI ii. 44, 55, gis zag-sal-sisa is rendered by išar[tu], 'the straight'. a word translating sisa, and therefore qualifying gis zag-sal.

(b) Dulcimer.-A primitive form of this (? before 2500 B.C.) is shown in a relief, originally inlaid, found at Bismaya (E. J. Banks, Bismya, New York, 1912, p. 268). It had 5 or 7 strings and was played with a plectrum. For the Assyrian form see Rawlinson, ii. 161; A. H. Layard, Monuments of Nineveh, London, 1849, i., plates 12, 22 (held with a downward slope), and 73. In the first case the musicians welcome Aššur-naşir-pal on his return from a bull-hunt, while the last celebrates the capture of the Babylonian city Dilmu (Dailem) by Sennacherib.

(c) Lyre and cithara.-Though these must have been among the earlier stringed instruments, the extant pictures are mainly of a late date (see Rawlinson, ii. 154-155, 158). Being rather rectangular in form, they did not, apparently, give a wide range of notes. The number of strings of a lyre varied from 4 to 10 (see Rawlinson, ii. 154, also 153, where a 4-stringed lyre similar to a trigon' is shown). A band of foreign musicians (captives) is shown in Rawlinson, ii. 164 = Brit. Mus., Assyrian Saloon, no. 14. Notwithstanding their likeness to the Israelitish captives in Sennacherib's Lachish reliefs, it is doubtful whether they were of that nationality, since those sent by Hezekiah were not captives. That the yoke or cross-piece of their instruments terminates in swans' (?) heads, however, would probably present no difficulty.

(d) Guitar. This rather resembles the Egyptian nefer, or banjo, which may have been borrowed from Babylonia (F. W. Galpin, in Stainer, Music of the Bible, p. 45). This was known in Babylonia at a very early date (ib. pl. i. B). The form is similar to that used later in Assyria (Rawlinson, ii. 156). iii. WIND INSTRUMENTS.-(a) Pipe, whistle, or flute.-This probably became a favourite instrument after the arrival of the Sumerians in Babylonia, where reeds are plentiful. The common Sumerian word for the pipe or flute is gi-gid, 'long reed,' or gi-sir, 'reed of music,' in Semitic melilu, from elelu, to make a joyful sound.' According to the Descent of Istar into Hades, rev., lines 49 and 56, this was the instrument of Tammuz, who possessed one made of the sacred lapis lazuli. When Tammuz joyfully played thereon, the mourners, male and female, played with him, and then the dead might arise and enjoy the incense. The flutes depicted on the Assyrian monuments are played in pairs (Rawlinson, ii. 157). The Elamite double flute was similar to the Assyrian (Layard, ii., plates 48, 49).

(b) Trumpet.-Pictures of this instrument are rare, the most noteworthy example being that in the hands of a director of works in Sennacherib's time (Layard, ii. 15-Brit. Mus., Nineveh Gallery, 51, 52). It was probably not a speaking-trumpet, but an instrument conveying orders by its note.

(c) Horn.-Though the inscriptions do not seem to refer to this as a musical instrument, it is probable that they used it as such. The Sumerian for 'horn' is si, Sem. qarnu, Heb. geren.

(d) Other wind instruments. According to Langdon, the Sum. tigi, Sem. tigû, was an ordinary reed-flute, and something similar is to be understood by the gi-erra-qan bikiti, or 'reed of wailing.' The gi-di, Sem. takaltu, and the Sem. pagû, he suggests, may be bagpipes.

3. Singers. Of these there were many classes, but we have still much to learn about them. Some were probably for secular performances only.

(a) Náru (fem. nártu). These are expressed by the character lul, preceded by the determinative for 'man' or 'woman.' Lul was possibly the picture of a harp or lyre, to which they seem to have sung. According to Sennacherib, a choir of these, male and female, was sent to Nineveh by